Proceedings of The 35 International Symposium CIB W062 on Water

Proceedings
of
The 35th International Symposium
CIB W062
on
Water Supply and Drainage for Buildings
7-9th September 2009
Düsseldorf, Germany
Organized by
Faculty of Building Services and Environmental Engineering
Fachhochschule Gelsenkirchen, University of Applied Sciences
Gelsenkirchen, Germany
Editors
Prof. Dr.-Ing. Mete Demiriz
Nina Jablonski
Robert Schütz
Fachhochschule Gelsenkirchen, University of Applied Sciences
Publisher:
Fachhochschule Gelsenkirchen, University of Applied Sciences
Prof. Dr.-Ing. Mete Demiriz
Neidenburger Str. 43
D-45877 Gelsenkirchen, Germany
ISBN 9783980 723954
Disclaimer
The organizers and the editors assume no responsibility whatsoever for the
accuracy, completeness or usefulness of the information contained in these
proceedings. The authors alone are responsible for the contents of their
papers.
Publication Date: 7th September, 2009
Foreword
The 35th International Symposium on Water Supply and Drainage for
Buildings, CIB W062 2009, is organized by the Department of Building
Services and Environmental Engineering of the Fachhochschule
Gelsenkirchen University of Applied Sciences with the support of the
International Council for Research and Innovation in Building and
Construction (CIB). The Symposium is held in Düsseldorf from 7th to 9th
September 2009. Further support is given by VDI, the Association of
German Engineers, and DVGW, German Technical and Scientific
Association for Gas and Water. It is partially financed by the
Fachhochschule Gelsenkirchen University of Applied Sciences and the
sponsors.
The proceedings contain thirty-four papers presented in eight sessions.
They cover the research and development in the fields of hot and cold
water supply and drainage systems, water conservation, corrosion and
materials including trends, applications, evaluation and management.
We thank all authors for their participation and contributions to this
symposium.
We would also like to thank the organizing committee and the international
scientific committee for their advice on the selection and organization of
the sessions as well as the VDI and DVGW for their special support.
We would also like to express our gratefulness to the administration of the
Fachhochschule Gelsenkirchen University of Applied Sciences and all the
sponsors for their generous financial support in the current global crisis.
Prof. Dr. Markus Thomzik
Dean
Prof. Dr. Mete Demiriz
Organizer
i Organizing Committee
Chairman:
Programme Manager:
Secretary:
Mete Demiriz
Cécile Julien
Nina Jablonski
Robert Schütz
International Scientific Committee
K. De Cuyper
L.S. Galowin
O.M. Gonçalves
Y. Assano
E. Petresin
J.A. Swaffield
F. Derrien
L.T. Wong
M. Demiriz
Belgian Building Research Institute, Belgium
National Institute of Standards and Technology,
USA
Escola Politecnica University of Sao Paulo, Brasil
Shinshu University, Japan
University of Maribor, Slovenia
Herriot Watt University, Scotland,
United Kingdom
Centre Scientifique et Techniquie du Batiment,
France
The Hong Kong Polytechnic University,
Hong Kong, China
Fachhochschule Gelsenkirchen,
University of Applied Sciences,
Gelsenkirchen, Germany
Partners
VDI - The Association of German Engineers - Society of Building Services
DVGW - German Technical and Scientific Association for Gas and Water
Sponsors
Geberit
Viega
Gebr. Kemper GmbH
World Plumbing Council
Jung Pumpen GmbH, Pentair
Studor Group
Saint-Gobain HES
IAPMO
ii Content
Session I - Water Supply I
I.1
TREND IN AND RECENT RESEARCH INTO DIRECT
WATER SUPPLY SYSTEMS IN JAPAN
Noriyoshi Ichikawa, Shizuka Hori, Tsutomu Nakamura,
Tamio Nakano, Sadahiko Kodera, Satoshi Nakayama,
Japan
Page 10
I.2
FIXTURE UNITS AT CHOICES OF REFERENCE
DESIGN FLOW RATES FOR SIMULTANEOUS
DEMAND PROBLEMS OF LARGER WATER SUPPLY
SYSTEMS OF HONG KONG
L. T. Wong, K. W. Mui, China
Page 26
I.3
APPLICATION OF FUZZY LOGIC TO THE
ASSESSMENT OF DESIGN FLOWRATE IN WATER
SUPPLY SYSTEM OF MULTIFAMILY BUILDING
Lúcia Helena de Oliveira, Liang Yee Cheng,
Orestes M. Gonçalves, Pedro M. C. Massolino,
Brasil
Page 39
I.4
THE EVALUATION OF THE PER CAPITA WATER
DEMAND OF THE UNIVERSITY OF SÃO PAULO
THROUGHOUT TEN YEARS OF THE WATER
CONSERVATION PROGRAM OF THE UNIVERSITY
OF SÃO PAULO
G. Silva, H. Tamaki, G. Correia, O. Gonçalves,
Brasil
Page 55
iii Session II – Water Supply II
II.1
PROACTIVE CRISIS MANAGEMENT OF URBAN
INFRASTRUCTURE
Nekrep Perc, Slovenia
Page 69
II.2
WATER SYSTEMS IN HIGH RISE RESIDENTIAL
BUILDINGS, GUIDE LINES FOR DESIGN AND
CONSTRUCTION
Walter van der Schee, Netherlands
Page 85
II.3
EVALUATION OF BOOSTER PUMP SYSTEM IN
OFFICE BUILDING
Saburo MURAKAWA, Kazuhiko SAKAMOTO ,
Yasuo KOSHIKAWA, Hiroshi TAKATA,
Japan
Page 105
II .4
CONSIDERATION ON THE WATER ENVIRONMENT
PERFORMANCE OF ARCHITECTURE, CITY AND
EARTH
Hiroyuki Kose, Japan
Page 125
iv Session III – Hot Water Systems
III.1
STUDY ON THE SYSTEM EFFICIENCY OF THE
LATEST HOT WATER SUPPLY SYSTEM UNDER
PRACTICAL USE CONDITION
Masaharu Itagaki, Saburo Murakawa, Japan
Page 135
III.2
CALCULATION METHOD FOR LOADS OF HOT
WATER DEMAND WITH THE HOT WATER
STORAGE TANK SYSTEM IN HOUSES
Part 2 Modeling of the loads of reheating bathwater and
experimental evaluation of CO2 heat pump water heater
Hiroshi TAKATA, Saburo MURAKAWA, Akiko TAKAAZE,
Hiroki KITAYAMA, Yasuhiro HAMADA, Minako
NABESHIMA, Japan
Page 151
III.3
HOTSPOT FREE DESIGN, BUILDING &
INSTALLATION OF DRINKING WATER
INSTALLATIONS
Ing. O.W.W. Nuijten, Netherlands
Page 164
III.4
AN INVESTIGATION INTO FACORS AFFECTING
THE DESIGN TECHNIQUES USED TO CONTROL
LEGIONELLA IN WATER SYSTEMS
John Turner, United Kingdom
Page 182
III.5
A NEW HYGIENE SYSTEM FOR COLD AND WARM
DRINKING WATER INSTALLATIONS
U. Petzolt, M. Demiriz, Germany
Page 201
v Session IV – Water Conservation
IV.1
WATER CONSERVATION AND QUANDARIES
Dr. Lawrence S. Galowin,
United States of America
Page 218
IV.2
BUILDING RAINWATER HARVESTING.DOUBTS
AND CERTAINTIES
Silva-Afonso, Armando, Portugal
Page 238
IV.3
WATER CONSERVATION AT INTERNATIONAL
AIRPORT OF SÃO PAULO IN BRAZIL: THE
HIDROAER PROJECT
Lúcia Helena de Oliveira , Wilson Cabral de Sousa Júnior,
Marina S. O. Ilha, Orestes M. Gonçalves,
Brasil
Page 247
IV.4
WATER EFFICIENCY OF PRODUCTS. OUTCOME OF
APPLYING A CERTIFICATION AND LABELLING
SYSTEM IN PORTUGAL
Carla Pimentel-Rodrigues, Armando Silva-Afonso,
Portugal
Page 259
IV.5
SYSTEM-DERIVED CHANGES TO THE TIMEDEPENDENT CHARACTERISTICS OF WATER,
RAINWATER
AND
DRAINAGE
FLOWS
IN
BUILDINGS
Dr L.B. Jack, United Kingdom
Page 271
vi Session V – Drainage I
V.1
EMPIRICAL STUDY ON TERMINAL WATER
VELOCITY OF DRAINAGE STACK
Dr. C.L. Cheng, Ms. W.J.Liao, Dr. K.C. He ,
Ms J.L.Lin, Taiwan
Page 283
V.2
DRY DRAINS: MYTH, REALITY OR IMPEDIMENT
TO WATER CONSERVATION
Prof. J.A. Swaffield, United Kingdom
Page 300
V.3
DESIGNING SEWERS FOR
WATER FLOWS
Jeff Broome, United Kingdom
REDUCED
WASTE
Page 313
Session VI – Drainage II
VI.1
STUDY OF THE PERFORMANCE EVALUATION OF
DRAINAGE SYSTEMS WITH AN AIR ADMITTANCE
VALVE FOR SINGLE-FAMILY HOUSING
Effects of Swivel Fittings with Air Admittance Valve Function
to Improve Drainage Capacity
Kazutoshi Suzuki, Masayuki Otsuka, Norihiro Hongo, Koichi
Kawasaki, Japan
Page 326
VI.2
A STUDY OF A PREDICTION METHOD FOR
DRAINAGE PERFORMANCE OF DRAINAGE STACK
SYSTEMS USING A HORIZONTAL FIXTURE DRAIN
BRANCH SYSTEM WITH AN AIR-ADMITTANCE
VALVE
Masayuki Otsuka, Zhe Zhang, Japan
Page 340
vii VI.3
APPLICATION
OF
COMPUTATIONAL
FLUID
DYNAMICS TO SIMULATE HE BEHAVIOR OF
FLUIDS INSIDE VERTICAL STACK OF BUILDING
DRAINAGE SYSTEMS
Eric Wai Ming Lee, China
Page 355
Session VII – Drainage III
VII.1
TARGET DEFINITION AND ASSESSMENT OF
PERFORMANCE- COST: BENEFIT EVALUATIONTARGET DEFINITION AND ASSESSMENT OF RISKSENGINEERING
DEVELOPMENTSIKT
COMPARATIVE PRODUCT TESTS AS A BASIS FOR
ASSET MANAGEMENT INVESTMENT DECISIONS
Dr.-Ing. Bert Bosseler, Germany
Page 363
VII.2
FROM
DESKTOP
TO
PLANT
ROOM:
DEVELOPMENTS OF AN INNOVATI VE SYSTEM
FOR MAPPING AND ASSESSING TRAP SEAL
VULNERABILI TIES IN BUILDING DRAINAGE
SYSTEMS – LESSONS FROM THE FIELD.
Dr. M.Gormley, C. Hartley, United Kingdom
Page 379
VII.3
REDUCING THE RISK OF CROSS-CONTAMINATION
FROM THE BUILDING DRAINAGE SYSTEM USING
THE REFLECTED WAVE TECHNIQUE TO IDENTIFY
DEPLETED TRAP SEALS
D.A. Kelly, United Kingdom
Page 393
VII.4
TEST
METHOD
OF
PRESSURE-RESISTANCE
PERFORMANCE OF TRAP
Kyosuke Sakaue , Hiroshi IIzuka, Motyasu Kamata,
H. Kuriyama, Japan
Page 406
viii Session VIII – Miscellaneous
VIII.1
ANALYSIS OF CORROSION MECHANISM OF
SPRINKLER PIPING AND CORROSION
PROTECTION
Toshihiro YAMATE, Saburo MURAKAWA, Japan Page 420
VIII.2
A STUDY OF SANITARY EQUIPMENTS INSTALLED
ON LIGHT-WEIGHT PARTITIONS
Dr. M.C. Lee, Dr. R.Z. Wang, Dr. C.L. Cheng, Y.C. Yu,
Z.Y. Shih, Taiwan
Page 441
VIII.3
A STUDY ON PERFORMANCE EVALUATION FOR
TOILET SYSTEMS WITH AN UROFLOWMETER
Yuta Takahashi, Masayuki Otsuka, Hironori Yamazaki,
Japan
Page 456
VIII.4
THE POSSIBILITY OF WET SLUDGE UTILIZATION
FOR GREEN HOUSE EFFECT GAS EMISSION
REDUCTION
Prof. Yoshiharu Asano, Japan
Page 473
VIII.5
FORMULATION OF A SYNTHETIC GREYWATER TO
EVALUATE THE PERFORMANCES OF ON-SITE
GREYWATER RECYCLING TECHNOLOGIES
Fanny Hourlier, Anthony Massé, Pascal Jaouen, Abdel Lakel,
Claire Gérente, Catherine Faur, Pierre Le Cloirec,
France
Page 486
VIII.6
MATERIALS IN CONTACT WITH DRINKING WATER
F. Derrien CSTB, France
Page 502
ix Session I – Water Supply I
I.1
Trends in and Recent Research into
direct Water Supply Systems in Japan
N. Ichikawa
nichi@tmu.ac.jp
Department of Architecture, Graduate School of Urban Environmental Sciences,
Tokyo Metropolitan University, Tokyo, Japan
S. Hori (ZO Consulting Engineers Inc.)
S. Kodera (Urban Renaissance Agency)
T. Nakamura (Suga Corporation.)
S. Nakayama (Tokyo Metropolitan University)
Abstract
The penetration of water service in Japan is 97.3% (now in March, 2008). Right now,
higher quality of safe water is supplied to each building. In recent years, to achieve
supplying more “Delicious Water”, review of clean water processing and range
expansion of direct water supply system is advanced. In such a flow, our country’s
water supply and draining system in buildings will be facing with various changes.
The latter method which first stores water in a tank and after distributes on demand has
been adopted in middle to large-scale buildings. However, the collaborative works
among the public, private and academic sectors with various perspectives resulted in the
implementation of the direct water supply system for middle scale buildings in Japan.
10 The author had worked to review the direct water supply system for about 20 years in
many related committee. The committee the author participated is starting with
examination of flowing quantity at moment in direct increasing pressure water supply
system, Bureau of Waterworks Tokyo Metropolitan Government, Yokohama
Waterworks Bureau, etc. and committee in the statutory board such as countries and
municipalities. Historical details of the water service administration, trend of direct
water supply system, the research cases that author is doing and examination problem in
the future and so on were described in this paper.
Keywords
Direct Water Supply System, Estimate for Peak Flow Rate of Water Supply
1. Introduction
1) Historical Background
The background to the administration of waterworks in Japan is shown in Fig. 1. The
first pressurized waterworks (hereinafter referred to as “waterworks”) was the
Yokohama city waterworks, design and construction of which was supervised by H.S.
Palmer (U.K.) in 1887. This marked the beginning of modern waterworks in Japan. The
planned water supply rate at the time was approximately 90L (20 gallons)/[person/day],
about one quarter of today’s planned water supply rate. Following this, waterworks
spread in stages mainly to harbor cities such as Nagasaki, Osaka, Tokyo, and Kobe.
2) Penetration
1957 saw the promulgation of the Waterworks Law in Japan. The penetration of water
service at the time of 40.7% increased to 53.4% in 1960, making water supply available
to more than half the Japanese population. The penetration of water service at the
present time is 97.4%, a level that can be said to represent completion of waterworks.
11 3) Water Quality
Water quality standards for tap water (drinking water) were established by edict by the
Ministry of Health, Labour and Welfare (MHLW) in 1958. Since that time, appropriate
reviews and revisions of these water quality standards have continued up to modern
times. The 26 items of the water quality standards of December 1992 have increased
significantly to 46 items. At the present time, a total of 50 items have been enacted
1887
Start of modern water works (Yokohama City)
1957
1957
Promulgation of water works low
Diffusion of the water supply system 40.7%
1958
1960
Issued water quality standard by MHLW
Diffusion of the water supply system 53.4%
1991
Suggestion to introduce of DBWS system by MHLW
1992
Start of experimental DBWS system (Yokosuka City)
2009
2009
Diffusion of the water supply system 97.3%
Introduced in141 local governments
Fig.1 Background to the administration of waterworks in Japan
under the water quality standards.
4) Water Supply Systems
To supply water to small-scale buildings such as detached houses or buildings
approximately two stories high, a direct pressure water supply system has been used,
while a receiving tank system has been used for other types of buildings.
Recent years have seen an increase in the scope of application of direct water supply
systems. The reason for this was the announcement in 1991 by the Ministry of Health,
Labour and Welfare of “Long-term goals for the organization of waterworks looking
toward the 21st Century” as a national policy. This was because the following items
linked to the expansion of the scope of direct water supply systems were set out as
concrete goals for the organization of waterworks.
a) Concrete promotion of the use of direct water supply systems in building
approximately 4-5 stories high.
b) Elimination of hygiene-related problems such as small-scale receiving tanks not
properly maintained and managed.
12 Against this background, multi-faceted investigations were initiated by the government,
universities and private-sector enterprises acting in collaboration. However, at the time
in question, investigations were focused on the introduction of the direct booster water
supply (DBWS) system, so that consideration of the direct pressure water supply system
has only begun recently.
2. Direct Water Supply System Trends
In recent years, the direct booster water supply system has been generally adopted for
use in medium-rise apartment housings in metropolitan areas. N. Ichikawa reported on
the details such as the background to introduction of the direct booster water supply
system and an overview of the system and methods of design at the CIB-W62
International Symposium held in France in 2004 (1). Moreover, the author also talked
about initiation of investigations by Tokyo into expanding the scope of meter caliber
from 50A to 75A (80A). At the present time, the number of local authorities using the
75A caliber is on the increase.
Meanwhile, the direct pressure water supply system was principally used for water
supply to small buildings such as detached houses and buildings approximately two
stories high. Recent years have seen the beginning of consideration of application of this
system to housing complexes of 4-5 stories. As shown in Table 1, direct water supply
systems (direct pressure and direct booster water supply systems) has many advantages.
Furthermore, up to the present time, receiving tank systems have generally been used
for water supply to schools due the advantage of being able to store water for
emergencies. However, because of deterioration in water quality caused by such factors
as long school holidays and a decrease in the number of children, the use of direct water
supply systems as shown in Fig. 2 is being promoted for drinking water. In metropolitan
Tokyo, renovations are underway at 400 public primary schools, approximately one
third of these schools, to convert to direct water supply systems by the year 2010.
Although, as the above indicates, the scope of use of direct water supply systems in
increasing in Japan, this entails a wide diversity of problems from the standpoint of
water-supply equipment design, and research has been initiated aimed at resolving these
problems. Trends in direct water supply systems with the focus on examples of research
conducted by the authors in recent years are set out below.
13 Table.1 Advantages of direct water supply system
Advantage
/ Keeping water quality
/ Energy-saving
/ Miniaturization of space
/ Reduction of running cost
Disadvantage
/ Difficult to service when the accident happened
/ Influence of the pollution by flow backwards
Elevated Strage tank
closet
Faucet (for drinking)
Receiving tank system
Receiving tank
pump
From Water
service pipe
M
Direct pressure water supply system
Fig.2 Water supply system for school
2.1
Trends toward Expansion of the Scope of the Direct Pressure Water Supply
System
Regarding the expansion of the direct pressure water supply system (application to
buildings 4-5 stories high), factors such as pressure fluctuations in main water pipes, the
building usage status and the effects on faucets in units of simultaneous use of
appliances need to be considered.
(1)
Understanding the Status of Existing Housing Complexes
The authors conducted actual measurements and investigations of an existing 4-story
apartment housing where a direct pressure water supply system was used with the aim
of ascertaining the status of usage of water within the building. This research project,
14 requested by the Yokohama Waterworks Bureau, spanned the two years between April
2007 and March 2009.
The main goals of the survey were (a) to ascertain the status of usage of water within
the building and (b) to elucidate the effects of pressure fluctuations in main water pipes
on faucets in units in the building. An overview of the survey target apartment housing
is shown in Fig. 3. The building comprised at total of 16 units (11 of which were
occupied) featuring application of header system piping. Measuring devices capable of
measuring items such as the pressure and application-specific flow rates on each floor
404
403
M
M
20A
20A
303
40A
M
20A
M
20A
M
M
50A
40A
M
M
20A
20A
103
102
Vacant
room
Vacant
room
M
20A
20A
50A
Water service pipe
20A
201
M
20A
101
M
20A
50A
Fig.3 Overview of survey target
15 M
50A
M
50A
301
50A
202
M
Vacant
room
20A
Vacant
room
20A
50A
104
20A
50A
20A
203
M
302
M
50A
204
40A
M
50A
304
401
Vacant
room
40A
20A
402
Number of fixtures in
simultaneous use
0.30
0.25
0.20
0.15
0.10
0.05
0.00
3
4
5
6
7
8
Summer (weekday)
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
hour
Summer (holiday)
Winter (weekday)
0
1
Winter (holiday)
2
Fig.4 Number of fixtures in simultaneous use
were set up in the surveyed building. During the survey, these measuring devices were
used to automatically record signals from sensors on a laptop computer and this data
was subsequently analyzed. The water usage status shown by the results of the survey
are set out in Fig. 4. The results of the survey provided the following findings on the
water usage status.
1) The timeframe during which the highest number of appliances was used
simultaneously both in the summer and winter months were between 6:00 and 7:00
am on weekdays.
2) The maximum average number of appliances used simultaneously per timeframe
was recorded during weekday mornings and the value was 0.281.
3) During peak load times, the frequency of occurrence of the number of appliances
used simultaneously obtained from the measurement results was close to the results
deduced from the Poisson distribution based on the average number of appliances
used simultaneously.
4) Measurement showed that an average of three appliances was used simultaneously
in units on a daily basis. Based on the average number of appliances used
simultaneously, the number of appliances used simultaneously estimated using the
Poisson distribution matched the number of appliances used simultaneously
projected with a significance level of 0.01.
16 (2)
Consideration of Effects of Pressure in Main Water Pipes on Terminal
Appliances
In Japan, water pressure at branch points from water pipes to sites and buildings is
regulated at 0.15 MPs or more. However, depending on the water supply system
performance, there are cases where the occurrence of insufficient pressure makes it
impossible to achieve suitable capacity. This phenomenon often occurs when multiple
appliances are used simultaneously. In considering expansion of the scope of the direct
pressure water supply system, there are many problems that need to be investigated such
as the basic conditions of main pipe pressure and piping in buildings, water outflow
rates and water usage intervals.
With the aim of investigating the effects of pressure fluctuations in main water pipes on
appliances in units, the authors conducted an experiment based on the concept of a 4story apartment housing. This research project, requested by the Yokohama Waterworks
Bureau, spanned the two years between April 2007 and March 2009.
The experiment used part of the 100 m tower of the Urban Renaissance Agency. An
overview of the experiment model is shown in Fig. 5. The experimental facility featured
a single piping system capable of supplying water to two units per floor up to and
including the 4th floor. In order to set the water supply pressure as a parameter for the
experiment, a reduced pressure valve was installed in the water supply service point to
adjust pressure. Additionally, three appliancesd (water closet (built in flush valve),
single faucet and shower) were installed on the top floor based on interior piping for one
unit as the model. For other units connected to the vertical piping system, a fixed-flow
valve was installed to produce a water supply load. For experiment parameters, water
supply pressure was set to 0.25 - 0.50 MPa and the water outflow volume for each
appliance and pressure in the proximity of appliances was measured.
17 P
: Pressure sensor
M
: water meter
: reduce pressure valve
R
A
A
P
M
: constant flow regulating valve
40
25
40
32
40
32
40
40
25
P
R
0.10
0.05
0.00
24
22
20
18
16
14
12
10
8
6
4
2
0
0.6
0.4
0.2
0.25
0.3
0.35
0.4
0.45
0.5
Pressure on service pipe [
MPa]
Flow:using a closet
Flow:using 3 appliances
(without water supply load)
Flow:using 3 appliances
(with water supply load)
Flow on the other house
Flow
0.8
Pressure
0.0
Pressure on fixture [
MPa]
0.15
Flow on fixture[L/min]
0.20
Pressure on fixture [MPa]
0.25
25.0
22.5
20.0
17.5
15.0
12.5
10.0
7.5
5.0
2.5
0.0
ba
sic
N
o
us
in
12 g
L/
m
in
24
L/
m
i
36 n
L/
m
i
48 n
L/
m
in
Flow on fixture [L/min]
Fig.5 Overview of the experiment model
Pressure:using a closet
Pressure:using 3 appliances
(without water supply load)
Pressure:using 3 appliances
(with water supply load)
Fig.6 Results of the experiment
An example of the results of analysis of the results of the experiment is shown in Fig. 6
and findings from these results are shown below.
1) When the three appliances were used simultaneously, the measured value for the
water closet (built in flush valve) at a water supply pressure of 0.25 MPa was
below the absolute minimum required pressure for the appliances regardless of
the water supply load status in other units.
2) Results showed that the absolute minimum pressure was achieved for the water
closet (built in flush valve) at a water supply pressure of 0.30 MPa.
3) The reduction in water supply pressure when appliances are used simultaneously
is largely due to pressure in piping within the units, and this point to a need to
conduct a detailed investigation of the design of piping between target sections.
18 (3)
Consideration of Effects of User Comfort (Irritation)
It is supposed that the direct pressure water supply system presents the possibility of
reduced water outflow rate due to pressure fluctuations when multiple appliances are
used simultaneously. It can be surmised that this reduced water outflow rate leads to
dissatisfaction in users and serves to create the feeling of uncomfortable. Levels of
dissatisfaction in fields involving heat sensitivity are assessed using PPD (Predicted
Percentage of Dissatisfied).
Based on this concept, the authors commenced research into levels of dissatisfaction
relating to water usage behavior from the year 2007 and this research is still ongoing
today. This research also represents one of the research themes selected for a grant in
aid for scientific research by the Ministry of Education, Culture, Sports, Science and
Technology (Representative researcher: Noriyoshi Ichikawa) for the three years from
2008 - 2010. The level of dissatisfaction that causes irritation marks the boundary
within which it is possible to maintain comfortable use of water. By ascertaining levels
of dissatisfaction, the authors believe that it is possible to integrate standards for
determination of design of future water supply systems and performance assessment
into one.
2.2
Trends in Expansion of the Scope of the Direct Booster Water Supply System
Introduction of the direct booster water supply system started in areas such as Europe
and America in the 1960’s. The reason for the delay in introduction of this system in
Japan was regulation by the Waterworks Law to the effect that “Pumps that pose the
risk of affecting waterworks piping must not be directly connected to waterworks
service piping.”
19 ARV
M : WATER METER
*BFP : BACKFLOW PREVENTER
BPU : BOOSTER PUMP UNUT
ARV : AIR RELEASE VALVE
EST : ELEVATED STRAGE TANK
*BPU
BFP
M
Fig.7 Direct booster water supply system
As shown in Fig. 7, this system was put into trial use in Yokosuka City in 1992 and this
marked the starting point for full scale introduction in major metropolitan areas such as
Osaka and Tokyo. Since then, use has increased rapidly with many local authorities
introducing the system. Although, at the time of introduction, meter caliber was set to
maximum of 50A, recognition by Tokyo authorities of the need for meter caliber up to
80A (75A) from the year 2004 onward due to expansion of the scope of usage of the
system has led to the acceptance of 80A (75A) by more than 50 local authorities at the
present time.
Since the introduction of direct water supply systems, a wide range of technical
standards (e.g. Japan Water Works Association Standards JWWA-B-129 “Valve
Backflow Preventers for Water Supply,” JWWA-B-130 “Pressure Booster for Direct
Water Supply” and JWWA-B-134 “Reduced Pressure Principle Backflow Preventers
for Water Supply”) have been enacted and development of various devices undertaken.
Some examples of developments and research undertaken relating to the direct booster
water supply system are detailed below.
(1)
Development of Backflow Preventers
It is extremely important that water contamination in main pipes be prevented in
waterworks supply systems. The installation of backflow preventers in primary booster
pumps in the direct pressure water supply system is mandatory, as shown in Fig. 7.
However, the significant pressure loss of about 0.1 MPa in this device has given rise to
considerable controversy. At the present time, compared to previous devices, products
that achieve a reduction in pressure loss of approximately 20% are in the process of
realization. As well as fully ensuring backflow prevention performance, positive
20 consideration in future of the reduction of resistance loss, while seemingly contrary, is
essential in terms of realizing energy conservation by making effective use of the
energy in main pipes.
(2)
Consideration of Application in High-rise, Large-scale Building
The application of the direct booster water supply system in even higher, larger
apartment housings is currently under consideration. The following experiments were
conducted in the 100 m high Urban Renaissance Agency tower (fig.8) .For applications
in high-rise buildings, the system involved serial installation of booster pumps on the
first and intermediary floors. (2) For applications in large-scale buildings, a “parallel
booster water supply system” where booster pumps were installed in parallel was used.
Based on this kind of investigation, introduction of booster pumps in serial multi-stage
Fig.8 Overview of the experiment system
water supply systems and parallel water supply systems was initiated in Tokyo from
February 2009. This means that direct water supply systems with boosters can be used
in a wider range of buildings. However, there are a great many issues to be closely
considered and the authors feel that introduction of these systems at the present time
would be somewhat premature.
(3)
Consideration of Estimates for Peak Flow Rate of Water Supply
In recent years, in apartment housings in particular, factors such as changes in the
number of family members per unit and the spread of water-saving devices have
produced changes in the water usage status, leading to the need to review calculation
21 formulae. However, since the costs of investigations preparatory to a full review would
be prohibitive, no such initiative has been undertaken up to the present time.
Fortunately, the authors have been able to conduct an actual survey of water usage
status in housing complexes and to make a start on carrying our research into reviewing
estimates for peak flow rates of water supply matched to actual circumstances. This
research project, requested by the Bureau of Waterworks, Tokyo Metropolitan
Government, spanned the two years between April 2007 and March 2009.
During the survey, which targeted 230 apartment housings in metropolitan Tokyo that
use the direct pressure and direct booster water supply systems, peak flow rates were
measured. To carry out measurements and analysis, signals from electronic water
meters were recorded on a data logger every second and analysis carried out after
accumulating one minute’s worth of data. Examples of some of the results are shown
below.
1) Based on regression equations obtained from the results of measurements, additional
calculation formulae were acquired with the parameter of satisfying all measured
values (hereinafter referred to as “peak simultaneous use rate formula”)
N  30
31 N
Q = 28N0.29
Q = 13N0.51
N: Number of persons, Q: Peak simultaneous use rate [L/min]
22 600
Peak flow rate [L/min]
500
(Present)
Estimate for peak flow rate
400
300
Actual measurement
(Proposal )
Once a year
Once a half year
Once every three month
Once a month
Estimate for peak flow rate
200
100
0
0
100
200
300
400
500
Number of persons
600
700
800
Fig.9 Estimate for peak flow rate
2) For safety factor settings, an attempt was made to use Gumbel distribution (extreme
value statistics) utilized in fields such as earthquakes and rainfall to estimate
maximum values.
3) The approach to the concept of safety rates is extremely important. For this reason,
data from different apartment housings where long-term measurements were in
progress was used to estimate maximum values per period (1 month, 3 months, 6
months, 1 year) and it was proposed that these be added to safety factors.
4) The graph shown in Fig. 9 was produced from the results of reflection in the peak
simultaneous use rate formula of safety factors per probability.
5) It is clear that the current calculation formula allows much greater latitude than the
peak simultaneous use rate formula acquired from the results of actual measurement
during this research.
23 3. Conclusion
This report focuses on research conducted by the authors into recent trends in direct
water supply systems. The authors judge that the content of this report adequately
represents the current state of direct water supply systems in Japan. In addition to
continuing in their pursuit of the research presented herein, the authors plan to conduct
investigation into the following areas.
1) Construction of a new simultaneous peak flow calculation method
2) Assessment of the efficacy of energy in water pipes in direct water supply systems
3) Elucidation of levels of dissatisfaction arising from various water usage behaviors
4) Elucidation of the various required performance factors for the direct pressure
water supply system
4. Acknowledgements
This study was conducted as a part of the “Study on direct pressure water supply system
aiming at the most suitable design” program 2009 (Representative; N. Ichikawa)
supported by Ministry of Education, Culture, Sports, Science and Technology of Japan
(MECSST). The authors would like to thank for the MECSST.
5. Reference
1) N. ICHIKAWA et al., Examination of Direct Booster Water Supply System in
Japan, Proceedings of the 31st CIB - W62 International Symposium, 2004
2) A. INADA, S HORI, N. ICHIKAWA et al., Study on City-Pressure Water Supply
System (Part 1)- (Part 3), Technical Papers of Annual Meeting, SHASE, 2008-2009
3) T. NAKANO, S HORI, N. ICHIKAWA et al., The research of the computation of
the design quantity of water supply at the decision of the bore of the crane in the
tenement houses (Part 1)- (Part 4), SUMMARIES OF TECHNICAL PAPERS OF
ANNUAL MEETING, JIA, 2008-2009
24 6. Presentation of Author
Dr. Noriyoshi Ichikawa is professor at Tokyo Metropolitan
University, Graduate School of Urban Environmental Sciences,
Department of Architecture and Building Engineering. And he is
President of JSPE(Japan Society of Piping Engineers). He is
conducting various researches on his major field of study of
water supply and drainage system in buildings. He is also
actively involved in governmental and academic institutions and
committees related to his field of study as chief coordinator and
board member.
25 I.2
Fixture units at choices of reference design
flow rates for simultaneous demand problems
of larger water supply systems of Hong Kong
L. T. Wong1, K. W. Mui
Department of Building Services Engineering, The Hong Kong Polytechnic University,
Hong Kong, China
1
Tel: (852) 2766 7783; 1Email: beltw@polyu.edu.hk
Abstract
Fixture unit approach used for estimating the probable maximum simultaneous demands
in building water supply systems is based on a fact that a given simultaneous reference
design flow rate may be produced by different numbers of identical appliances
characterized by the appliances’ discharging flow rates and discharge probabilities.
Each appliance is represented by a fixture unit value, which indicates the appliance
associated with the same simultaneous demand of a number of base case appliances
characterized by the base case discharging flow rate and discharge probability. The
validity of the selected reference design flow rate and its sensitivity to the probable
maximum simultaneous demand for water systems in high-rise residential buildings are
examined in this paper. In particular, fixture units and the estimated probable maximum
simultaneous demands due to appliances attributed by discharge probabilities and
discharging flow rates ranged from 1/8 to 8 times the based case attributes are
considered. Estimated demands from the fixture unit approach are compared with
computational results for an example water supply installation by Monte-Carlo
simulations. The results showed that the existing choice of a reference flow rate at 10
Ls−1 for the fixture unit approach would be sufficient in determining the probable
maximum simultaneous demands not to exceed a probable failure rate of 1% for 900
26 pairs of WC-and-washbasin installations in residential buildings. An increased reference
design flow rate would be required for the applications of the fixture unit approach in
demand analysis of larger water installations in similar densely built environment.
Keywords
Demand analysis, water supply system, fixture unit approach, reference design flow rate
Introduction
Probabilistic approaches for estimating the usage patterns of water appliances and their
associated instant demands at any points of a water supply system have been adopted in
many practical building installations [1]. The simultaneous demand problems of a water
supply system were addressed from the binomial theory for frequency analyses of
usages [2]. Actuations of an appliance in the installation occur randomly and
intermittently with variable magnitudes and they can be described by the probability for
appliance discharging events. The analysis provided a means for quantifying a probable
‘not to exceed’ failure rate in fulfilling certain instant demands. The design approach is
practical because water supply main is very unlikely to address the simultaneous
demands of all installed appliances. The installations may be ‘overloaded’ with certain
number of appliances operating simultaneously where a small failure probability to the
theoretically maximum demand is allowed [3]. The validity of the allowable maximum
failure rate can be investigated through field measurements of in-use installations. It
was reported that the observed maximum demands in some water supply systems did
not exceed a failure rate of 1% derived from geometric demand patterns of observed
demands in a study [4].
A fixture unit approach was used to evaluate the probable maximum simultaneous
demand problems in building water supply systems. This approach is based on the fact
that a given simultaneous reference design flow rate can be produced by different
numbers of identical appliances characterized by the appliances’ discharge flow rates
and discharge probabilities [5]. The probable discharge flow rate of an appliance can be
equivalent to a number of base case appliances and assigned appliances with fixture
units. The choice of this reference design flow rate is an assumption needed to be
studied in detail for large water supply systems.
Apart from solving the problem with the fixture unit approach, Monte-Carlo simulations
can also be used to determine the probability density function of system failures to meet
the instant water demands [6]. A stochastic model for estimating the instant water
demands in a water supply system was developed where modelling parameters were
27 obtained by Monte-Carlo sampling technique without an assumption of the reference
design flow rate [7]. However, the fixture unit approach is simple to use and many
designs employed the Hunter’s probabilistic method for practical water pipe and plant
sizing [8]. From the actual building usage patterns, data, and extended laboratory
research results, the piping requirements using the probabilistic approach can be applied
for both water supply and discharge systems in buildings with tabulations and design
curves as specified in some design guides [9]. Codification was resulted from fixture
units for probable instances in building water pipe/plant sizing. Water supply loading
tables in plumbing design applications were based upon loads in fixture units for
practical applications.
In this study, the variability of the probable maximum simultaneous demands in water
supply installations due to the choice of various reference design flow rates is
investigated. Estimated demands from the fixture unit approach are compared with the
computational results by Monte-Carlo simulations. Appropriate choices of water supply
systems in high-rise buildings of Hong Kong are recommended.
Simultaneous demands and fixture unit approach
For a base case appliance having repeated cycles of discharge operation with a mean
discharge period d (s) and the mean time interval between discharges w (s), the
probability of the appliance discharge p at any time is [2],
p 
d
w
… (1)
Assume the appliance operations are binomially distributed, and the probability p of N
base case appliances operating out of M identical base case appliances installed in the
installation, M pN is given by, where, (1p) is the probability of the appliance not
operating and C MN is the binomial coefficient,
N
pN  CM
N p 1  p 
MN
M
; CM
N 
M!
N ! M  N  !
… (2)
In some water supply system designs, piping systems are designed for a maximum
acceptable risk of failure in order to minimize the cost of the system with design
number of N (out of M installed, say, M>30) base case appliances operating
simultaneously. This design implies that the plants and piping systems might be
‘overloaded’ when serving all the M appliances operating simultaneously, i.e., the
theoretical maximum simultaneous flow rate. When more than N appliances are
operating, the acceptable level of the system in terms of reliability is defined as
28 ‘engineering unsatisfactory’ (i.e., the occurrence of ‘failure’). The failure rate  is
determined by the sum of the probabilities that more than N appliances are operating
simultaneously,
  p N  1  p N  2   ...  p M  1  p M  
M

i  N 1
pi ; N < M
… (3)
The number of appliances N that are operating simultaneously can be determined by the
probability p at an acceptable failure rate , which would be approximated by the
Sterling’s formula for an ‘engineering acceptable’ limiting failure rate. The probable
number of appliances operating simultaneously can be expressed by Equation (4) with z
= 1.82255 [10] for  = 1%, which is recommended in some designs,
N  Mp  z 2 Mp 1  p 
… (4)
The corresponding probable maximum simultaneous demand qd (Ls−1) due to the
installations of M appliances is then determined by Equation (5), where q (Ls−1) is the
discharging flow rate of the base case appliance,

q d  Nq  q Mp  z 2Mp 1  p 

… (5)
Equation (5) can be used to determine the design flow rate of an installation consisted of
2 or more appliance types using the fixture unit approach. The fixture unit approach is
established for estimations of the probable maximum simultaneous demands in
plumbing and drainage systems in buildings [9]. Specifically, the reference
simultaneous flow rate of an installation due to a number of installed identical
appliances, say qref = 10 Ls−1, would be produced by a number of the base case
appliances with the base case usage characteristics. Each appliance is then determined
with a fixture unit value, which indicates the appliances associated with the same
simultaneous demand of base case appliances. The level of reference design flow rate
qref (Ls−1) was determined by professional judgement and its sensitivity to the probable
maximum simultaneous demand is evaluated in this study.
It is noted that each appliance is attributed by the discharge probability and the
discharging flow rate, i.e. Ab(pb,qb) and Ai(pi,qi). The same reference design flow rate
qref (Ls−1) would be produced by an installation of Mi number of appliances Ai or Mb
number of the base case appliances Ab. The fixture unit Ui at the choice of the reference
design flow rate qref (Ls−1) for the appliance type Ai is given by, taking the fixture unit
of the base case appliance Ub=1,
Ui 
Mi
Mb
… (6)
q ref
29 Figure 1 illustrates the idea of using a base case appliance characteristics Ab(pb,qb) with
the base case discharge probability pb and the base case discharging flow rate qb (Ls−1)
as shown in Figure 1(i) to approximate an appliance Ai with the discharge probability
and the discharging flow rate Ai(pi=2pd,qi=qd) or Ai(pi=pd,qi=2qb). Ideally, the 2 base
case appliances should be operated without simultaneous discharging or simultaneous
discharging exactly in phase in order to approximate a single operation of the appliance
Ai as shown in Figure 1(ii), i.e. the ideal cases of approximation. However, probable
cases in random discharge patterns of a number of base case appliances were not
excluded in the fixture unit approach as shown in Figure 1(iii), i.e. the non-ideal cases
of approximation. Indeed, the fixture unit was not only dependent on the attributes (pi
and qi) of an appliance Ai but also the choice of reference design flow rate qref.
pb qb + qb Preferred approximation Other possibilities 2qb 2qb pb pb qb <2pb >2pb qb <2qb 2pb (ii) Appliance i pb + pb qb (i) Base case (iii) Model cases
pb: base case discharge probability Figure 1: Models of discharging appliance Ai(pi,qi) using a base case appliance
Ab(pb,qb)
30 Results and discussions
A base case appliance in an existing design guide attributed by the discharge probability
pb=0.0282 and the discharging flow rate qb=0.15 Ls−1 was used for discussion, i.e.
Ab(pb,qb)~[0.0282, 0.15]; and the corresponding base case fixture unit was Ub=1 at the
base case reference design flow rate qref=10 Ls−1 [5]. In order to illustrate the sensitivity
of the fixture units due to the choice of the reference design flow rate qref (Ls−1),
appliances attributed by discharge probabilities and discharging flow rates ranged from
1/8 to 8 times the based case attributes were considered, i.e. Ai=Ai(pi,qi), where pi=kpb,
qi=kqb and k[0.125, 8] respectively. Values of the fixture units were evaluated at
various reference design flow rates qref (Ls−1).
1000
k=8 4 2
Fixture unit Ui 100
10
1
0.1
0.01
pi=kpb; qi=kqb 1 0.5 0.25 0.125 0.001
1
10
100
1000
Reference design flow rate qref (Ls−1) Figure 2: Fixture units of appliances references to a base case appliance of
discharge probability of 0.0282 and discharge flow rate of 0.15 Ls−1
Figure 2 shows the fixture units Ui of appliance Ai with reference to Ab at reference
design flow rates qref between 1 Ls−1 and 1000 Ls−1. It was noted that a unity based case
fixture unit Ub was defined for all reference design flow rates. Fixture units Ui of
appliances Ai(kpi,kqi) at k=0.125, 0.25, 0.5, 1, 2, 4, 8 were 0.009, 0.04, 0.193, 1, 5.6, 33,
200 at a reference design flow rate qref=1 Ls−1; Ui=0.013, 0.054, 0.229, 1, 4.5, 21, 101 at
qref=10 Ls−1; and Ui=0.015, 0.06, 0.243, 1, 4.2, 18, 74 at qref=100 Ls−1, respectively. It
was observed that the reference design flow rates qref had some influences on the fixture
units of Ai.
The fixture unit ratio  i , q
ref
indicates the variations of the values of fixture units Ui of an
appliance i at a selected reference design flow rate qref as compared with the base case
reference design flow rate qref=10 Ls−1 and is expressed by an equation below. Ideally,
31 the fixture unit ratio  i , q
ref
of an appliance is ideally close to ‘unity’ over a range of qref
−1
(Ls ), which the selected reference design flow rate is insensitive to the fixture units.
i , q ref 
U i , q ref
… (7)
U i ,10
The results showed that the fixture unit ratio of appliances of k=0.125, 0.25, 0.5, 1, 2, 4,
8 times the base case attributes were  i ,q 1 =0.67, 0.74, 0.84, 1, 1.24, 1.57, 1.98 at a
ref
−1
reference design flow rate qref=1 Ls ; and  i ,q
ref
100
=1.14, 1.11, 1.06, 1, 0.92, 0.83, 0.74
Fixture unit ratio i
at qref=100 Ls−1, respectively. Apparently, the choice of a smaller reference design flow
rate, e.g. at qref =1 Ls−1, resulted a larger variation of i.
max k=8 4 2 2
pi=kpb; qi=kqb 1 0.125 1
8 0.125 0
1
10
min 100
1000
Reference design flow rate qref (Ls−1) Fixture unit ratio i
(a)
pi=kpb; qi=qb 2
1
k[0.125, 8] 0
1
10
100
1000
Reference design flow rate qref (Ls−1) (b)
32 Fixture unit ratio i
pi=pb; qi=kqb 2
k=8 4 0.125 1
0.125 8 0
1
10
100
1000
Reference design flow rate qref (Ls−1) (c)
Figure 3: Fixture unit ratios i for appliances Ai(kpi,kqi), k[0.125, 8]
Figure 3 shows the fixture unit ratios  for appliances Ai=Ai(pi,qi), pi=kpb, qi=kqb,
k[0.125, 8], grouped into 3 cases in (a), (b) and (c). It was noted that the maximum
and minimum fixture unit ratios max, min and ranges of pi and qi were shown in the
figure. Fixture unit ratios determined from the results presented in Figure 2 were
showed in Figure 3(a) for pi=kpb, qi=kqb. It confirmed that a smaller variation of  was
found when a larger reference design flow rate (e.g. qref100 Ls−1) was selected as
compared with a smaller reference design flow rate.
Figure 3(b) showed that the fixture unit ratios were less sensitive to the discharge
probability range pi=kpb for an appliance at the base case discharging flow rate qi=qb;
the corresponding discharge unit ratios i were between 0.94 and 1.03. However, the
ratios were sensitive to the discharging flow rates qi=kqb for an appliance operating at
the base case discharge probability pi=pb as shown in Figure 3(c); the corresponding
discharge unit ratios were from i=0.67 to 2.17 at qref=1 Ls−1, i=0.69 to 1.15 at qref=100
Ls−1 and i=0.61 to 1.20 at qref=1000 Ls−1.
Comparison with a stochastic model
The probable maximum simultaneous demand of an installation with the number of
appliances can be evaluated by a stochastic model [7]. This model was applied to
evaluate the probable maximum simultaneous water demands of domestic washrooms at
complex usage patterns, where the appliances in the same washroom would or would
not operate simultaneously. The model parameter can be identified from some
descriptive distribution functions.
33 In order to compare the influences due to the choice of the reference design flow rate in
the fixture unit approach on the probable maximum simultaneous demand of an
installation, the stochastic model takes a constant discharge probability pi and a constant
discharging flow rate qi (Ls−1) for a number of appliances Ai=Ai(pi,qi) of the same type;
i=1…ni in the installation. The discharge operation is described by a random number
p*[0,1].
*
 0 ; p  pi
qi  
*
qi ; p  pi
… (8)
In each simulation j, the simultaneous discharging flow rate qd,j (Ls−1) is determined by,
ni
q d , j   q i ; i=1…ni;
… (9)
i 1
The probable maximum simultaneous demand q *d (Ls−1) is determined by the
q (Ls−1) from all
distributions of all simulated simultaneously discharge flow rates ~
d
simulations j=1…ns, where the allowable failure rate  of 1% taken in some practices
adopting the fixture unit approach,
q*d
~
q  F   ;   1  qd dqd

*
d
… (10)
0
The required number of simulations ns can be determined with reference to the
improvement on errors by further simulation steps. Two expressions of errors are used;
the absolute modelling error a is determined by the modelled number of simultaneous
discharging appliances for 99% cases N*, corresponding to =0.01 in Equation (4),
a 
N*  N
N*
; N  Mp  z 2 Mp 1  p 
… (11)
And the relative modelling error at ns simulations expressed by the change of model
output due to an increment of 1 simulation and is given by,
r  1 
N*ns 1
… (12)
N*ns
Regarding a discharge probability p[0.01,0.05], it was reported that the maximum
absolute modelling error a would remain unchanged for simulations ns>10000, the
corresponding relative modelling error r was 0.00810−3.
34 The fixture unit approach at the reference design flow rate qref[1,1000] was used to
determine the probable maximum simultaneous demands of installations which
composed of 2 different appliance types A1 and A2, operating at a residential discharge
pattern as shown in Table 1. The probable maximum simultaneous demands determined
by the fixture unit approach at a reference design flow rate q d ,ref (Ls−1) were then
compared with those q *d (Ls−1) determined by the stochastic model. The percentage
deviation between the probable maximum simultaneous demands  f ,ref is given by,
q

f ,ref   d ,*ref  1 100%
 qd

… (13)
Table 1: Example fixtures
Appliances
Discharge
probability, p
Discharging
flow rate,
(Ls−1)
Washbasin
0.028
0.15
1
1
WC
0.050
0.10
1.12
1.17
Fixture unit at Fixture unit at
q qref=10 Ls−1
qref=250 Ls−1
Figure 4 shows the percentage deviations of the fixture unit approach for an installation
size from 100 to 10000 washbasin-and-WC pairs in residential buildings. In the figure, a
positive value indicates an over-estimate by the fixture unit approach, this overestimation of the probable maximum simultaneous demands would be considered as
satisfactory that the design of not-to-exceed the maximum allowable failure rate =1%.
The results showed that the choice of reference design flow rates had significant
influence on the predicted probable maximum simultaneous demands and hence a wide
range of deviations f,ref (Ls−1) were reported. The deviations varied between −7% to
5%.
Taking the existing practice of using a reference design flow rate of 10 Ls−1 as an
example, the results showed that the fixture unit approach would give satisfactory
predictions of the probable maximum simultaneous discharge flow rates for installation
sizes of 900 residential washbasin-and-WC pairs. Within these range limits, an overestimate by the fixture unit approach at a reference design flow rate of 10 Ls−1 would
not be more than 3% as compared with the ones determined by the stochastic model. It
is noted that the installation sizes for an 80-storey high-rise residential building in Hong
Kong and a housing estate of 40 high-rise residential buildings are about 1200 and
10,000 respectively. An increased reference design flow rate would be required for the
35 Percentage deviation f,ref design criterion of 1% failure probability allowed for water supply systems in buildings.
This study showed the reference design flow rates qref=100 Ls−1 would be adequate for a
residential installation of size up to 10,000. Table 1 gives the example fixture units for
appliances at a reference flow rate of 100Ls−1. The existing fixture units used for some
buildings are shown for comparison. The results suggested that fixture units can be used
for some appliances in high-rise buildings.
6%
1000
2%
25
-2%
10
5 -6%
-10%
qref=1 Ls−1 2.5
-14%
100
1000
10000
Installation appliances M Figure 4: Percentage deviations f,ref of the design flow rates by the fixture unit
approach at reference flow rates qref
Conclusion
Fixture unit approach has been used for estimating the probable maximum simultaneous
demands in water systems for a lot of buildings for many years based on a reference
design flow rate of 10 Ls−1. This paper reported that the selection of the reference
design flow rate would have significant influence on the estimated probable maximum
simultaneous demand. The existing choice of the reference design flow rate would
underestimate the demands of water supply systems in some high-rise buildings, i.e.
more than 1% probability for the demands to exceed the estimated probable maximum
simultaneous demands. The existing assumption of the reference design flow rate
adopted in fixture unit approach would give good estimation for an installation up to
900 WC-and-washbasin pairs in residential buildings. The reference design flow rate
would be increased for larger installations in high-rise buildings so that a good estimate
can be made. This paper presents useful information in for the application of fixture unit
approach in estimating the probable maximum simultaneous demands in water systems
36 of high-rise buildings and enables further studies on water supply system designs for
similar built environment having a high population density.
Acknowledgment
The work described in this paper was partially supported by a grant from the Research
Grant Council of the HKSAR, China (PolyU5305/06E and 533709) and by a grant from
The Hong Kong Polytechnic University (GU551, GYG53).
References
1. Konen T.P. and Goncalves O.M. (1993). Summary of mathematical models for the
design of water distribution systems within buildings. Proceedings of CIBW062
International Symposium of Water Supply and Drainage for Buildings.
2. Hunter R.B. (1940). Methods of estimating loads in plumbing systems, Report
BMS65. National Bureau of Standards, Washington.
3. Oliveira L.H., Goncalves O.M. and Uchida C. (2009). Performance evaluation of
dual-flush WC cistern in a multifamily building in Brazil. Building Services
Engineering Research and Technology, February, 27-36.
4. Mui K.W., Wong L.T. and Yeung M.K. (2008). Epistemic demand analysis for fresh
water supply of Chinese restaurants. Building Services Engineering Research and
Technology, May, 183-189.
5. Wise A.F.E. and Swaffield J.A. (2002). Water, sanitary and waste services for
buildings (5th ed.). London: Butterworth Heinemann.
6. Courtney R.G. (1972). A Monte-Carlo method for investigating the performance of a
domestic water system. Proceedings – 1st International Symposium on Water Supply
and Drainage for Buildings CIBW062 (pp.97-104). 19-20 September, BRE, UK.
7. Wong L.T. and Mui K.W. (2008). Stochastic modelling of water demand by domestic
washrooms in residential tower blocks. Water and Environment Journal, June, 125-130.
8. Plumbing services design guide (2002). The Institute of Plumbing, Essex, UK.
9. Galowin L.S. (2008). “Hunter” fixture units development. Proceedings - 34th
International Symposium on Water Supply and Drainage for Buildings CIBW062
(pp.58-80). 8-10 September, The Hong Kong Polytechnic University, Hong Kong.
10. Wong L.T. and Mui K.W. (2007). Modeling water consumption and flow rates for
flushing water systems in high-rise residential buildings in Hong Kong. Building and
Environment, May, 2024-2034
37 Presentation of Authors
Dr. L. T. Wong is an associate professor at the Department of
Building Services Engineering, the Hong Kong Polytechnic
University.
Dr. K. W. Mui is an assistant professor at the Department of
Building Services Engineering, the Hong Kong Polytechnic
University.
38 I.3
Application of fuzzy logic to the assessment of
design flow rate in water supply system of
multifamily building
L. H. Oliveira (1), L. Y. Cheng (2), O. M. Gonçalves (3), P. M. C. Massolino (4)
(1) Department of Construction Engineering of Escola Politécnica, University
Paulo, Brazil, e-mail: lucia.oliveira@poli.usp.br
(2) Department of Construction Engineering of Escola Politécnica, University
Paulo, Brazil, e-mail: cheng.yee @poli.usp.br
(3) Department of Construction Engineering of Escola Politécnica, University
Paulo, Brazil, e-mail: orestes.goncalves@poli.usp.br
(4) Department of Construction Engineering of Escola Politécnica, University
Paulo, Brazil, e-mail: pedro.massolino@poli.usp.br
of São
of São
of São
of São
Abstract
The traditional approaches to determine the design flow rates of water supply system of
buildings are empirical and probabilistic ones. In Brazil, the empirical formulations are
the most used because its application is recommended by Brazilian standard
NBR 5626/1998. On the other hand, the probabilistic methods have been considered
more suitable because they take into account more realistic conditions of each design,
which is represented as random variables intervening in the use of the water inside the
buildings. Nevertheless, as the usage of some sanitary appliances may depend on the
subjective behaviors of the users, the aim of this work is to present a model that uses
Monte Carlo simulation for random variables and fuzzy logic for fuzzy variables to
achieve a more accurate assessment of design flow rate in water supply system of
multifamily buildings. As the first stage of the study the results of the application of
fuzzy logic to determine the duration of showers, which affects remarkably the flow
rate, is shown here in. For the validation of the model, all fixtures of an apartment with
three students were monitored during a period of 20 days with the use of a water
measurement system in real time. The comparison of the time of shower of one student,
obtained by fuzzy logic with that collected in real time measurement is carried out, and
show good agreement between the calculated and the measured duration of shower.
39 Keywords
Design flow rate; water supply system; fuzzy logic; Monte Carlo simulation.
1.
Introduction
The traditional approaches to determine the design flow rates in building water supply
system can be classified into empirical and probabilistic ones [1]. The empirical
formulations are the most used. However, the probabilistic ones are more suitable
because they take into account more realistic conditions of each design, which is
represented as random variables intervening in the use of the water inside the buildings.
Nevertheless, the probability theory only considers random variables, which depend on
future events. It does not treat fuzzy variables that better represent the behavior of the
user, which is required in a more accurate assessment of flow rates in the feeding
branches.
On the other hand, amount all the sanitary appliances used in multifamily buildings the
shower generates the largest impact in the value of the design flow rate of the
apartment’s feeding branch, in case of water submetering system. This is because of its
longer duration of usage and its larger unitary flow rate. Moreover, the duration of
usage of shower is strongly affected by subjective behavior of the users, which may be
modeled by means of fuzzy approach, instead of the probabilistic ones.
In this way, the aim of this work is purpose a model that applies Monte Carlo simulation
and Fuzzy Logic to establish the design flow rate for the water system of the residential
buildings. While fuzzy logic deals with subjective variables, in especial, to assess the
duration of the shower usage, the Monte Carlo simulation uses the distributions of the
variables for the determination of the instant of use of the sanitary appliances.
As it is very difficult to carry out flow rate measurements for a typical family, the
approach present herein is validated by using the measurements obtain from the
apartment of three students. For the sake of simplicity, only the morning shower is
considered because the peak period clearly defined and the instant of usage is more
critical due to time to leave the apartment for work or study.
2.
Models for the determination of design flow rates in Brazilian
water supply system
According to Gonçalves [1], the empirical models commonly used for the assessment of
the design flow rates in the cold water supply system include those technique based on
the use of tables, graphs and empirical formula established from the experience and
judgment of its authors. Meanwhile, the probabilistic models are based on the weights,
graphs and mathematical expressions derived from probabilistic concepts.
In the following sections, an overview of these two models is presented as well as the
basic concepts of Fuzzy Logic, which is an approach adopted proposed herein to
40 development a more accurate model for the assessment of design flow rates of water
systems.
2.1 Deterministic model
The Brazilian standard NBR 5626/1998 [2] recommends the use of a deterministic
method for the assessment of the design flow rates in cold water system, whose
expression is given by Equation (1):
Q  0,3
P
(1)
Where:
Q is the design flow rate in a given section, L/s.
∑P is the sum of the relative “fixture units” of all the fixtures, installed downstream of
the pipe section.
Despite of its simplicity, this method does not take into account the influence of the
users’ activities, which, in turn, are function of:
 the type of the building and the characteristics of the user;
 the characteristics of the building, which is defined for the size and the distribution of
the population and,
 the characteristics, flow rates and intensity of use of the sanitary appliances.
2.2
Probabilistic model
In order to consider the real conditions of each design situation Gonçalves [3] proposed
a model for the determination of the water demands in the water supply systems, which
is briefly shown as what follows.
In building water supply systems the flow rates depends on the interaction between the
user and the sanitary equipment system, and are affected by the following factors:
 activities of the users, which is a function of:
o the type of building (residential, school, hotel etc.);
o the characteristics of the users, determined for physiological, regional,
cultural, social and climatic aspects;

characteristics of the building, which is a function of:
o the population (size and distribution) and
o the space organization;

characteristics of the set of sanitary appliances, which is a function of the types and
the number of sanitary appliances.
The intervening variables, which consider the above mentioned factors are grouped
in the model as follows:

intensity of use of the set of sanitary appliances;

unitary flow rates of each type of sanitary appliance.
The intensity of use of the set of sanitary appliances is considered in the model by using
the following variables:
41 
duration of the discharge of a sanitary appliance, t;

interval of time between consecutive discharges of a sanitary appliance, T;

number of sanitary appliances installed downstream of the pipe section, n.
The unitary flow rate of a type of sanitary appliance is denoted by q.
The duration of the discharge of an appliance (t) consists of the period between the
beginning of the discharge and the end of the water supply for the discharge. It can be
determined by in loco measurements and calculation mean and the variance of a set of
data or by the three points estimative method: a minimum value (tmin), a most probable
(tprov) and a maximum (tmax), using the Gamma distribution, according to Gonçalves [3],
the mean (μt) and the variance (σ2t), may be determined by Equation (2) and (3),
respectively.
t 
t mín  3  t prov  t máx
5
(2)
2
σt 
(t máx  t mín ) 2
25
(3)
Similarly, the unitary flow rate of each appliance (q) can also be determined by means
of field surveys and its mean value and variance might also be calculated by using the
three points estimative method.
The interval between two consecutive usages (T) depends on several factors, and can be
represented by the following variables:
 number of appliances of the considered type installed in the sanitary room, n;
 number of usages per capita of a type of appliance during the peak period, u;
 population served at the sanitary room in which the appliance is installed, P.
The number of uses per capita of each type of sanitary appliance in the peak period (u)
can be determined similarly to the ones presented for the unitary flow rate and for the
duration of the discharge.
Despite the probability model gives more precise results than deterministic ones, the
usage of sanitary appliances is strongly dependant on the users, whose subjective
behavior may be better modeled as fuzzy variable [4]. Therefore, beside the
deterministic and probabilistic models, approaches based on the Fuzzy Logic might be
effective for the assessment of the flow rates, especially of the showers, whose usage is
highly subjective.
In this way, in order to develop a more reliable and accurate method to estimate the
duration of the usage of the sanitary appliances, the present work applies the Monte
Carlo simulation to model the random variables such as the instant of usage of sanitary
42 appliances and applies Fuzzy Logic to model the behavior of the users as a set of fuzzy
inference rules for the usage of shower, which is particularly related to the instant of
usage and air temperature. The approach adopted herein is presented in the following
sections.
2.3
Monte Carlo simulation
Monte Carlo simulation [5] is considered in the present research to take into account
random events of the use of the sanitary appliances such as the wash basin, water closet,
sink, sink laundry and washing machines. Different from the shower, as will be shown
in the next sections, the use of these sanitary appliances does not show clear relation to
some intervening variables.
2.4
Fuzzy logic
Of all the sanitary appliances used in residential buildings, the shower generates greater
impact in the value of the design flow rate of a section, due to the longer duration of use
and relatively high unit flow rate. Moreover, the behaviors of its users are quite
subjective, which might be better modeled by using Fuzzy Logic. This fact motivated
the authors of this work to use the fuzzy logic to determine the duration of use of the
shower during the bath of users.
According to Von Altrock [6], fuzzy logic is a technology that translates natural
language descriptions of decisions policies into algorithm using a mathematical model.
It is a reasoning system that involves fuzzy propositions and the procedure of the fuzzy
inference consists of three major steps: fuzzification, inference and defuzzification, as
presented in Figure 1.
Input variables
(li
i i
Results Fuzzy inference
i bl )
(li
i i
i bl )
Linguistic Defuzzification
Fuzzification Numerical level
Input variables (
i l
Output variables
i bl )
(
i l
i bl )
Figure 1: Structure of a fuzzy logic system [6]
Fuzzification: It is the initial step, where the fuzzy sets are used to translate numerical
variables into linguistic variables, and the membership degree of the associated
linguistic values are obtained.
43 Fuzzy inference: After translating all input numerical values into linguistic variable
values, the fuzzy inferences are carried out by applying the if/then fuzzy rules that
define system behavior. This step yields a linguistic value for the linguistic variable. For
example, the linguistic result for “duration of shower” (linguistic variable) could be
“very fast” (linguistic value).
Defuzzification: In this step, the fuzzy variable obtained by the inference rules is
converted into a discrete numerical value that better represents the inferred values of the
linguistic variable, which is the output of the fuzzy inference. There are many
desfuzzification methods available in the literature. However, according to Cox [7],
method of baricentre and method of average of the maximum are the most used ones. In
this work the method of baricentre has been adopted.
3.
Methodology
This section shows the methodology and the fuzzy approach adopted herein. The data
obtained in the apartment of the students will be used for the establishment of the
distribution of the following variable: average values of flow rates, average duration of
use and schedule of the beginning of use of the sanitary appliances. It is highlighted that
the duration of the showers is obtained by fuzzy logic.
Since it is very difficult to obtain authorization to install the measurement equipments in
loco because the cabling and the sensor bother the users, as an alternative,
measurements have been carried out in an apartment where three students live.
3.1
Model using Monte Carlo simulation and Fuzzy logic
Figure 2 shows the flowchart of the model to be developed. The following variables will
be used:
 deterministic – quantity of sanitary appliances (n);
 random – average flow rate of sanitary appliances (qm); average duration of use of
sanitary appliances (tm) except the electric shower, which will be obtained by fuzzy
logic; instant of beginning of the activities related to the water use.
 fuzzy – linguistic values of the variable “instant of shower”, “air temperature” and
“duration of shower”.
By means of the fuzzy logic determines the duration of the shower and with the use of
Monte Carlo simulation is determined the start of use of the sanitary appliance.
Based on the quantity and type of sanitary appliances and the users, Monte Carlo
simulation is carried out to determine the instant of usage of all sanitary appliances by
each user. The duration of usage is also determined by Monte Carlo simulation, except
in the case of shower. Because of its longer duration and the influence of the users´
behaviour, the duration of shower is determined by using fuzzy logic as explained in the
section 3.2. In the case of overlapping of the shower usage (more than one user at the
44 same time), a strategy that establishes the priority amount the users is adopted to solve
the conflict.
Start
Monte Carlo simulation Instant of appliance use for “n” n, n ‐ 1, n ‐ 2…n ‐ n Queuing with priority Redefining of instant of appliance use
Fuzzy logic N Is queuing ended? Y Determination of flow rate x time Figure 2: Step of the model to determine the design flow rate using Fuzzy Logic
and Monte Carlo simulation
45 3.2 Measured duration of shower
At first the duration of shower times of each user have been measured in a period of 20 days, and with the use of a real time water measurement system, constituted of a volumetric
water meter class C, nominal flow rate of 1.5 m3/h, connected to a Cyble pulse and a
data acquisition system (MGCplus – HBM).
The local where the measurements have been carried out was an apartment in the city of
São Paulo-Brazil. Three male students live there. One is a under graduating student, one
is master degree student and another one of PhD degree student. These students are
from two regions of Brazil: North and Northeast, which present different climatic
conditions, and both hotter than São Paulo. This might influence the habits of the users
and the duration of showers. The apartment has three dormitories, one bathroom with
one electric shower, one close-coupled toilet with nominal volume of discharge of 12
liters (WC) and one wash basin. The climatic conditions during the period of data
collect was very hot, with maximum and minimum air temperature of a day ranging
from 16 oC to 34 oC, and the temperature in the morning varied from 20 oC to 22 oC.
3.3 Determination of the duration of shower by fuzzy logic
Figure 3 shows an overview of the fuzzy reasoning system for the assessment of the
duration of the morning shower.
All students leave the apartment in the morning and come back only at the end of the
day. The peak period for the use of the shower in students’ apartment is from 7:30 to
10:00 a.m.
For the sake of simplicity, amount the variables that affect the duration of shower in the
peak period in the morning, two main variables “instant of shower” and “air
temperature” are considered herein. Based on these two variables, the set of linguistic
variables for “instant of shower”, “air temperature” and “duration of shower” are
defined by interviews applied to the users.
46 Output Input Rule 1: If the user takes a shower very early and the temperature is very cold then... 1.Instant 2.Temperature Rule 2: If the user takes a shower very late and the temperature is cold then... R l 3 If h
i
i
1. A very fast shower. 2. No shower. 3. A normal shower. d
Figure 3: Fuzzy reasoning system to assess the duration of the morning shower as
function of instant and air temperature
The questions carried through in the interview have been:
•
Which the air temperatures you consider very cold, cold, pleasant, hot, very hot?
• During the peak period, what are the time intervals associated with the following
expressions: very early, early, in time, delayed and very delayed.
• Associate the duration of your shower, in minutes, with the following expressions:
very fast, fast, normal and long.
3.3.1 Linguistic variables
The interviews generates data that present some scattering, typical of personal
subjectivity of the users. Thus, mean value obtained from the three users have been used
to establish the sets of fuzzy linguistic values corresponding to “instant of shower”, “air
temperature” and “duration of shower”, which are shown in Figure 4, 5 and 6
respectively.
47 very early early on time delayed very delayed 7:30 8:00 8:30 9:00 10:00 Figure 4: Linguistic values of the variable “instant of shower” of the students
very cold cold pleasant hot very hot 15oC 20oC 23oC 27oC 30oC Figure 5: Linguistic values of the variable “air temperature” of the students
no bath very fast fast normal long 0 min 3 min 5 min 8 min 12 min Figure 6: Linguistc values of the variable “duration of shower” of the students
48 3.3.2 Rules of the fuzzy reasoning
Also, from the interviews, a matrix of fuzzy reasoning rules is obtained for each user.
As the matrices present different pattern due to different habits of the users, they are
shown separately in Tables 1, 2 and 3.
Table 1 - Fuzzy matrix for the establishment of the shower time of the user A
Time very early early on time delayed very delayed very cold normal normal normal fast very fast cold normal normal normal fast very fast pleasant long long long fast very fast
hot long long long fast very fast
very hot long long long fast very fast
Temperature Table 2 - Fuzzy matrix for the establishment of the shower time of the user B.
Time very early early on time delayed very delayed very cold fast fast
fast
very fast no shower
cold fast fast
normal
very fast no shower pleasant normal
normal
normal
fast very fast hot long normal normal fast very fast very hot long long normal fast very fast Temperature 49 Table 3 - Fuzzy matrix for the establishment of the shower time of the user C
Time very early early on time delayed very delayed very cold long long normal fast very fast cold long normal normal fast very fast pleasant normal normal normal fast very fast hot fast fast fast fast very fast very hot fast fast fast fast very fast Temperature 4.
Results and discussions
Results obtained by measurements of the flow rate and duration of the use of the water
closet (Figure 7) and the use of wash basin (Figure 8) with their respective distribution
of use of these sanitary appliances by the three students.
Frequency (%)
15
0,30
10
0,25
5
Flow rate (L/s)
0,20
0,00
Fri 6Mar 2009
21:54
21:55
21:56
21 ‐ 22
18 ‐ 19
15 ‐ 16
12 ‐ 13
9 ‐ 10
6 ‐ 7
0,05
3 ‐ 4
0,10
0 ‐ 1
0
0,15
Time (h)
21:57
Time (h:min)
Figure 7: Water closet flow rate profile and respective distribution of water
closet use
Frequency (%)
0,06
Flow rate (L/s) 0,05
0,04
0,03
0,02
0,01
0,00
8:02
Sat 7 Mar 2009
8:03
8:04
8:05
Time (s)
8:06
8:07
15
10
5
0
0 ‐ 1
3 ‐ 4
6 ‐ 7
9 ‐ 10
12 ‐ 13
15 ‐ 16
18 ‐ 19
21 ‐ 22
0,07
Time (h)
8:08
Figure 8: Wash basin flow rate profile and respective distribution of wash basin
use
50 One of the results obtained by measurements of the duration of the bath of the student A
is presented in Figure 9. It gives a measured time series of the flow rate due to the
shower usage of the student A. The data show the duration of the shower. Since only the
student A has the habit of taking shower in the morning, the validation of the duration
of shower determined by fuzzy logic has been carried out using his data.
0,14
0,10
0,08
0,06
0,04
Flow rate (L/s)
0,12
0,02
0,00
8:43
8:44 Fri 27Feb 2009
8:45
8:46
8:47
8:48
Horário [s]
8:49
8:50
8:51
8:52 Time (h:min) license
catmanEasy Presentation
Figure 9: Shower profile of the student A collected in field
For the calculation of the duration of shower, by using Mamdani relation, a software in
Java language has been developed. The results are giving in Figure 10 for the student A.
As mentioned before, the student A had the greater number of shower during the
morning, therefore measured data that corresponds his showers have been used for the
validation of the fuzzy approach.
Figure 10: The duration of morning shower of the student A obtained by the
fuzzy approach
51 Table 4 presents the time instant and the air temperature of the events recorded by the
measuring system regarding to student A.
Table 4 - Instant of the shower of the student A and the corresponding air
temperature obtained by measurement
Shower event Time (h:min) Temperature (oC) A 7:33 22 B 8:44 22 C 8:49 22 D 8:57 22 E 9:10 22 F 9:27 22 Figure 11 shows the duration of the shower obtained from in loco measurements and
that obtained by means of the fuzzy logic for the cases shown in Table 4. Since the
measurement has been realized in a period whose patterns of daily temperature variation
are almost the same, all the temperatures recorded in the peak periodo are very close to
22 oC.
Real shower
Shower time (min)
10
9
8
7
6
5
4
3
2
1
0
Fuzzy shower
B
C D
D
C
E
F
A
7:30
8:00
8:30
9:00
Time (h:min)
9:30
10:00
Figure 11: Shower times of the sudent A obtained by measurement in field and
in real time by fuzzy logic
52 In Figure 11, it is observed that the measured and the calculated durations of the shower
for the event A are quite different. There are two reasons for the discrepancy: the first
one is that this shower was much more early than usual time of the student A and the
second is that this shower have occurred in the weekend while the others had been in
normal days of activities.
Thus, without considering this atypical event, which is expected because of some
random events, the calculated duration of the shower of all other cases agrees well with
the measured ones. This result gives some insight about the effectiveness of the
approach based on Fuzzy Logic to estimate the duration of the usage of the sanitary
appliances.
5.
Final considerations
This paper presented the first stage of a research for the determination of design flow
rate in a water supply system of multifamily buildings. This model considers Monte
Carlo simulation for random variables and Fuzzy Logic to assess the duration of the
shower in a residential apartment, due to it better represent the behaviour of the users.
This first stage is focused on the assessment of the duration of shower by using fuzzy
logic.
The fuzzy approach was validated by available measured data, and the results show
good agreements between the measured and calculated ones. Thus, one of the great
advantages of the approach is the ability of taking into account the influence of the
subjectivity of the user on the usage of the sanitary appliances, mainly the showers.
In the second stage of the research, the simulation of the random events of the usage of
all sanitary appliances will be carry out by using Monte Carlo simulation. The model
will be validated by data of another apartments but the approach is very effective and
serves as a basis for the development of a more accurate method to assess the design
flow rate for the water supply system of the residential buildings.
Acknowledgements
The authors thank the Fundação de Apoio à Pesquisa do Estado de São Paulo - Fapesp
for the financial support the research and the students of the monitored apartment.
53 6. References
[1]
[2]
[3]
Gonçalves, O.M. Contribuições para a Economia e Qualidade dos Sistemas
Prediais, Memorial circustanciado (Livre-docência junto ao Departamento de
Engenharia de Construção Civil) - Escola Politécnica da Universidade de São
Paulo, Brasil, 1997.
Associação Brasileira de Normas Técnicas, NBR 5626: Instalação predial de água
fria. Rio de Janeiro, Brasil, 1998.
Gonçalves, O.M. Formulação de modelo para o estabelecimento de vazões de
projeto em sistemas prediais de distribuição de água fria, Tese (Doutorado em
Engenharia Civil) – Escola Politécnica, Universidade de São Paulo, São Paulo,
Brasil, 1986.
[4]
Kiatake, M.; Cheng, L.Y.; Petreche, J.R.D. Knowledge Representation of Architectural Design: A Case Study by Using IBIS‐FRS System. Journal of Advanced Computer Intelligence and Intelligent Informatics, Fuji Technology Press, Japan, 9(6), 677‐683, Nov. 2005. [5]
Shamblin, J.E.; Stevens Jr, G.T. Operation research – a fundamental approach.
McGraw-Hill Book Company. New York, 1974.
[6]
Von Altrock, C. Fuzzy‐logic and neurofuzzy applications in business and finance, Prentice Hall PTR, New Jersey, United States, 1997. Cox, E. Fuzzy logic for business and industry, Charles River Media Inc, Massachusetts, 1995. [7]
7. Presentation of Authors
Lúcia Helena is a professor at Department of Construction Engineering of Escola Politécnica of
University of São Paulo where she teaches and conducts researches on building services.
Liang Cheng is a professor at Department of Construction Engineering of Escola Politécnica of
University of São Paulo where he teaches and conducts researches on fuzzy analysis applied to
designs, CFD simulations and graphical geometry for Engneering.
Orestes Gonçalves is a professor at Department of Construction Engineering of Escola
Politécnica of University of São Paulo where he is the Head of the Building Services Research
Group.
Pedro Massolino is a graduate student at engineering, Escola Politécnica of University of São Paulo. 54 I.4
The evaluation of the per capita water
demand of the University of São Paulo
throughout ten years of the Water C
Conservation Program of the University of
São Paulo
G. Silva, H. Tamaki, G. Correia, O. Gonçalves
gisele.silva@poli.usp.br, humberto.tamaki@poli.usp.br,
gabriele.correia@poli.usp.br, orestes.goncalves@poli.usp.br
Depart. of Civil Eng., Polytechnic School, University of São Paulo, Brazil
Abstract
During the twelve years of its existence, the Water Conservation Program of the
University of São Paulo - PURA-USP - became a reference as a program of water
demand management at university campus. According to the system thinking,
throughout this period, technological activities (leaks elimination, replacement of
conventional sanitary appliances by water saving ones and minimization of wastes in
processes), mobilization activities (program spreading, awareness campaigns and
trainings) and water demand management activities (demand follow-up and actions over
water consuming systems) have been accomplished. In this study, the water demand is
evaluated in 29 educational, research and administrative Units of the University
between 1998 and 2007. The main result of the PURA-USP, throughout these ten years,
is the reduction of 40% of the water demand (from 135,999 to 82,207 m³/month). This
becomes even more expressive considering growth of 12% of population and 16% of
built area. Thus, the per capita water demand was reduced by 46% (from 113 to
61 L/cap/day), 5 Units had their demand increased and the remaining 24 units had their
demand decreased (ranging from 1% to 76% of reduction). It is also important to
highlight that in 12 Units the per capita water demand was reduced by at least 50%. In
some Units, other indicators were also evaluated, for instance, the demand per hospital
bed at the University Hospital, in which the reduction was 20% (from 2,138 to
55 1,710 L/ hospital bed/day), and the demand per student at the School of Application, in
which the reduction was 27% (from 112 to 82 L/student/day). The evaluation of these
indicators is essential for the management of the water demand and, consequently, for
the water conservation, making possible not only the evaluation of the results of the
actions, but also the definition of priorities for the demand reduction and even for the
development of forecasting models of future demands.
Keywords
Water conservation; water demand management; per capita water demand
1. Introduction
The establishment of populations near water bodies has increasingly shown its
deleterious effects and changed the perspective of the water problem, formerly limited
only to its scarcity in arid areas. Some habitation centres have grown, become
urbanized, intensified and diversified their water uses, and consequently reduced the
availability of water volume. In 1996, the São Paulo State Water Utility Company
(SABESP) in agreement with the Polytechnic School of USP (EPUSP) and the Institute
for Technological Research (IPT) has started the Water Conservation Program (PURA),
for meeting demands of actions at the building systems level and introducing itself in
the change of the paradigm of exclusive water offer management to one for water
demand management too. In 1997, as a case study in a university campus, the Water
Conservation Program of the University of São Paulo - PURA-USP - has started (Silva;
Tamaki; Gonçalves, 2002). In this paper, after the introduction of PURA-USP and its
actions, it is evaluated the per capita water demand evolution, important to the demand
management and to the results evaluation of a water conservation program.
2. The University of São Paulo and the PURA-USP
The University of São Paulo (USP), founded in 1934, the largest public university in
Brazil, is responsible for 28% of scientific production of the country. USP develops its
activities in education, research and extension programs in various campi, distributed in
six cities in the state of São Paulo, other than the capital where the main campus is
located - The University City Armando de Salles Oliveira (CUASO). The University
has 229 undergraduate courses, dedicated to different areas of knowledge, distributed in
40 Units and offered to approximated 56 thousand students. As for the research, there
are 230 graduate programs, containing 22 thousand attending students, being
responsible for the graduation of a quarter of the doctors in Brazil (USP, 2009).
56 In this work, it is presented the PURA-USP as implemented at the campus CUASO. By
means of an agreement between USP and SABESP, the University undertook to keep
payments of water and sewage charges up-to-date and to implement the PURA in
exchange for a concession of a discount of 25% on the charges.
2.1. Planning
The implementation of PURA-USP was preceded by a phase of planning (Silva, 2004):
 Surveying motivation: in 1997, costs of water supply and sewage disposal amounted
to US$ 1.66 million/month considering USP as a whole. This excessive value and the
interest of the University in promoting research justified the execution of PURA-USP.
On the other hand, it was an interest to the public concessionaire to operate a program
of this nature and allow a greater availability of water for the metropolitan region of São
Paulo, in which the availability of water is 200 m³/cap/year, much lower than the value
of 1,500 m³/cap/year adopted as critical by the United Nations (UN).
 Determination of purposes: besides reducing the water demand, the implementation
of the PURA-USP had as goals to maintain the reduced demand all along, to implement
a structured system for water demand management and to develop a methodology that
could be applied to other places later on.
 Deepening of the diagnosis of the situation: in 1998, the CUASO had fixed
population - 54,886 people, built area - 739,073 m², about 200 buildings (mostly built in
the 60’s and 70’s), water external networks - 36,837 m, 467 storage tanks, 19,181 points
of use and average water demand - 135,999 m³/month. The campus water supply is
provided by SABESP mains. It should be noted that the CUASO besides having wide
open areas, many built area, large population and water demand, it has great diversity of
uses - schools, laboratories, hospital, lodging, restaurants, sports club, museums,
theatres, offices, car parks, green areas and areas for animal raising. Since the
construction of the campus, its systems have not undergone significant intervention and
were found different designs, installations and states of conservation.
 Program organization: an organization for work was adopted comprising USP
Rectory; PURA-USP Commission (with technical coordination of the EPUSP) and the
PURA-Unit Commissions - one professor or researcher, one administrative employee
and one maintenance employee.
In order to determine priorities, the Program pre-execution included the finding of the
major consumers, types of use and location of Units, thus determining the execution
phases. The implementation of PURA-USP on the campus was carried out in two
phases: Phase 1 (7 Units - 1998 and 1999), first aiming at the largest water consumers
(more than 50% of the campus) that presented different use typologies and Phase 2 (21
Units - 2000 and 2001). Since 2002, PURA-USP has started actions at Units external to
the campus, while still maintaining permanent activities at the CUASO.
57 2.2 Execution
The implementation of PURA-USP comprised (Silva, 2004):
 Stage 1 - General diagnosis: deepening of the survey of characteristics of the
institution, occupancy, constructions, building systems and water uses. In this stage,
8,302 water use points were recorded in Phase 1 and 10,879 in Phase 2, which enabled
us to evaluate the amount of work.
 Stage 2 - Physical losses reduction: updating of the registries, detection (by means of
listening stick, geophone, and leak noise correlator) and elimination of leaks in external
water supply networks and in storage tanks (especially in ball valves and overflow
devices). Plenty of leaks were detected in the networks stating the precariousness of the
buried pipes, some subjected to pressures up to 700kPa.
 Stage 3 - Consumption reduction at the water use points: detection and elimination
of leaks at use points - leakages were found in 2.5% of total sanitary appliances (mainly
in urinals and flushing valves) - and replacement of conventional appliances for water
saving models, adopting replacement and specification guidelines that comply with the
standards, quality programs and users’ requirements - self-closing taps, swinging spout
taps with articulated aerators, low flushing toilets and self-closing valves for urinals.
 Stage 4 - Characterization of habits and rationalization of water-consuming
activities: surveying users’ habits in activities that take place in kitchens, laboratories,
gardens (watering) and in general cleaning and in places where there is specific use of
water. Information on more efficient procedures is provided in order to minimise waste
without losing quality of use. In order not to provide erroneous solutions, it is necessary
to become familiar with users’ activities, to understand how the systems work and to
check the available technology. In laboratories, it was stated a high waste of water
during water distillation processes. It was verified, at the CUASO, 240 distillers,
representing a loss of almost 3,000 m³/month. There was, however, a variation from 10
to 200 L of cooling water lost for each litre of distilled water produced. Tamaki et al.
(2007) observed that the correct adjustment of the water input contributes to the
minimization of losses. The replacement of distillation for technologies such as ion
exchange and reverse osmosis, whose loss of water is minimal (if not null), is seen as
more permanent solutions.
 Stage 5 - Program spreading, awareness campaigns and trainings: in order to reach
both the primary users (maintenance employees) and the final users (professors,
students, employees, visitors), it can be highlighted the preparation of advertising
materials (folders, stickers, posters, newspaper articles), creation of e-mail
(pura@poli.usp.br) and a page on the Internet (www.pura.poli.usp.br), training sessions
and the participation in technical events.
At the post-execution phase, in order to guarantee the reduced levels of water demand
or even increase the reduction, the permanent character of Stages 4 and 5 shall be
established, as well as the water demand management. The follow-up of demand
throughout these twelve years enabled us to ratify the importance of the permanent
nature of PURA-USP.
58 2.3 Water demand management
Water demand management, whose goals are efficient use and water saving, goes
beyond the mere follow-up of consumption and includes, besides the data organization
and surveying of graphs, the data evaluation, the determination of control parameters
(monthly demand, per capita demand, flow rate profiles, e.g.) and the system feedback
in the form of elimination of a leak or revision of processes (Tamaki, 2003).
After verification of the SABESP water connections (about 120, which capacities range
from ¾ to 4 inches) and the places by them supplied, the efforts of collecting consume
data were continued. The permanent monitoring of water bills and regular contact with
Units, Department of Finance of USP’s Rectory and SABESP, intending to verify
elevated demand or possible errors have been important.
For the demand management at the CUASO, Tamaki (2003) and Tamaki, Silva,
Gonçalves (2004) present simple, such as water bills and in loco readings, and more
sophisticated instruments such as:
 Submetering: installation of more water meters (about 60 units), besides those used
for billing, which allows greater correspondence between demand and consuming unit,
a more precise location of leaks and charging the water used by private luncheonettes.
 Remote reading: acts on the data collecting - a more reliable acquisition of a greater
amount of information pieces, besides a faster detection of anomalies; as in the
determination of control parameters - providing flow rate profiles along the day. The
CUASO system is a digital field bus system - M-BUS standard - including electronic
water meters, communication network and a remote reading management centre
(Tamaki et al., 2005). In July 2009, the implemented remote reading system had about
100 water meters.
By concluding the management cycle, there should be intervention procedures and
accountability in case of anomalies in the water demand. It is recommended the
recording of leaks that occurred, to make it possible to check the areas more subject to
leaks, the more recurrent problems, the materials that show more defects, etc.
3. Results of the Water Conservation Program
Twelve years on, since PURA-USP on CUASO was implemented, a series of effects
could be seen in addition to the big impact it caused - the reduction of the water demand
(Silva, 2004):
 Changes in the systems of cold water supply and sanitary appliances: up-to-date
registry of the networks, storage tanks and water meters; restorations and adequacy of
cold water systems; restoration and modernization of sanitary appliances, including
replacement for saving models - it can also be seen an increase in functionality,
uniformity of models and application of quality concepts (compliance with the Brazilian
Program of Quality and Productivity of the Habitat - PBQP-H).
 Changes in the routine of building maintenance: training of teams; reduction of calls
and loss of materials; and creation of a feedback cycle between water managers and
59 maintenance staffs, with consequential fast elimination of leaks. According to Silva et
al. (2007), the Program is an opportunity for correcting maintenance faulty points
(regarding to characteristics of the phases of the project, construction and
use/operation), as well as for guiding future constructions. Between August 2001 and
March 2009 it was verified that 277 leakages occurred, being 31% of them at the
building mains. The types of leakages found show the importance of counting on better
qualified maintenance staff: countless events resulting from badly constructed systems
and improper repair work of old leaks.
 Changes in administrative routines: creation of patterns for acquisition of water
saving appliances - descriptions, standards and quality programs to be followed; and
closer relation with administrative staff, which enables helping with issues relating to
water and its conservation.
 Changes in design parameters: among the recommendations of the Program, the
installation of saver appliances both for retrofitted and new buildings has been widely
adopted. In the future, it is expected that the sizes of the water and sewage systems will
also be reviewed which would also reduce the cost of execution of buildings. Silva et al.
(2007) alert that it isn't enough to detect and eliminate the leakages, being necessary the
minimization of its occurrences from the first. Thus, the projects and executions in the
building systems have been feedbacked, based on the problems verified during
operation and maintenance.
 Technological development of appliances: through partnerships with manufacturers,
when a problem occurred systematically, due to a flaw in the design, a technician of the
design sector could be called and the design could be revised. Still, tests were conducted
with innovative appliances in places of restrict and controlled use - waterless urinals and
toilets with dual flush systems.
 Introduction of alternative water sources: introducing alternatives water sources
should be encouraged, however, appropriate care must be taken to avoid compromising
both the populations’ health and the development of activities and cause environmental
unbalances. The use of well and rain water and the reuse of water lacks studies and
legislation. Its introduction depends not only on technical solutions but also on
procedures and accountabilities for maintenance and management of the systems. As
examples at the CUASO, may be cited: use of water wells for research, use of water of
the rowing lane for practice of sports and its reuse for garden and football field
irrigation, and the reuse of cooling water from the distillers by some laboratories.
 Users’ behavioural changes: end users were receptive to adoption of water saving
appliances. A striking change in user habits is also seen in attitudes like the introduction
of a system for water recirculation, notices warning about leakages and incorrect use,
and search of information on how to cooperate. In order to guarantee the continuity of
the behavioural changes, it is important to emphasise the need to inform the user about
the results attained. The campaigns shall be permanent and focused not in a momentary
lack of water, but in the actual possibility of water scarcity.
The CUASO’s water demand evolution, presented at the Table 1 and Figure 1, allows us
to testify, as the greatest impact of PURA-USP, its expressive reduction. Even after the
conclusion of the implementation of Stages 1 to 3 of the Program, further reduction can
be observed, which confirms the importance of the permanent character of the Program
and mainly of the water demand management.
60 140
Phase 1
Phase 2
Water demand (1,000 m³/month)
120
100
80
60
40
20
0
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
Period
Figure 1 - Evolution of CUASO’s water demand
Table 1 - Evolution of CUASO’s water demand, population and built area
Unit 1998 1999 2000
2001
2002
2003
2004
2005 2006 2007
Water Demand (m³/month) Phase 1 80,224 59,440 48,657 44,699 45,726 44,604 39,578 38,652 38,499 39,848
Phase 2 55,775 55,775 58,120 47,691 47,445 45,935 40,898 40,348 43,825 42,359
Total 135,999 115,215 106,777 92,390 93,170 90,539 80,475 79,000 82,324 82,207
Population (people) (1) (2) Phase 1 31,518 31,778 31,734 32,399 34,787 35,436 35,545 35,785 35,695 35,517
Phase 2 23,369 23,177 23,319 23,056 24,102 24,801 24,852 25,584 25,626 25,842
Total 54,886 54,955 55,053 55,455 58,889 60,237 60,397 61,369 61,321 61,359
Built Area (m²) (1) Phase 1 340,573 340,573 340,573 342,698 343,716 343,868 377,237 377,237 378,794 384,696
Phase 2 398,501 398,501 402,653 414,984 418,795 422,105 462,834 460,887 464,651 469,975
Total 739,073 739,073 743,226 757,682 762,511 765,973 840,071 838,124 843,445 854,671
(1)
(2)
Source: USP (1999-2008).
Undergraduate and graduate students, professors and employees.
Between 1998 and 2007, the CUASO’s demand had a reduction of 40%, ranging from
135,999 to 82,207 m³/month. Considering the Units from Phase 1, this reduction was
even more expressive, reaching 50%. The total amount of water saved in this period was
4.8 million m³, which could supply 240 thousand homes for a month. These data
become even more significant when considered the increase of 12% of the fixed
population and 16% of the built area in the CUASO, in this same period, as changes
also presented at Table 1.
Finally, as for the financial impact, there is expressive difference between the
US$ 20.0 millions that should have been expended with water and sewage services, in
2007, and the US$ 9.8 millions that were effectively expended (considering USP as a
whole). The accumulated economic net benefit was of US$ 77.6 millions in these ten
years.
61 4. The evaluation of the per capita water demand
The time-based evaluation of control parameters, such as monthly demand, daily per
capita demand and flow rate profiles, allows not only to alert a situation of consumption
anomaly, but also the assessment of impacts of the interventions and, consequently, of a
Water Conservation Program.
In this paper, the evolution of daily per capita water demand, between 1998 and 2007,
of each CUASO’s Units will be evaluated. In Table 2 and Figure 2, the daily per capita
water demand of CUASO’s Units, in 2007, is ranked in ascending order of demand.
There is a wide range of demand values, with values from 7 to 1,533L/cap/day, which
represent the great diversity of water uses resulting from various activities at the
University.
Table 2 - Evolution of the daily per capita water demand of CUASO’s Units
Unit (1) Use 1998 1999 2000 2001 type(2) Values up to 40 L/cap/day (2007) (6) FFLCH H 15 14 11 12 FAU H 21 24 13 14 IME H 40 43 44 23 ECA H 48 41 54 33 FEA H 31 22 21 23 FE H 40 41 39 34 IAG H 21 21 26 26 IB M 73 84 60 57 MAE H 127 118 180 140 (3)
2002 2003 2004 2005 2006 2007
WD
P (4) (5) PCWD 10
14
17
28
20
33
25
52
152
9
16
16
31
19
38
17
47
116
7
13
12
29
16
43
16
40
34
6
15
12
29
20
35
12
34
27
6
13
17
20
21
31
30
31
31
2,321
763
992
1,351
1,568
1,346
520
1,145
182
15,873 7 2,526 14 2,924 15 3,933 16 4,192 17 2,043 30 760 31 1,641 32 228 36 Values between 41 and 100 L/cap/day (2007) (6)
IO M 71 71 45 40 48
IF M 64 63 80 47 81
FCF L 211 98 79 72 60
EP M 98 70 67 55 57
MAC H 119 125 178 119 125
IGc M 165 159 132 98 90
EEFE M 46 47 58 55 43
CCE H 222 168 189 58 55
RUSP H 242 247 260 169 121
IP H 93 90 155 107 51
FO M 73 70 79 63 60
47
34
76
51
130
67
36
55
111
72
72
37
22
69
54
124
43
40
56
83
64
64
35
21
57
58
54
70
33
81
92
62
74
39
54
54
58
67
80
42
78
85
81
84
534
2,853
2,035
11,044
138
1,352
1,360
310
2,056
2,880
3,266
580 2,395 1,703 8,936 101 898 903 202 1,101 1,465 1,506 42 54 54 56 62 68 68 70 85 89 99 Values between 101 and 200 L/cap/day (2007) (6)
FMVZ L 136 135 169 136 139
IQ L 573 276 240 217 184
CCS H 393 398 397 249 177
IEE M 89 90 153 220 155
ICB L 404 340 250 213 244
149
218
170
141
216
114
192
178
98
191
96
152
195
116
187
107
138
219
127
185
3,986
4,741
541
398
5,773
1,348 1,553 174 124 1,332 134 139 141 146 197 62 Values above 201 L/cap/day (2007) (6)
PCO O 311 322 347 223
HU O 489 500 361 323
COSEAS O 1,120 1,175 963 926
CEPEUSP O 1,049 1,118 1,202 1,901
182
346
959
2,177
193
302
935
2,234
189
252
823
2,330
203
246
851
1,879
251 261 784 2,082 2,246 453 12,056 1,726 9,833 602 4,619 137 225
317
742
1,533
(1)
FFLCH - School of Philosophy, Literature and Human Sciences; FAU - School of Architecture and Urbanism;
IME - Institute of Mathematics and Statistics; ECA - School of Communications and Arts; FEA - School of
Economy, Administration and Accounting; FE - Education School; IAG - Institute of Astronomy, Geophysics and
Atmospheric Science; IB - Bioscience Institute; MAE - Museum of Archeology and Ethnology; IO - Oceanographic
Institute; IF - Institute of Physics; FCF - School of Pharmaceutical Sciences; EP - Polytechnic School; MAC Contemporary Art Museum; IGc - Institute of Geosciences; EEFE - Physical Education and Sport School; CCE Electronic Computer Center ; RUSP - Rectory of USP; IP - Institute of Psychology; FO - School of Dentistry; FMVZ
- School of Veterinary Medicine and Zootechny; IQ - Chemistry Institute; CCS - Social Communication
Coordination; IEE - Electrotechnical and Energy Institute- ICB - Institute for Biomedical Sciences; PCO - CUASO
Operations Services ; HU - University Hospital; COSEAS - Social Assistance Coordination; CEPEUSP - USP Sports
Center.
(2)
H - Human use; M - Mixed use; L - Laboratorial use; O - Other use.
(3)
WD - Water demand (m³/month); P - Population (people); PCWD - Per capita water demand (L/cap/day).
(4)
Source: USP (1999-2008).
(5)
Undergraduate and graduate students, professors and employees.
(6)
Considered 22 days of use of water per month.
up to 40 L/cap/day 41 ‐ 100 L/cap/day
101 ‐ 200 L/cap/day
above 201 L/cap/day
400
742
1533
350
Per capita water demand (L/cap/day)
Phase 1
317
Phase 2
300
250
225
197
200
134
150
100
50
7
14
15
16
17
30
31
32
36
42
54
54
56
68
62
68
70
85
89
139
141
146
99
0
Unit
Figure 2 - Per capita water demand of the CUASO’s Units in 2007
However, it’s verified that, as with the water use typology, the Units can be grouped in
characteristic ranges:
 Human use (H): up to 40L/cap/day - considering mainly the use in pantries,
restrooms and for cleaning;
 Mixed use (M): ranging between 41 and 100 L/cap/day - with the same use of the
human use plus use in laboratories;
 Laboratorial use (L): ranging between 101 and 200 L/cap/day - intensive use in
equipments and others processes;
 Other use (O): above 201 L/cap/day - intensive use in facilities services (PCO),
hospital (HU), residential buildings for students, day care and restaurants (COSEAS)
and sports club (CEPEUSP).
63 Still, not always the Units' demands follow the expected range to its use typology,
which can be due to characteristics or conditions of the systems and their maintenance
staffs or variations in their uses (congresses, renovation, etc). Thus, the correct
assessment of the per capita demands requires also much knowledge of the Units:
 Demand below the expected: IB - substitution of distillation by more efficient
technologies, FCF - efficient maintenance staff;
 Demand above the expected: IP - recurrent issues in ball valves and at external
networks (which demand complete substitution), IEE and CCE - expressive growth of
built area.

Besides, in some Units, the fixed population (composed by students, professors and
employees) shows low correlation with the water usage and, consequently, with the
demand. These cases denote the need of adoption of consuming agents other than the
fixed population:
 Demand above the expected: MAC, CCE, RUSP and CCS - visitors or external
public;
 Demand as expected: HU - external public and hospital beds, COSEAS and
CEPEUSP - external public.
In Table 3, as an example, it is presented the evolution of the per capita demands at
three places where the fixed population is not the most characteristic consuming agent.
Thus, are observed the per capita demands, in 2007, of: 1,710 L/bed/day, at University
Hospital (HU); 82 L/student/day at School of Application (of FE); and
191 L/habitant/day at the Residential Buildings for Students (of COSEAS).
Table 3 - Evolution of per capita demand in specific places
Unit (1) 1998 1999 2000 2001
2002
2003
2004
2005
2006 2007 Water demand per agent (L/agent/day) (2) (3) Hospital (4) 2,138 2,102 1,543 1,444
School (5) 112 112
99 81
(5)
201
200 168
Res. Building 228 1,758
81
248
1,592
99
198
1,306
118
248
1,362
95
229
1,429 1,710 88 82 203 191 (1)
University Hospital (HU); School of Application (of FE); Residential Buildings for Students (of COSEAS).
Source: USP (1999-2008).
(3)
Consuming agent: number of hospital beds, students and residents, respectively.
(4)
Considered 30 days of use of water per month.
(5)
Considered 22 days of use of water per month.
(2)
Other authors also identified the need of a system approach to the water demand,
concluding that it varies not only according to the fixed or external population, but it
also depends on other variables. Cheng; Hong (2004), in their study at primary schools,
state, for example, that there is an increase of the water demand in schools that have
swimming pools, kitchens or large gardens. Pedroso (2008), in her study also at schools,
obtained consumption indicators ranging from 5.56 to 49.85 L/student/day. The
64 authoress ratifies the need of evaluation of other variables: permeable and nonpermeable area; number of classrooms and restrooms; number of employees and
teachers; number and type of sanitary appliances (and leaks); daily maximum
temperature.
Besides various per capita demand values, it can be verified that its variation during
these ten years was quite distinct to each Unit. In Figure 3, it can be observed that, from
the 29 Units considered, 24 are showing reduction of the per capita demand, from 1998
to 2007, ranging from 1% to 76% (12 presented reductions above 50%).
100
Variation of per capita water demand (% )
80
65
60
40
50
47
46
35
20
0
-1
-20
-4
-16
-25
-40
-28
-34
-35
-35
-41
-43
-60
-80
-45
-48
-51
-56
-56
-58
-61
-64
-65
-67
-69
-72
-74
-76
-100
Unit
Figure 3 - Variation of per capita demand of the Units from 1998 to 2007
Additionally, it can be verified that the obtained percentages are distributed in a quite
uniform manner. As for the other 5 Units, they presented increase of the per capita
demand, ranging from 35% to 65%, some of which are related to the increase of the
constructed area and, consequently, a more intensive usage, which reinforces the
necessity of system approach of the demands. Recollecting the example of differentiated
consuming agents, reductions of 20% were obtained at the University Hospital (from
2,138 to 1,710 L/bed/day) and of 17% at the School of Application (from 112 to
82 L/student/day) from 1998 to 2007.
In Table 4 and Figure 4, it is presented the total daily per capita water demand for
Phases 1 and 2 of the PURA-USP and for CUASO (Total). The reduction of the water
demand, from 1998 to 2007, allied to the increase in population during this period,
resulted in a reduction of 46% to the per capita water demand at the CUASO, from 113
to 61 L/cap/day. Considering the Units from Phase 1, this reduction was even higher,
reaching 56%.
65 Table 4 - Evolution of CUASO’s per capita water demand
Unit 1998 1999 2000 2001 Per capita water demand (L/cap/day) Phase 1 116 85 70 63 Phase 2 108 109 113 94 Total 113 95 88 76 2002
2003
2004
2005
2006 2007 60
89
72
57
84
68
51
75
61
49
72
59
49
78
61
51 75 61 120
Per capita water demand (L/cap/day)
Phase 1
Phase 2
Total
100
80
60
40
20
0
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
Period
Figure 4 - Evolution of CUASO’s per capita water demand
Finally, it can be concluded, from the Table 4 and Figure 4, that a reduction occurred in
the per capita water demand at the CUASO, throughout the ten years evaluated, in
which it has established itself at a new level of 61 L/cap/day, and has maintained this
value in a consolidated manner since 2004.
We believe that for a new reduction to happen on the level of the per capita demand,
new structural intervention will be necessary in the systems, including the regeneration
and modernization of large cold water networks at the CUASO.
5. Final considerations
The expressive results of the PURA-USP obtained throughout the twelve years of its
existence are consequence of the permanent character of the Program, its system
approach and the association of technologic, mobilization and, mainly, water demand
management actions. In this context, the determination of the daily per capita water
demand for the CUASO's Units and their uses for the water demand management at the
campus were essential for PURA-USP. It is believed that the methodology developed at
PURA-USP can be applied to similar systems, since their particularities are respected.
Similarly, special care must be taken while extrapolating the data obtained in this paper
to other systems, since many factors differentiate the uses of the water.
66 6. References
Cheng, C. L.; Hong, Y. T. (2004). Evaluating Water Utilization in Primary Schools.
Building and Environment, 39, 837-845.
Pedroso, L. P. (2008). Estudo das Variáveis Determinantes no Consumo de Água em
Escolas: O Caso das Unidades Municipais de Campinas, São Paulo (227p.). Campinas,
Brazil: Unicamp.
Silva, G. S. (2004). Programas Permanentes de Uso Racional da Água em Campi
Universitários: O Programa de Uso Racional da Água da Universidade de São Paulo
(328p). São Paulo, Brazil: EPUSP.
Silva, G. S. et al. (2007). Eliminação de Vazamentos em Redes Externas no Contexto de
Programas de Uso Racional da Água - Estudo de Caso: Universidade de São Paulo.
Proceedings - X Simpósio Nacional de Sistemas Prediais (12p.). São Carlos, Brazil:
ANTAC.
Silva, G. S.; Tamaki, H. O.; Gonçalves, O. M. (2002). Water Conservation Programs
in University Campi - University of São Paulo Case Study. Proceedings - 28th
International Symposium Water Supply and Drainage for Buildings CIB W062 (13p.).
Iasi, Romania: Technical University “GH.ASACHI”.
Tamaki, H. O. (2003). A Medição Setorizada como Instrumento de Gestão da Demanda
de Água em Sistemas Prediais - Estudo de Caso: Programa de Uso Racional da Água
da Universidade de São Paulo (151p). São Paulo, Brazil: EPUSP.
Tamaki, H. O. et al. (2005). The Application of Remote Reading of Water Meters at
University Campi Regarding Water Conservation Programs - Case Study: University of
São Paulo. Proceedings - 31st International Symposium Water Supply and Drainage
for Buildings CIB W062 (14p.). Brussels, Belgium: BBRI.
Tamaki, H. O. et al. (2007). Minimização de Desperdícios de Água em Processos de
Purificação de Água - Estudo de Caso: Universidade de São Paulo. Proceedings - X
Simpósio Nacional de Sistemas Prediais (12p.). São Carlos, Brazil: ANTAC.
Tamaki, H. O.; Silva, G. S.; Gonçalves, O. M. (2004). Submetering as an Instrument
of Water Demand Management in Building Systems - University of São Paulo Case
Study. Proceedings - 30th International Symposium Water Supply and Drainage for
Buildings CIB W062 (12p.). Paris, France: CSTB.
USP - Universidade de São Paulo (2009).
A USP. Available
<http://www4.usp.br/index.php/a-usp>. Access in: July 22 2009.
____ (1999-2008). Anuários Estatísticos USP. São Paulo, Brazil: USP.
67 in:
7. Presentation of Authors
Orestes Marraccini Gonçalves concluded the undergraduate course
on Civil Engineering, obtained a Masters of Science and Ph.D. on
Civil Construction Engineering at the Polytechnic School of
University of São Paulo. Currently, he is a Full Professor of the
Polytechnic School, where he coordinates the group of Building
Systems Engineering. He is also the coordinator of the PURAUSP. Gisele Sanches da Silva has a degree on Civil Engineering and a
Masters of Science on Building Systems Engineering, both
awarded by Polytechnic School of University of São Paulo. She
currently works as engineer at the PURA-USP.
Humberto Oyamada Tamaki has a degree on Civil Engineering and
a Masters of Science on Building Systems Engineering, both
awarded by Polytechnic School of University of São Paulo. He
currently works as engineer at the PURA-USP.
Gabriele Malta Correia is an undergraduate student in
Environmental Engineering at the Polytechnic School of University
of São Paulo. She works as a trainee at the PURA-USP.
68 Session II: Water Supply II
II.1
Proactive crisis management of urban
Infrastructure
M. P. Nekrep
matjaz.nekrep@uni-mb.si
University of Maribor, Faculty of Civil Engineering, Slovenia
Abstract
As a member of European Action COST C19 I would like to present main objectives of
the action "Proactive crisis management of urban infrastructure" in my present work.
The main objective of the Action is multi-disciplinary research on urban infrastructure
vulnerability and handling of crisis situations. Obviously there are connections between
hazards on the one side, and design, maintenance and operation of infrastructure
systems (and components) on the other side.
The complete life-cycle crisis management (e.g. planning, design, construction,
operation, maintenance and rehabilitation of urban infrastructure,) will be discussed, but
here limited to water and electricity supply and road transportation.
We can divide life-cycle into three rough topics:



Proactive risk management
Planning on how to handle acute crises
Management of acute crises
Work is based upon results of the COST C-19 action.
Keywords
Risk management/ urban infrastructure / COST C-19
69 1. Introduction
Introduction to COST Action C19
This work is based upon results of COST C19 “Proactive crisis management of urban
infrastructure” EU action. The action was initiated and supported by COST (European
COoperation in the Field of Scientific and Technical Research). The action was
approved and officially started the 19th of March 2004 and last until the final
conference/meeting in February 2008. COST Actions brings together researchers and
experts in different countries working on specific topics. It finances networking of
nationally funded activities in supporting meetings, conferences, short term scientific
meetings etc. COST does not fund research itself.
The main objective of the Action is to define current knowledge gaps and identify
possible measures to improve the multi-disciplinary research on urban infrastructure
vulnerability and handling of crisis situations. Additional objective of the Action is to
present a state-of-the-art on current know-how demonstrating its application in direct
relation to crisis situations.
Modern developed societies are heavily dependent on urban infrastructures like
transport, railway, gas, electricity, and water supply. The breakdown of one of such
critical infrastructures may cause serious consequences for the safety of the citizens.
Managing the safety and security of these infrastructures taking into account prevention
and preparedness, crisis interventions and restoring normal conditions using optimised
and risk based approaches are important. Urban infrastructure is highly vulnerable to
earthquakes, political conflicts, terrorism, droughts, floods, and other natural and
societal disasters. Failures of some of these structures, such as water supply and other
pipeline systems, roads and bridges, cable communications and energy supply may have
major impacts in terms of human lives and economic losses.
The members of the actions represent 15 countries in Europe (Czech Republic, Finland,
Germany, Italy, Netherlands, Norway, Portugal, United Kingdom, France, Switzerland,
Serbia Montenegro, Slovenia, Cyprus, Iceland and Romania.) and with expertise
covering different infrastructures (road and railway transportation, water and
wastewater, electricity supply, gas). The members of COST C19 consist of MC
members and WG members. AT some of the meetings also invited experts has joined
the meeting..
The work in the action was organized in two working groups (WG):


WG A: Risk analysis and risk management with focus on theories, methods and
tools for risk assessment.
WG B: Planning of handling of acute crisis
70 The work in C19 covers most of the phases in the risk-/crisis management wheel as
illustrated in
Figure 0, from prevention, preparedness, warning and alarm, response, limiting the
extent of loss and recovery.
The strength of the action is the multi-disciplinary approach evaluating many types of
urban infrastructure. Different disciplines can learn from each other both even though
each sector is unique and has its own challenges.
Figure 0: Risk management cycle
Results from the action
The results from COST action C19 consist of a wide variety of outputs.
One important outcome of the action is the networking among different
disciplines/infrastructures. Even though each sector/discipline is unique and has its own
challenges, learning from each other has been valuable.
71 At present, a number of methods for analysing risks of technological systems that are
suitable for hazard identification and risk assessment along with the criteria of their
selection are known for the different infrastructures. Within C19 a state of the art report
on risk analysis (RA) methodology for different infrastructures has been developed and
discussed in plenary.
There are few attempts on developing a RA methodology with a interdisciplinary
perspective/crosswise different infrastructures. As a part of the action a prototype tool
for identification and estimation of risk related to critical infrastructure has been tested
within the action. The method primarily focuses on the identification and estimation of
the risk. The tool can be used for analysing risk and vulnerabilities between different
sectors/infrastructures and the interrelationship between the urban infrastructures. The
risk is assessed and plotted directly into standard risk matrixes. The consequences might
be expressed with respect to different aspects: life and health, environment, economy,
manageability, political trust, supply failure (both quality and availability of delivery).
The tool can be used for overall analysis at national, region or local level.
Figure 1: Prototype tool for risk analysis of urban infrastructures
The work in COST action C19 has identified knowledge gaps and future research needs.
The identified gaps results from questionnaires, case studies, papers and discussions
with all C19 members.
The final report is divided into two main parts. The first session deals with generic
methods for risk analysis with illustrative cases where some of the methods have been
72 applied on different infrastructures. The second part of the report deals with acute crisis
and the handling thereof.
2. Critical infrastructure - Water supply
Introduction
Water supply covers all elements from water source to consumers tap. An illustration of
a typically water supply system is shown in Figure 3. The water source might be
groundwater, surface water (lakes and rivers) or desalination. After abtraction the raw
water is normally purified and disinfected in a water treatment plan. Treated water are
tansported to the consumers either by gravity or the water is pumped. The water
distribution systems consist of many elements like pipes, tanks, valves, pumps etc. Once
water is used, wastewater is typically discharged in a sewer system and treated in a
wastewater treatment plant before being discharged into a river, lake or the sea. The
abstraction source for drinking water supply varies from contry to country, e.g. close to
100% groundwater abstraction in Denmark to about 92 % surface water abstraction in
Norway1.
Coagulant
Tanks/reservoir on the network
pH
Consumer
Surface water
UV
Excess water
Figure 3 Water supply illustration (illustration adapted from www.bergenvann.no)
The service quality of the water supply has many different dimensions: continuity/water
quantity; water quality; water pressure; and serviceability (e.g. responding to customers
complaints).
The way of organizing the water supply sector varies from country to country due to
economic, legal, political, structural, historical conditions. Preferred solutions in one
country can not be automatically be transferred and adopted in another country. Public
1
International Statistics for water supply, IWA Specialist group on Statistics and Economics, Beijing 2006.
73 ownership of the water supply infrastructure is the common solution, with England and
Wales as one clear exception, where both ownership and operation are on private hands.
In all countries public tenders for limited operations contracts are getting more and
more common. There is a tendency of restructuring of the sector, leading to fewer units
(e. g. cooperation between municipalities or full mergers). Most countries apply a
system of full cost coverage for financing the services, but also systems with price cap
regulation exists like in England and Wales. Even though the sector mostly is organized
under public control, most countries and political leaders want to modernize the sector
and make it more efficient. In many countries testing of different ways of organizing the
sector is under way, assuming that the new solutions are better for solving the existing
problems (outsourcing of operation, maintenance and construction, public – private
partnership etc).
Typically a great variety of institutions have responsibilities in water supply. A basic
distinction is between institutions responsible for policy and regulation on the one hand;
and organizations in charge of providing services on the other hand. The service
providers might also be splitted into owner of the assets and the operator.
Water supply systems are major assets and essential infrastructures in every modern
society. Therefore they are considered as a critical infrastructure. In fact, water supply is
a vital lifeline with topmost priority in any crisis situation and the prompt availability of
potable and fire-fighting water is a fundamental prerequisite to mitigate the impacts of
the crisis. The disruption of such systems, besides the effects through the direct
consumers, can perturb the performance of other infrastructures, as vital as hospitals or
other emergency systems, and can induce economic losses in systems like industrial
facilities or generation and distribution of electric power.
During the last decade, there are several examples of water supply and sewer system
damage due to river flooding. Recent catastrophes like in Moldau and Elben in August
2002 and later several rivers in Central Europe and United Kingdom in December 2002
have been widely documented in the daily press. Another example is the enormous
floods in Moravia during summer 1997. A number of municipalities were without
public drinking water for several days. “Crisis staff” under command of special army
officers coordinated all activities. The financial damage was extremely high and a lot of
water resources, water treatment plants, wastewater treatment plants, etc. had to be
reconstructed after that incident.
74 In the UK several incidents of cast iron water pipe bursts have occurred with massive
losses of water because the bursts could not be isolated. Radio and television were used
to warn and advise city populations. In some cases it had been necessary to mobilize the
army as there were insufficient operational resources in the water company to manage
the problem. In one company, water supply problems due to drought and associated
inadequate management forced the key directors to resign. Water utilities all over
Europe had to take special precautions after the attack on the WTC in New York on
11th September 2001.
Similar situations derive from severe or prolonged droughts (e.g. Spain, Israel) or floods
(Poland, Germany) that force the utilities to implement short, medium and long term
solutions to face the problem in a sustainable way and bring the attention of the research
community to the problem. Some serious events on water supply pollution (i.e.
Walkerton, Milwaukee, Hertfortshire) have brought the focus to scenarios of water
pollution due to accidents, human failures or even terrorism.
Challenges
Decision-makers should have in mind that there are population’s expectation pointing
toward a good performance, even under many types of adverse conditions, for the water
infrastructures. This means that infrastructures are supposed to be robust and resilient
under a wide range of possible future scenarios. Therefore decision-makers have to
decide how these structures should be built, hardened or operated to cope with risky
situations. Besides particular measures that have to be taken to avoid criminal
opportunities, some measures should be taken into account at the planning, design or
management level of such systems. Water systems are very costly and represents huge
replacement costs.
Recruiting new personnel with high competence has become an emerging challenge in
many water companies.
75 Table 0. CRITICAL INFRASTRUCTURES - SOME DEFINITIONS
Definition Source Critical infrastructures consist of those physical and information technology EU facilities, networks, services and assets which, if disrupted or destroyed, would have a serious impact on the health, safety, security Systems and assets, whether physical or virtual, so vital to the United States that US ‐ the incapacity or destruction of such systems and assets would have a debilitating Patriot impact on security, national economic security, national public health or safety, or act any combination of those matters. «an infrastructure that is so vital that its incapacitation or destruction would have US‐ a debilitating impact on defence or economic and the national critical PCCIP infrastructures» National Critical Infrastructure refers to those physical and information Canada technology facilities, networks, services and assets, that if disrupted or destroyed, would have a serious impact on the health, safety, security or economic well‐
being of Canadians or the effective functioning of governments in Canada. Critical infrastructure is those facilities, services and information systems which NATO are so vital to nations that their incapacity or destruction would have a debilitating impact on national security, national economy, public health and safety and the effective functioning of the government. Organisations or institutions which are of essential importance to society, the Switzer‐
failure or disruption of which will cause long‐term bottlenecks in the supply chain land or have other dramatic consequences for large sections of the population Critical infrastructures (CI) include organizations and systems with tremendous Germany importance for the society that, if disrupted, would have sustained influence on supply chains and public safety and could lead to further dramatic consequences. Infrastructures are deemed critical if they constitute an essential, indispensable Neder‐
facility for society, and if their disruption would rapidly bring about a state of lands emergency or could have adverse societal effects in the longer term. Those assets, services and systems that support the economic, political and social UK life of the UK whose importance is such that any entire or partial loss or compromise could: cause large scale loss of life, have a serious impact on the national economy, have other grave social consequences for the community or be of immediate concern to the national government Critical infrastructure is those assets and systems which are essential to uphold Norway the critical functions of the society which again meets the basic needs of society and the feeling of security in the population. Critical infrastructure includes the following infrastructures: electric power, electronic communication, water and 76 wastewater, transport, oil/gas and satellite‐based infrastructure. Critical society functions includes banking, food, health and social security services, police, rescue and emergency service, crisis management, government, court of justice, defence, environmental monitoring and waste management. 4. Introduction to risk analysis
Basic concepts
Generally, risk management is defined (IEC 300-3) as a “systematic application of
management policies, procedures and practices to the tasks of analyzing, evaluating and
controlling risk”. It will comprise (IEC definitions in parentheses):


Risk assessment, i.e.
o Risk analysis (“Systematic use of available information to identify
hazards and to estimate the risk to individuals or populations, property or
the environment”), and
o Risk evaluation (“Process in which judgments are made on the
tolerability of the risk on the basis of risk analysis and taking into
account factors such as socio-economic and environmental aspects”)
Risk reduction/control (Decision making, implementation and risk monitoring).
There exists no common definition of risk, but for instance IEC 300-3 defines risk as a
“combination of the frequency, or probability, of occurrence and the consequence of a
specified hazardous event”. Most definitions comprise the elements of probabilities and
consequences. However, some as Klinke and Renn (2002) suggest a very wide
definition, stating: “Risk refers to the possibility that human actions or events lead to
consequences that affect aspects of what humans value”. So the total risk comprises the
possibility of a number (“all”) unwanted/hazardous events. It is part of the risk analysis
to delimit which hazards to include. Further, risk refers to threats in the future,
involving a high degree of uncertainty.
We use Kaplans (1991) definition of risk as the conceptual risk definition. This risk
definition is valuable both when communicating the risk picture, and when assessing the
risk picture. Kaplan takes all possible scenarios, s, as a starting point. A scenario in this
context is a sequence of events that could occur, and lead to an accident or undesired
event. Then the frequency (f) of this event expresses how often (or probable) such a
scenario would take place. Finally the consequence, or result of the scenario could be a
multidimensional vector of end consequences, say x. The risk picture is described by the
set of all triplets <si, fi, xi>.
77 It is important to emphasise that this definition does not say how to measure (quantify)
risk. In order to quantify the risk we usually need to i) identify hazards, ii) perform a
frequencies and root cause and frequency analysis, iii) perform a consequence analyse,
and iv) compile the total risk picture.
In Figure 4 we have illustrated the relationships between risk analysis, risk evaluation,
risk assessment, risk reduction/control and risk management according to the definitions
used in IEC 60300-3-9.
Figure 4 Risk assessment & management (IEC 60300-3-9)
78 Figure 5Figure 5 shows basic steps in the risk analysis and assessment procedure.
Planning and
organising
What is
acceptable risk?
Description of
object
Hazard
identification
Frequency
analysis
Consequence
analysis
Risk
evaluation
Acceptable?
Risk reducing
measures
No
Yes
Other measures
desirable?
Figure 5 Risk Analysis and Assessment Procedure
Several methods exist for each of the main “boxes” in Figure 5. Hazard identification or
identification of the undesired event is supported by e.g., use of Checklists, Preliminary
hazard analysis, FMECA, HAZOP and use of Event data sources. The frequency or
causal analysis is supported by e.g. Fault tree analysis (FTA), Reliability block
diagrams (RBD) Influence diagrams, FMECA and use of Reliability data sources. The
consequence analysis is supported by e.g., Event tree analysis (ETA), Consequence
models, Evacuation models, Fire & explosion models, Monte Carlo Simulation models,
Hydraulic models and Traffic flow models.
There exist several standards related to risk analysis, some of these are listed below:

IEC60300-3-9: “Risk analysis of technological systems”

EN1050: “Safety of machinery – Risk assessment”

EN50126: “Railway applications – The specification and demonstration of
reliability , availability, maintainability and safety (RAMS)”

ISO17776: “Petroleum and natural gas industries – Offshore production
installations – Guidelines and tools for hazard identification and risk assessment
79 
EN1441: “Medical Devices - Risk Analysis”

ISO (Draft) Guide: 73: Risk Management – Vocabulary - Guidelines for the use
in Standards
Quantitative risk measure
In many presentations and standards risk is defined as the probability of an undesired
event multiplied with the consequence of that event. We do not find this definition
sufficient for our presentation. When giving a definition of risk, it is important that the
definition is broad enough to capture the elements we want to put into it. This will
always be the starting point, then we may at a later stage be more pragmatic, and narrow
the scope of the definition. In the following we will argue in the lines of Kaplan (1991).
To express the risk we basically ask three questions: i) what can go wrong? ii) how
likely is it?, and if it goes wrong, iii) what are the consequences?
Safety versus security
In the literature there are some differences between the words ‘safety’ and ‘security’.
Often the term security is used for protection against deliberate incidents and the term
safety is used for protection against unintended incidents.
More precisely we define:

Safety is protection against random incidents. Random incidents are unwanted
incidents that happen as a result of one or more coincidences.

Security is protection against intended incidents. Unwanted incidents happen
due to a result of deliberate and planned acts.
Security incidents are results of planned actions, i.e. acting in order to achieve a wanted
outcome (e.g. money), while safety accidents are unplanned (no one plans to cut his
finger in a sawing machine). The security incidents are mainly malicious and criminal
acts. Safety accidents are seldom, if ever, malicious, but they can be criminal as they
often are violations against regulations.
The hazard identification in Figure 5Figure 5 is usually used in a context where “safety”
is the scope of the analysis. When we deal with security problem, the word ‘hazard’
might be replaced with the term ‘threat’. Often we also use the word ‘threat agent’. A
threat agent could be a person, a group of persons, a method etc. used to exploit a
vulnerability in an information system, operation, or facility, fire, natural disaster and so
forth. Threats within the field of security can always be tracked down to humans.
80 Incidents are a result of a person or a groups will. The threats within security can be
divided into external threats and insider threats (i.e. inside the organisation). The
external threats (e.g. hackers) principally imply deliberate incidents. This external threat
makes the picture of threats more complex and adds uncertainty. It is impossible to
control an external threat, the only measure the organisation may implement to protect
oneself. Further it is difficult to predict the threat (where, when and how the attack
appear), and it might be difficult to find the responsible after an incident since the
attacker seldom leave any tracks behind (information security).
Acceptable risk and value trade off
A baseline for the discussion in this paper is that risk is never acceptable
unconditionally, risk is undertaken only if a benefit is desired. See for example Fischoff
et al (1981) and Vatn (1998). In other words, this means that risk it self is not
acceptable, but the decision or action leading to risk may be acceptable. Different
principles exist for treating the acceptable risk problem, e.g. the MEM, the GAMAB
and the ALARP principles (EN 50126). For example the ALARP principle split risk
into three regions. In the unacceptable region risk cannot be justified exept in
extraordinary circumstances. In the ALARP region (As Low As Reasonably
Practicable) risk is tolerable if risk reduction is impracticable or its cost is grossly
disproportional to the improvements gained. In the lowest region there is no need to
demonstrate ALARP. In case of applying the ALARP principle we need to define the
border line between unacceptable risk and the ALARP region. There are several such
borderlines which are denoted risk acceptance criteria (RAC). Since risk is only
acceptable if some gain could be achieved by the operation or activity leading to risk
RACs are hard to establish unconditionally. Often risk acceptance criteria are
established based on historical risk. The main argument is then that a new activity
should not increase the risk level, and we should stick to the historical risk level when
the RACs are established. However, since the new activity usually is not equal to the
historical activity with respect to benefits and other inconveniences, the historical risk
might not be very relevant. The discussion about whether explicit RACs should be used
in the risk management process is not pursued further in this report.
81 5. Conclusions, knowledge gaps and future research needs
The work in COST action C19 has identified knowledge gaps and future research needs:
-
Application of risk based methods might contribute for risk managers to achieve
better solutions. Some scepticism still surrounding the application of decision
models can be understood by the challenge of formulating and solving realistic
models to express all the features of the decision problems. Even if there are
research needs for improving the methods available, nowadays there are very
powerful tools that can incorporate all the facets occurring in real-world
situations. One reason for the so far poor application of risk based methods is the
missing orientation of relevant research towards decision makers as
“consumers” of risk management research.
-
The interactions of urban infrastructure systems are physically very complex.
There is need for better methods to address the complex interactions of urban
infrastructure systems, physical environment, level of services and social factors.
A prerequisite for an integrated approach is the ability to evaluate and measure
the interactions between physical infrastructure, delivered services and social
consequences. The complex consequences on urban services, human health,
economy resulting from malfunctioning of urban infrastructure and the
interaction of all functions is difficult to address.
-
Of increasing importance is the exchange between different infrastructuresectors and different utilities to i) communicate the interdependence of
infrastructure systems and ii) to transfer knowledge from those sectors/utilities
which already successfully apply risk based methods (e.g. oil industry).
-
Circulation of information and mutual understanding (e.g. between experts and
concerned citizens) are crucial for efficient risk management.
82 4 References
1. Røstum et al, Proactive Crisis Management of Urban Infrastructure, COST office,
Brussels 2008
2. IEC 300-3-9, “Dependability Management – Part 3: Application guide – Section 9:
Risk analysis of technological systems”
3. IEC 812, “Analysis techniques for system reliability – Procedure for failure mode
and effects analysis (FMEA)”
4. Tuhovčák,, L., Kučera, T., Ručka, J., Svobod, M., Sviták, Z. (2006): Technical audit
of the water distribution network, Water Science & Technology: Water Supply Vol
6 No 5, ISSN: 1606-9749, p.129-138
5. www.WaterRisk.cz; official website of the Czech national research project
“Identification, quantification and management of risks of public water-supply
systems – WaterRisk”
6. World Health Organization, “Guidelines for drinking-water quality 3rd edition”,
WHO Geneva, (2004), ISBN 92 4 154638 7
7. AWWARF (2002). Cost of Infrastructure Failure. American Water Works
Association Research Foundation. ISBN 1583212647.
8. Balducelli, C., Bologna, S., Di Pietro, A., Vicoli, G. (2005). Analysing
interdependencies of critical infrastructures using agent discrete event simulation.
International Journal of Emergency Management. 2(4). 306 – 318.
9. Chang, S. (2003). Evaluating disaster mitigations: methodology for urban
infrastructure systems. Natural Hazard Review, 4(4), 186-196.
10. Mil van B.P.A., A.E. Dijkzeul, R.M.A. van der Pennen (2006). Risk modelling
handbook: selection of models and methods fot conducting risk analysis. Ministery
of Economic Affairs, The Netherlands, available at: http://www.minez.nl/ .
11. Whelton, A., Wisniewski, P., States, S., Birkmire, S., Brown, M. (2006). Lessons
learned from drinking water disaster and terrorism exercises. Journal American
Water Works Association, 98(8), 63-73.
12. Kaplan. S. 1991. Risk Assessment and Risk Management - Basic Concepts and
Terminology. Hemisphere Publ. Corp., Boston, Massachusetts, USA. In Risk
Management: Expanding Horizons in Nuclear Power and Other industries, pp.11-28.
13. Klinke, A. & Renn, O. 2002. A new approach to risk evaluation and management:
Risk-based, precaution-based and discourse-based strategies. Risk Analysis, 22(6),
1071-1094
14. Fischhoff, B., S. Lichtenstein, P. Slovic, S.L. Derby, and R.L. Keeney. 1981.
Acceptable Risk. Cambridge University Press, New York.
15. IEC 300-3-9 Dependability management - Part 3: Application guide - Section 9 Risk
analysis of technological systems. International Electrotechnical Commission. 1995.
16. IEC 812, standard Analysis Techniques for System Reliability - Procedure for
Failure Mode Effects Analysis
17. Vatn, J. 1998. A discussion of the acceptable risk problem. Reliability Engineering
and System Safety, 61(1-2):11-19.
83 5 Presentation of Author
A
Matjaž Nekkrep Perc, B.SC.
B
Civill. Eng. is a Senior Leccturer at thee
Faculty forr Civil Enggineering, Head
H
of Cenntre for Hy
ydrotechnicss
and authoor of manny workingg solutionss of CAM
M in civill
engineeringg.
8
84 II.2
Water systems in high rise residential
buildings, guide lines for design
and construction
W.G. van der Schee (1)
(1) w.g.vd.schee@wolterendros.nl
Abstract
This paper is based on the study and gives a brief summery of the report “Water systems
in high rise residential buildings, guide lines for design and construction”. This paper
focuses on two of the three aspects: pressure zones with boosting units and hot water
production with distribution.
The design of a water system for high rise buildings differs on two important objects
from standard installations: the pressure and the hot water distribution. Over the world
several solutions are available and described. These solutions depend on the local
practices, costs for energy, history and culture.
Keywords
High rise buildings
1.
Introduction
In The Netherlands more and more high rise residential buildings are under
construction. A building is called a high rise residential building if the height is more
than 70 meters and the main function of the building is residential. Up till now
85 knowledge and experience to design and construct water systems in high rise residential
buildings is mainly practical, guide lines or standardization are not readily available.
The requirement this knowledge is growing, due to the growth of high rise buildings.
The associations TVVL and Uneto-VNI have initiated a study on the specific problems
and requirements on potable water systems in high rise residential buildings. The result
is a report “Water systems in high rise residential buildings, guide line for design and
construction”.
This paper defines the basic design data and conditions water systems in general.
Chapter 3 focuses on the design of the hydraulic system; different pressure zones,
pressure reduction valves and types of pressure boosting units. Chapter 7 focuses on the
hot water production and distribution. Solutions for both local hot water production and
central hot water production are also described.
2.
Basic design data and conditions
Water systems are constructed in the Netherlands according NEN 1006 “General
regulation for water systems”. NEN 1006 describes the requirements for water systems.
For a practical translation to design and construct water systems in the Netherlands
there are the “Water guidelines”. The main regulations and guidelines for water systems
are stated below.
2.1
Pressure
There is no strict regulation on the pressure in water systems in the Netherlands. But,
one of the most important aspects to design a well operating water system is the
pressure. Different pressures levels are necessary for the various appliances in of the
system. The following pressures are mentioned in the various guidelines:




The pressure delivered by the water company differs per water company. The
minimum pressure after the flow meter is 200 kPa. (Water company)
The minimum design pressure on the taps depends on the application of the tap.
Guidelines for the minimum pressure of taps for domestic use are stated in table 1
(according the “Water guidelines”).
The maximum static pressure on the tap is 500 kPa (according NEN-EN 806-2)
The maximum pressure on the tap is 600 kPa (according the “Water guidelines”).
For high rise buildings it’s not desirable to maintain this pressure.
86 Table 1 Minimum pressure on domestic taps
tap Minimum pressure (kPa) Shower, bath, washbasin, 100 kitchen Fire hose reel 150 Luxury shower Specification manufacturer, if unknown 200 Luxury sanitation Specification manufacturer 2.2
Temperature and availability hot water
Regulation and guidelines with regard to the temperature have as object safety (avoiding
scalding and the growth of Legionella) and functionality. The Regulation and
guidelines with regard to the temperature are as follows:






2.3
The water company supplies water with a minimum temperature of 4°C and a
maximum of 25°C.
In a house installation without circulation system the temperature at the tap is at
least 55°C.
In a house installation or a collective water system with circulation system the
temperature on the tap is at least 60°C.
In hot water systems with circulation system the temperature at the return is at least
60°C.
Drinking water is not warmer than 25°C.
For hot water at the tap the maximum waiting time is 20 + 15 seconds. 20 seconds
for the pipe and 15 seconds for the hot water producer
Noise
Noise is generally described as “unwanted sound”, so the designer of the water system
needs to address unwanted noise. Requirements on noise are set up to avoid or reduce
noise nuisance. The maximum sound level in houses caused by the sanitary installations
is 30 dB(A) according to NEN 5077. This requirements with respect to piping systems
are:


In residential buildings the maximum water velocity in pipes for cold and hot water
generally is 1.5 m/s. With an absolute top limit of 2 m/s.
For the hot water circulation system the maximum water velocity is 0.7 m/s.
87 3.
Hydraulic lay out
Beside the above described guidelines to design a correct hydraulic water system the
following issues are also important:


The height of the building. The fixed height of the taps.
The friction loss in the distribution pipes
3.1
Friction loss in the distribution pipes
In distribution pipes the friction loss is mainly relevant in very tall buildings. Because of
the total length of the distribution pipe, the sum of the friction loss achieves high values.
It is possible that on the highest floors a relative big difference occurs between the static
pressure and the user pressure. Since the hydraulic design for boosting units is often
based on the minimum user pressures, the maximum static pressure is actually less.
In distribution pipes the friction losses per meter are limited. As a result of the relative
high calculated maximum flow, pipes with big diameters are selected which results in
lower friction losses. Table 1 shows the friction losses for the basic design. Table 1 Design friction loss
Water velocity in riser (m/s) Design friction loss (kPa/m) 1 0.25 1,5 0.6 2 1.25 3.2
Maximum pressure in the installation
The height of the building, the minimum user pressure for the index tap and the friction
losses in the main determines the maximum static pressure in the pipe system. This
maximum static pressure is necessary to select the quality of the pipe material.
The maximum pressure in the pipe system is determined as follows:
1. Determine the maximum height hmax (m) from the booster pumps to the highest
leading tap.
2. Determine the minimum user pressure pdyn,tap (kPa) for the index tap.
88 3. Determine the friction losses in the index pipe work ∆pdyn (kPa/m)
4. Calculate the maximum pressure pmax (kPa) with:
(1)
p max  p dyn ;tap  h max  (10   p dyn )
Example
A high rise residential building, index tap on 150 metres above the boosting pumps.
Minimum use pressure index tap on 150 metres 200 kPa, friction losses in the mains at
2 m/s velocity are estimated on 1.25 kPa/m. Maximum pressure:
pmax = 200 + 150 * (10 + 1.25) = 1,889 kPa
3.3
Maximum height of a riser
In paragraph 3.2 the height of the building is a governing factor to determine the
maximum pressure in the installation. In this paragraph the maximum pressure of the
pipe, limited by the pipe pressure rating, is the basis to determine the theoretical
maximum height of the riser.
The maximum height of the riser determined as follows:
1. Determine the maximum pressure in the pipe material. For pipes there are three
ratings: 1,000 kPa, 1,600 kPa and 2,500 kPa.
2. Determine the minimum use pressure pdyn,tap (kPa) for the index tap.
3. Determine the friction losses in the riser ∆pdyn (kPa/m)
4. Calculate the maximum height hmax of the mains with:
hmax 
p max  p dyn ;tap
10  p dyn
(2) Table 2 shows for the most common situations for the maximum height hmax of the riser.
89 Table 2 Maximum height hmax of the riser in relation to the pressure rating
Max. rating Min. use pressure 1,000 kPa Max. rating Max. rating 1,600 kPa 2,500 kPa (kPa) Friction loss (kPa/m) Friction loss (kPa/m) Friction loss (kPa/m) 0.25 0.6 1.25 0.25 0.6 1.25 0.25 0.6 1.25 100 88 85 80 146 142 133 234 226 213 150 83 80 76 141 137 129 229 222 209 200 78 75 71 137 132 124 224 217 204 As explained in paragraph 3.1 in high rise buildings also the maximum static pressure
can lead. This occurs mainly if there is a high friction loss caused by elbows, changes in
diameter and devices in the pipes. In case of standard circumstances this is not a
problem: The static pressure will not exceed the maximum static pressure on the leading
tap. If the average friction loss at the total height of the riser is higher than indicated in
table 3 the difference between the user pressure and the static pressure will be too high.
In that case pressure reducing valves are necessary or it is necessary to reduce the
friction losses. In this case keep also in mind the losses in the hot water producer.
Table 3: Maximum average friction loss per meter riser.
Height riser Minimum pressure index tap (kPa) (m) 100 150 200 75 4.0 3.3 2.7 100 3.0 2.5 2.0 125 2.4 2.0 1.6 150 2.0 1.7 1.3 175 1.7 1.4 1.1 200 1.5 1.3 1.0 225 1.3 1.1 0.9 Remark: Maximum pressure on index tap 400 kPa 90 Figure 1 indicates the consequences if the friction loss in the riser is too high. With a
height of the riser of 150 metres, a average friction loss of 2.5 kPa/m and, on the index
tap a minimum user pressure of 150 kPa, the static pressure on this index tap will be 525
kPa. As shown in table 3 the average friction loss is limited at 1.7 kPa.
160 140 120 100 80
Height 60
40
20
0
0
200
100 kPa 400 600 800
525 kPa 1000 1200 1400 1600 1800 2000 2200
Druk [kPa]
Static pressure
user pressure
Figure 2 With an average friction loss of 2.5 kPa/m in high risers the difference
between static pressure and user pressure becomes to high
Remark on figure 1: On the total height of the riser the maximum flow differs from the
design maximum flow, therefore the relation between the user pressure and the height
is not proportional
4.
Pressure boosting principles
The following paragraphs describes the various pressure boosting principles and how to
determine the pressure zones.
1.
2.
3.
4.
4.1
Pressure zones without pressure reducing valves
Pressure zones with single pressure reducing valves
Pressure zones with serial pressure reducing valves
Pressure zones with break units with atmospheric pressure
Pressure zones without pressure reducing valves
The principle of pressure zones without pressure reducing valves is shown in figure 2. A
boosting unit is necessary for each different pressure zone. The pressure realised by the
91 water com
mpany is gennerally suffficient to feeed the low
wer floors. The
T minimuum and the
maximum user pressuure for the inndex taps deetermines th
he height off a pressure zone.
Figure 2 Pressure zones with
hout pressure reducingg valves
4.2
v
Presssure zones with singlee pressure reducing valves
The appliccation of preessure reduccing valves (PRVs) offfers flexibiliity for the ddesign. It is
possible too realize higgher pressurre zones andd the numbeer of boostiing units consequently
decreases, its sometim
mes possibble to use only
o
one boosting uniit. In addittion of the
booster puumps the eqquipment, piping,
p
valvves, fittingss must be designed
d
annd rated to
accommoddate the highh water presssures particcular at the base of the water pipinng system.
Safety meaasurements are necessaary to avoidd that the pressure
p
beccomes too hhigh in the
installationn in the residdential accoommodationns. These arre as follow
ws:



A maxximum presssure of 6000 kPa at thhe inlet of the pressuure reducingg valve by
creatinng more prressure zonees. See figgure 3. This solution leads to hiigh energy
losses.
More pressure
p
redducing valvees in serie.
A safety valve in the apartmeents. This saafety valve requires a drainage.
d
Seee figure 4.
The diagraam printed in figure 4 b results in hundreds of PRV
Vs for everyy building.
Access muust be provvided to seervice thesee valves. The
T costs of
o the valvees and the
additional costs for prroviding acccess is a decciding disad
dvantage. Thhis method is not very
popular. The
T diagram
m printed inn figure 4 b eliminates some of the PRVs bby making
branches and
a serve sevveral floorss with one PRV.
P
9
92 a b
Figure 3 Installation with sin
ngle pressu
ure reducin
ng valves
Figure 4 Installatioon with singgle pressurre reducingg valves and
d safety vallves in the
aparttments
4.3
Presssure zoness with seriaal pressure reducing valves
v
Downstreaam arrangeement of PR
RVs offers also more flexibility for the design of thee
distributioon system. When moree PRVs aree used less boosting ssystems are necessary..
The initiall investmennt costs are lower but the
t energy costs
c
increaases. Figuree 5 shows a
up-feed syystem with serial
s
conneected PRVs.
93 9
Figu
ure 5 Distrib
bution systtem using a master PR
RV and dow
wnstream P
PRV
4.4
Presssure zones with break
k units and
d top tank
It’s also possible
p
to create pressure zones with atmospheric tannks. Severall tanks are
installed onn different heights in the
t buildingg. See figurre 6. The lowest pump serves the
fixtures in the lowestt pressure zone and thee upper tan
nk. The capacity of thee basement
t total buuilding dem
mand. This normally used
u
is reguular if the
pumps is sized for the
w be too high for the pipe rating.. The appliccation of ann atmospherric tank has
pressure will
risks, the water
w
is in contact
c
withh the ambiennt air (by a filter),
f
so baacterial conntamination
is a threat. If the watter temperaature in thee tank increeases there is a risk foor growing
legionella. In the Neetherlands it’s not coommon to use gravityy tank sysstems with
atmospheriic storage taanks.
F
Figure
6 Seeveral presssure zones with tankss.
9
94 5. Height of the pressure zones
A pressure zone is defined as a pipe system where at the connected taps the required
minimum user pressure and maximum user pressure or maximum static pressure are
maintained. There are two principles: Without PRVs and with PRVs.
5.1
Without pressure reducing valves
Without the use of PRVs the minimum and maximum user pressure determine the
height of the pressure zone. The height of the pressure zone is determined by the lowest
value of the following two formulas:
Based on the minimum and maximum user pressure:
h
p dyn ;tap ;max  p dyn ;tap ;min
10  p dyn
(3)
And based on the minimum user pressure and maximum static pressure:
h
p stat ;tap ;max  p dyn ;tap ;min  h1  p dyn
10  p dyn
h
Maximum height of the pressure zone (m)
h1
Height of the start of the pressure zone (m)
pdyn;tap;max
Maximum user pressure
pdyn;tap;min
Minimum user pressure
∆pdyn
Average friction loss per meter
95 (4)
For the most situations formula 3 determines the maximum height. Table 4 shows the
maximum height of a pressure zone for the most common situations.
Table 4: Maximum height in metres of a pressure zone without pressure reducing
valves, based on a minimum user pressure and an maximum user pressure (300
kPa).
Friction loss (kPa/m)
Minimum user
0.25
0.6
pressure
1.25
100 kPa
20
19
18
150 kPa
15
14
13
200 kPa
10
9
9
Figure 7 shows an example with a minimum user pressure of 100 kPa, a maximum user
pressure of 300 kPa and a maximum static pressure of 400 kPa. The minimum and
maximum user pressures determine the height of the pressure zone. See also figure 8,
with the same design data, but the friction losses are higher. In figure 8 the minimum
user pressure and the maximum static pressure determine the height of the pressure zone
96 Height [m] 155
150
Static press user press.
145
h
140
135
130
125
120
0
100 200 300 400 500 600 700 800
Pressure [kPa]
Figure 7: The minimum user pressure and the maximum user pressure determine
the height of the pressure zone
Height [m] 155
static press
150
h
user press.
145
140
135
130
125
120
0
100
200
300
400
500
600
700
800
Pressure [kPa]
Figure 8: The minimum user pressure and the maximum static pressure determine
the height of the pressure zone
5.2
With pressure reducing valves
Item 4 describes that if PRV’s are used there is a primary pressure zone and a
secondary pressure zone. See figure 9. The height of the secondary pressure zone is
determined by the minimum and maximum user pressure, as explained in item 5.1
(formula 3). To adjust the pressure on the outlet of the PRV account must be taken for
the pressure loss in case of flow (pressure loss piping and PRV).
97 140
user press.
100 kPa
120
600 kPa Static press.
300 kPa
100
height [m]
80
60
40
20
0 200 400 600 800 1000 1200 1400 1600 1800
0
100 300 600 pressure [kPa]
Figure 9: Height of the secondary pressure zone with PRVs
6. Comparison of the energy consumption
Application of PRV’s result in energy loss. The energy needed for a water flow is
proportional with the multiplication of the flow (l/s) and the difference in pressure.
Assume a high rise building with apartments with a equal water demand, the energy
consumption is then proportional to the difference in pressure and the amount of
apartments or floors. Figure 10 shows two schemes where the energy consumption is
calculated. Table 5 shows a overview of the two calculated pressure zones. For this
comparison it’s assumed that the two boosting concepts have the same pump
technology.
To compare the energy consumption for both concepts the pressure difference for each
boosting system is multiplied by the amount of floors that’s served by the boosting
system. Result for the concept with two boosting systems:
Zone 1: 850 * 11 = 9,250
Zone 2: 520 * 11 = 5,720
Zone 3: 0 (Water company)
Total 15,070
Result for the concept with one boosting system:
Zone 1: 850 * 22 = 18,700
Zone 2: 0 (Water company)
98 The theorretical energgy consumpption of bo
oth conceptts are in prroportion off 15,070 too
18,700. That
T
is 1 : 1.24. Conccept 2 conssumes theorretically appproximately
y around a
quarter moore energy.
F
Figure
10: Boosting
B
cooncepts to compare
c
th
he energy cconsumptio
on.
Two booosting unitts
One boostiing unit
Heighht
(m)
zone 1
75
100
100
72
130
130
69
160
160
66 190
190
63 220
220
60 250
250
57 280
280
54 310
310
51 340
340
48 370
370
45 400
400
42 430
430
zone 2
zone 3
100
99 9
Zone 1
Z
Zone 2
39 460 130 460
36 490 160 490
33 520 190 520
30 550 220 550
27 580 250 580
24 610 280 610
21 640 310 640
18 670 340 670
15 700 370 700
12 730 400 730
9 760 430 110
760
110
6 790 460 140
790
140
3 820 490 170
820
170
0 850 520 200
850
200
The dynamic friction losses are negligible
Table 5 Comparison the energy consumption of pressure boosting concepts
To reduce the energy consumption of the booster pumps it is recommended to use a
variable-speed drive for each pressure zone. A variable-speed drive eliminates energyloss and avoid pressure fluctuations in the distribution system.
7. Hot water production
To produce hot water there are several principles:


Hot water production in the apartments
o With individual heaters, a gas-fired boiler or electric heater
o A heat exchanger connected to a central heating system or district heating
Central hot water production with circulation system
100 7.1
Hot water production per apartment
This is hot water production with individual gas-fired boilers, electrical heaters or heat
pumps in combination with exhaust air. In case of a gas-fired boiler account must be
taken for the requirements for the flue gas discharge. For hot water storage tanks keep in
mind the extra weight of the storage tank in each apartment.
7.2
Individual heat exchangers
Large city’s such as Amsterdam and Rotterdam have a district heating distribution
system operated by a private utility company. This district heating distribution system
supplies the heat for the heating in the apartment using an heat exchanger. To produce
hot drinking water there are double-wall heat exchangers to guaranty the water quality.
Both heat exchangers can be combined to one device.
The big advantage of this system is the simple lay out of the water supply system and
the individual metering. The water heaters are located close to the fixtures so there is a
minimum on heat losses in the hot drinking water system. A disadvantage is that by
maintaining a permanent high temperature of the district heating mains a heat loss
occurs in the shaft which warms the cold water pipes in the same shaft. One of the
regulations is to keep the cold water temperature lower as 25°C, so it is recommended
to separate the cold water distribution in “cold” pipe shafts.
To avoid high pressures in the central heating distribution system the systems is
seperated by several exchangers. See figure 11.
101 District heaating system Boiler heatingg system
Figure 11. Hydraulicc separatioon in the cen
ntral heatin
ng distribu
ution system
m to avoid
high static pressures..
7.3
Centtral hot waater producction
If a centrall hot water production
p
is used, thee hot water is
i distributeed from the central hot
water tank to the aparrtments withh a hot wateer pipe and a hot waterr circulationn pipe. The
f
are applicable:
general priincipals as follows

Avoid installing thhe circulation pipe in the
t apartments, use onlly supply piipes for the
users.
 The maximum
m
preessure in the distributioon system iss 600 kPa.
 The usser pressure for each appartment has a minimum
m of 200 kP
Pa.
 The maximum
m
preessure for thhe hot waterr storage tan
nk is 1000 kPa.
k
Each presssure zone neeeds a separrate hot watter system. See figure 12.
1 It’s not possible to
locate the hot water storage
s
tankks in the baasement beccause of thee high staticc pressure.
w
conceppt provides space
s
on thee floors neaar the apartm
ments.
This hot water
10
02 F
Figure
12. Central hoot water pro
oduction with
w circulaation system
m.
In a highh rise buildding where the water distributioon system iis zoned vertically too
maintain pressures
p
w
within
the maximum criiteria established, it is imperative that certainn
precautionns be obserrved in the design of the hot andd circulatioon water sy
ystem. Eachh
zone mustt be consideered as a coomplete sysstem with no
n interconnnection with
h any otherr
zone. Eacch zone muust have its own waterr heater, diistribution ppiping, circculation hott
water pipiing and circculation pum
mps. It is always desiraable to locaate the heateer at the topp
of the sysstem. The pressure
p
onn the heaterr and circullation pumpp are subjeected to aree
much less than at the base of the system.
Also for thhis system it
i is recomm
mended to seeparate the pipes in a hhot and cold
d shafts.
8. Concclusion
Many paraameters muust be considdered durin
ng the enginneering of domestic waater systemss
for high riise buildinggs, and manny possible solutions exist.
e
The eengineer mu
ust considerr
building height,
h
availlable municcipal water pressure, pressure
p
reqquirements not
n only onn
the upper floor but also
a
throughhout the buiilding, flow
w demand, bbooster pum
mp capacityy
f hot and cold, PRVss and spacee
and controol, pipe andd valve materials, riserr locations for
requiremeents in the building.
Also are im
mportant thhe economiccs, energy efficiency annd acousticss.
The enginneer has to realize
r
thatt the occupaants for thiss type of aapartments in
i high risee
buildings will be peopple paying top
t dollar an
nd expect a high qualitty plumbing
g system.
103 1
At the start of the design the plumping engineer has to determine carefully the pressure
zones to maintain the minimum and maximum pressures within the recommended
limits. The pressure on the fixtures and the demand flow determine to a significant
amount the comfort for the residents and the designer take this in accaount.
9. References
1. Water systems in high rise domestic buildings, 2008, TVVL Leusden, the
Netherlands
2. NEN 1006, General requirements for water supply installations, 2002, NEN, Delft,
the Netherlands
10. Presentation of Author
Walter van der Schee is a member of the Dutch Technical
Association for Building Installations (TVVL). He is a member of
the board of the department Sanitary Technologies (ST). The
objective of the association is to promote research and technology
in the field of building services. This is done by networking;
giving courses; lectures; organising symposia; research and cofinancing university-chairs.
For further information see www.tvvl.nl
Walter van der Schee is working at Wolter & Dros where he is
responsible for the engineering of installations in buildings.
For further information see www.wolterendros.nl
104 II.3
Evaluation of Booster Pump System in Office
Building
Saburo Murakawa (1), Kazuhiko Sakamoto (2), Yasuo Koshikawa (3) and
Hiroshi Takata (4)
(1) muraka@hiroshima-u.ac.jp
(2) sakamoto.kazuhiko@takenaka.co.jp
(3) koshikawa@hiroshima-u.ac.jp
(4) takatah@hiroshima-u.ac.jp
(1) Graduate School of Engineering, Hiroshima University, Japan
(2) Takenaka Corporation
(3) Graduate School of Engineering, Hiroshima University, Japan
(4) Graduate School of Education, Hiroshima University, Japan
Abstract
In recent years, booster pump system has increased as water supply system in high rise
buildings. Because it is very important to maintain the quality of drinking water that is
supplied by the waterworks, also the performance of pump has improved. However, in
this system, it is estimated to have many hours for pump operation. Therefore, we
measured the volume of supply water by booster pump system in an office building, and
also we clarified the condition of pump operation and the consumed amount of electric
power. We grasped the big difference in the volume of water supply between the
measurement data and the planning values. As for the estimation of suitable water
supply loads in the office building, we applied the Monte Carlo simulation technique
105 that has been developed by us. On the basis of the estimation loads which were
calculated by the different numbers of people in the building, we selected the suitable
capacity of booster pump system. In comparison with the results of the suggested
method by us and the existing method, we clarified that it was possible to control the
increase of the consumption of electric energy by using the suitable water supply system
based on the high accuracy for water supply demands estimation. The energy
consumption by the booster pump system showed the values of 1.5-2.9 times in
comparison with that by the gravity tank system.
Keywords
Office building, Water supply demands, Calculation method, Measurement, Booster
pump system, Energy conservation
1. Introduction
In recent years, booster pump system instead of the gravity tank system increased as
water supply system in high rise buildings because of the maintenance for drinking
water quality that is supplied by the waterworks and the improvement of the pumping
performance. However, in this system, it is estimated to have longer time for running of
the pump than that of the gravity supply system. Therefore, we have concern about the
energy consumption in the booster pump system.
From the view point of energy conservation in the buildings, it is very important to
design the water supply system by suitable capacity and control for the pumps on the
basis of the high accuracy estimation of the water supply demands.
In this paper, we clarify the electric energy consumption and water supply volume as
the running conditions of the booster pump system from the measurement results in an
office building through one week. Then, we have the estimation of the electric energy
consumption and water supply volume on the basis of the several calculating conditions,
especially as for the number of workers in the building by the simulation technique.
From these analyses, we try to evaluate the booster pump water supply system on the
energy conservation.
106 2. Running conditions of the booster pump system
2.1 Outline of the measurement
We carried out the measurement in a rental office building named as F office building
located at the central area in Hiroshima city.
Table 1 and Figure 1 show the outline of F building and water supply system
respectively. The water supply system has two kinds of system as drinking water supply
system and non-drinking water supply system. The water is supplied by the city
waterworks. However, the water supply system inside the building is divided into two
service lines. As for one system to supply water to hot water service rooms and basins
and slop sinks in the toilets, and another system to supply water to water closets and
urinals in the toilets, we describe the service lines as “drinking water (supply) system”
and “non-drinking water (supply) system” respectively.
Table 1-Outline of F office building
Location
Site area
Building area
Total floor area
Number of stories
Construction period term of work
Number of workers to an enrollment
Water supply system
Number of fixtures
Hiroshima city
2
2,036.79m
2
1,423.11m
2
18,256.39m
14 stories above and one under ground, one penthouse
July 2000 - January 2002
881 people ( Male : Female = 78.3 : 21.7 % )
Booster pump system with the receiving tank
Male toilet
Water closet:38, Urinal:38, Basin:38, Slop sink:13
Female toilet Water closet:38, Basin:38
Hot water service room Electric hot water tank with a tap:13
Single lever mixing tap:13
107 WC (Male)
Hot water
service room
WC (Female)
WC (Male)
Hot water
service room
WC (Female)
WC (Male)
Hot water
service room
WC (Female)
14F
.
.
.
5F
WC (Male)
Hot water
service room
WC (Female)
WC (Male)
Hot water
service room
WC (Female)
WC (Male)
Hot water
service room
WC (Female)
4F
.
.
.
1F
Water meter
Drinking water
receiving tank
Non-drinking water
receiving tank
Booster pump (Drinking)
Booster pump (Non-drinking)
B1F
Non-drinking water system
Drinking water system
Figure 1-Water supply system
Table 2 shows the outline of two kinds of water supply system. Each system has one
receiving tank and one unit of the booster pumps. The measurement was carried out in
November, 2005. The items and conditions of the measurement are shown in Table 3.
Table 2-Outline of the water supply system
Receiving tank
Material of the tank
Capacity
Number of pumps
Control system
Pump
Operation system
Discharge flow rate
Moter capacity
Pump head
Drinking system
FRP panel
3
3
71m
31m
3 pumps
4 pumps
Inverter driven induction moter control with the fixed
pressure at the end of piping system
Alternately parallel running
300L/min
780L/min
7.5kW × 2
7.5kW × 3
900kPa
800kPa
108 Non-drinking system
Concrete
Table 3-Items of the measurement
Item
Water supply volume
Electric consumpution
of the pumps
Inverter frequency
of the pumps
Water supply pressure
Measuring term
07(Mon.) Nov. 2005, 17:00 - 13(Sun.) 16:00
13(Sun.) Nov. 2005, 17:00 - 14(Mon.) 20:00
Recording interval
60 seconds
10 seconds
Measuring device
07(Mon.) Nov. 2005, 17:00 - 14(Mon.) 20:00
60 seconds
Watt-hour meter
14(Mon.) Nov. 2005, 06:30 - 20:00
Continuously
Video camera
14(Mon.) Nov. 2005, 06:30 - 20:00
Continuously
Video camera
Ultrasonic flow meter
2.2 Measurement results
As the results of measurement through one week, Figure 2 shows the hourly fluctuations
of water supply volume and electric energy consumption. The high demands of drinking
and non-drinking water appear in daytime. The non-drinking water supply volume
shows about five times in comparison with the drinking water supply volume. However,
the electric energy consumption in each system shows the almost same values between
2,000 Wh and 2,500 Wh.
As the more detailed fluctuations, Figure 3 shows the water supply volume and electric
energy consumption per one minute in the time zone of 07:00 and 12:00 on Monday. In
the time zone of 07:00 when the water supply volume is small quantity, there is not
difference on the electric energy consumption of pumps between less than 10 L/min and
about 20L/min of water volume. It consumes about 40 Wh in each water supply volume.
4000
3000
3000
Electric energy consumption
[ Wh ]
Water supply volume
[ L/h ]
In the time zone of 12:00 when the water supply volume is large quantity, the electric
energy consumption shows the same tendency of fluctuation toward the water supply
demands. It consumes 70 Wh on the electric energy for the maximum water supply
volume; 120 L/min.
2000
2000
1000
1000
0
0 12 0 12 0 12 0 12 0 12 0 12 0 12 0 [h]
Mon. Tue. Wed. Thur. Fri.
Sat.
Sun.
Non-drinking water system
0
0 12 0 12 0 12 0 12 0 12 0 12 0 12 0 [h]
Mon.
Tue.
Wed.
Thur.
Fri.
Sat.
Drinking water system
Figure 2-Hourly water supply volume and electric energy consumption
109 Sun.
100
60.0
50.0
80
40.0
60
30.0
40
20.0
20
10.0
0
0.0
7:00
7:10
7:20
7:30
7:40
7:50
Water supply volume
電力量
給水量
140
80.0
120
70.0
100
60.0
50.0
80
40.0
60
30.0
40
20.0
20
10.0
0
12:00
0.0
12:10
12:20
12:30
12:40
給水量
Electric energy consumption
12:50
電力消費量(
W h)
Electric energy
consumption
[Wh]
70.0
給水量(
L/m in)[L/min]
Water supply
volume
80.0
120
W h)
Electric電力消費量(
energy consumption
[Wh]
給水量(
L /m in)[L/min]
Water supply
volume
140
電力量
70.0
100
60.0
50.0
80
40.0
60
30.0
40
20.0
20
10.0
0
0.0
7:00
7:10
7:20
7:30
給水量
7:40
140
80.0
120
70.0
100
60.0
40.0
60
30.0
40
20.0
20
10.0
7:50
Water supply volume
電力量
50.0
80
0
12:00
0.0
12:10
12:20
12:30
Electric給水量
energy consumption
12:40
12:50
電力消費量(
W h)
Electric energy
consumption
[Wh]
80.0
120
給水量(
L/m in)[L/min]
Water supply
volume
140
電力消費量(
W h)
Electric energy
consumption
[Wh]
給水量(
L/m in)
Water supply
volume
[L/min]
Drinking system
電力量
Non-drinking system
Figure 3-Water supply volume and electric energy consumption
Figure 4 shows the fluctuation pattern of instantaneous flow rates per one minute as an
example of measurement for the non-drinking service line through one day. The
instantaneous flow rates fluctuate widely from about 07:00 to about 22:00.
Figure 5 shows a histogram at 10 L/min interval for the instantaneous flow rates in the
non-drinking system. The frequency of flow rates from 0.1 L/min to 10 L/min accounts
for 50 percent of all frequencies excepting at zero.
Figure 6 shows the fluctuation patterns of water supply volume, water supply pressure
and inverter frequency of the pumps in the non-drinking system as the mean values of
ten seconds interval. The water supply pressure drops according to the occurrence of the
water supply demands and the pumps start immediately to recover the dropped pressure.
The pumps run with the scheduled rotation successively. In spite of low demands from
0 L/min to 12L/min on the time zone of 07:00 and high demands from 1L/min to 27
L/min on the time zone of 12:00, the pumps run at about 50 Hz of inverter frequency.
Every time of decrease of the water supply volume, the pumps run and stop by turns. At
the alternate time of running pump, the water supply pressure drops some quantity.
However, the water supply pressure is kept at nearly constant value in spite of the
fluctuation of water supply volume.
Figure 7 shows the relationship between the water supply volume and electric energy
consumption. The electric energy consumption converted into one minute interval
shows small values because the duration of pump operation is less than one minute
110 when the water supply volume is less than 10 L/min. In case of the water supply volume
between 20 L/min and 60 L/min, the pump system consumes about 30 Wh- 45 Wh of
electric energy
給水量[L/m
Water supply
volumein]
[L/min]
160
160
140
140
120
120
100
100
80
80
60
60
40
40
20
20
00
0:00
0:00
4:00
4:00
8:00
8:00
12:00
12:00
16:00
16:00
時刻
[h]
20:00
20:00
800
800
100
100
700
700
90
90
80
80
600
600
70
70
500
500
60
60
400
400
50
50
300
300
40
40
200
200
30
30
20
20
100
100
累積度数[%] frequency
Accumulated
[%]
Frequency
頻度 [Times]
Figure 4-Fluctuations of water supply volume in the non-drinking system (08 Nov.)
10
10
00
0.1~
0.1-
10~
10-
20~
20-
30~
30-
40~
40-
50~
50-
60~
60-
70~
70-
80~89
80- 90以上
90-
00
給水量[L/mi
n]
Water supply volume [L/min]
給水量
Water
supply volume
給水圧
111 0.5
0.4
15
0.3
0.2
10
5
0.1
0
Water supply pressure
給水量
給水圧
1 2 :4 1
1 2 :4 0
1 2 :3 9
1 2 :3 8
1 2 :3 7
1 2 :3 6
1 2 :3 5
0
給 水pressure
圧 (M P a )[MPa]
Water supply
20
1 2 :3 4
7 :2 1
7 :2 0
7 :1 9
7 :1 8
7 :1 7
7 :1 6
7 :1 5
0
7 :1 4
0
7 :1 3
5
0.2
0.1
7 :1 2
10
0.4
0.3
7 :1 1
15
0.7
0.6
1 2 :3 3
0.5
25
1 2 :3 2
20
0.8
1 2 :3 1
0.6
30
1 2 :3 0
25
Water supply
volume
給水量
(L / m in[L/min]
)
0.8
0.7
給 水 圧pressure
(M P a ) [MPa]
Water supply
30
7 :1 0
Water supply
給 水 量volume
(L / m in[L/min]
)
Figure 5-Histogram for water supply volume of non- drinking system (08 Nov.)
N O .1ポンプ
NO .2ポンプ
No.2 Pump
No.1 PumpN O .4ポンプ
NO .3ポンプ
7:00 zone
No.3
Pump
NO
.1ポンプ
No.4
Pump
NO
.2ポンプ
NO .3ポンプ
1 2 :4 1
1 2 :4 0
1 2 :3 9
1 2 :3 8
1 2 :3 7
1 2 :3 6
1 2 :3 5
1 2 :3 4
1 2 :3 3
1 2 :3 0
7 :2 1
7 :2 0
7 :1 9
7 :1 8
7 :1 7
7 :1 6
7 :1 5
7 :1 4
7 :1 3
7 :1 2
7 :1 1
10
0
1 2 :3 2
40
30
20
80
70
60
50
40
30
20
10
0
1 2 :3 1
Inverter周frequency
波 数 (H z )[Hz]
60
50
7 :1 0
Inverter
周 frequency
波 数 (H z ) [Hz]
80
70
NO .4ポンプ
12:00 zone
Figure 6-Running conditions of the pumps and water supply volume
電力量[W
h]
Electric energy
consumption
[Wh]
in the non-drinking system
90
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
00
00
20
20
30
40
40
60
50
80
60
100
70
120
Water supply volume [L/min]
80
90
140
160
給水量[L/m in]
Figure 7-Relationship between the water supply volume and
the electric energy consumption of the pump
2.3 Analysis of pump operation
The required water supply volume and pressure in the non-drinking water supply system
of F building were planned by the parallel running of three pumps. The pump system
has the performance curves shown in Figure 8. In the case of less than 400 L/min for
water supply volume, only one pump operates. Figure 9 shows the performance of one
pump. The relationships between the discharge flow rates and electric energy
consumption are shown as the following equation because of the fixed pressure control
at the end of piping line by the inverter control pump system.
y  22.18x 2  7.67 x  2.12
(1)
112 In which x = water supply volume [m3/min]
x>0 ; and y = electric energy consumption [kW/h].
When the pump is run with more small discharge flow rates than the rated capacity, the
electric energy consumption of the pump is not large difference between the high loads
and the low loads. Though the water supply loads are very small, the electric power of
pump is required the energy to pump up water to the highest piping line of water supply
system.
When we plot the measurement flow rates ; from 0.1 L/min to 50 L/min which account
for 90 percent of all measurement frequencies, on the performance curve of the pump,
the calculated electric energy consumption corresponds to the measurement result
between 2.1 kWh and 2.5 kWh. From these results, it can be understand that the
duration time of water uses has a great influence on the electric energy consumption of
the booster pump.
As for the electric energy consumption per day, Table 4 shows the measurement result
and the calculation result which is followed with the equation (1). The both results
accumulated through a day correspond with each other. Therefore, we apply the
equation (1) to estimate the electric energy consumption for the booster pump by the
simulation technique.
Total head of water [m]
120
100
80
60
40
Operation of
one pump
20
Operation of
two pumps
Operation of
three pumps
0
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
Discharge flow rate [m3/min]
Figure 8-Performance curve of the pumps as an example of parallel running
113 Electric energy consumption [kWh]
Q-H
71
Head
loss
100
65
Fixed pressure control
at the end of piping system
Actual head
Total head of water [m]
100
95%
90%
85%
0
Q
8.7
7.1
5.8
y=22.18x 2+7.67x+2.12
3.9
2.1
0
0.16
0.27
0.33
Q
0.40
3
Discharge flow rate [m /min]
Figure 9-Performance curve of the pump as an example of one running
Table 4-Comparison of the electric energy consumption of
measurement and simulation
Electric energy consumption [ kWh/d ]
Measurement
Simulation
29.04
29.61
3. Estimation of the appropriate water supply volume
There is a wide difference between the design value and the measurement value for the
water supply volume. Therefore, the booster pump system that was designed in this
building does not operate with the full efficiency. In the case of the booster pump
system of which the worked conditions are influenced directly by the fluctuation of
water supply volume, it is very important to estimate the water supply demands
appropriately.
We have developed the calculation method of cold and hot water demands in some
kinds of buildings for use, which is applied by the Monte Carlo Simulation technique
[1]-[5].
In this paper, we apply the models that were suggested in office buildings [4], and
estimate the water supply demands for non-drinking system as to the difference of the
number of workers. On the basis of the results, we suggest the suitable capacity of the
booster pump system in this building.
114 3. 1 Calculation models of water demands
On the basis of the previous paper [4], we show the calculation models of water supply
demands in Table 5. The simulation is tried at hourly intervals for 100 times through
07:00 to 24:00. The calculation models for toilet are composed of four sub-models ;
arrival model to each fixture, occupancy model in each fixture, water volume model
discharged from each fixture and fixture operation model.
Table 5-Calculation models of water demands
Male
Female
Water closet Urinal Basin Water closet Basin
Arrival ratio
Arrival
model
Set in each times
[person/min]
[people/min]
Arrival interval
distribution
Duration time of
occupancy [sec]
Distribution
Duration time of
water discharge [sec]
Distribution
Occupnacy
model
Water volume
model
Fixture operation
model
Slop sink
Exp.
Erl.3
Erl.7 Hyp.2
6.5
5
Exp.
Flow rate [L/min]
120
Distribution
Frequency of
fixture operation
Erl.6
12
37
260
17
56
Erl.3
Hyp.2
Hyp.3
6.5
11
56
Exp.
Erl.3
Hyp.3
120
5
12
Erl.6
Erl.10
Erl.3
1.17
1
1
6
Erl.10 Erl.3
30
5
Erl.10 Erl.10
1.37
110
1
1
1.2
Frequency of fixture usage
[frequency/(people・
h)]
[frequency/(person
h)]
Frequency of fixture usage
[frequency/(people・
h)]
[frequency/(person
h)]
When we apply the simulation models shown in Table 5, we have to set up the arrival
ratio in each fixture through the calculation time zone. Therefore, the simulation is
supposed that the arrival phenomena to each fixture show the random patterns for uses.
The hourly frequencies of fixture usage per people occupied in each office floor are
shown in Figure 10 as to the average values. The mean arrival ratios in each fixture are
calculated by Figure 10.
a) Male
1
0.8
0.6
0.4
0.2
0
7
9
11
13
Water closet
15
17
Urinal
19
21
23 [h]
Basin
3
b) Female
2
1
0
7
9
11
13
Water closet at 12:00-13:00
Restaurants exist)
(
Water closet
Figure 10-Frequencies of fixture usage
115 15
17
19
21
23 [h]
Water closet at 12:00-13:00
No restaurant for workers)
(
Basin
3.2 Estimation of the water supply volume
3.2.1 Setting up for the number of workers
In case of office building, the hourly ratios for the number of workers in each office
floor are set in four models shown in Figure 11. In this paper, we apply the average
model as to the occupied ratio for the number of workers to an enrollment. The total
numbers of workers are set in three groups, which are 2,556 people, 1,278 people and
881 people.
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
Ratio of the number of workers
occupied in office to an enrollment
Ratio of the number of workers
occupied in office to an enrollment
The effective floor area multiplied by the density of occupied people; 0.2 people/m2 that
is used in general for an office planning makes the most numerous people ; 2,556
people. The fewest people ; 881 people is the measurement result in this building. The
number of 1,278 people mediates between two sides. The measurement result shows
about 20.7 m2 for the total floor area per one person, which is closely to the latest mean
value ; about 24 m2 investigated in Japanese office buildings recently.
7
9
11
Avg
13
15
Max
17
19
Min
21
23[h]
All
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
7
9
Avg
a) Male
11
13
15
17
Max, All
19
21
23[h]
Min
b) Female
Figure 11-Hourly ratios of the number of workers occupied
in the office to an enrollment
3.2.2 Daily water supply volume
Table 6 shows the daily mean values of water supply volume which were measured and
calculated in three groups of the number of workers. The calculation results show the
mean values that were accumulated for every one minute intervals from 07:00 to 24:00
on the basis of 100 trials of the simulation.
The measurement result is larger than the calculation result when we compare with the
same number of workers ; 881 people. It seems that the measurement result includes the
consumption by the guest people. The daily water supply demands are in proportion to
the number of workers to an enrollment.
116 Table 6-Daily mean water supply volume
Simulation
Measurement
881 people
881 people
1,278 people
2,556 people
25,944
22,803
32,432
63,708
Non-drinking water supply system [ L/d ]
3.2.3 Hourly water supply volume
Figure 12 shows the results for the hourly water supply volumes of the measurement
and simulation which were carried out on the basis of 881 people as the number of
workers. The measurement results have a little high value in comparison with the
simulation results. However, the hourly fluctuations of the measurement are similar to
the patterns of the simulation. A little different between two things will be interpreted
with the above-mentioned reason.
Water supply volume [L/h]
Figure 13 shows the hourly fluctuation patterns which were calculated with each
number of workers. The peak hour in each case appears at the time zone of 10:00.
4000
3000
2000
1000
0
0
2
4
6
8 10 12 14 16 18 20 22 24 [h]
Simulation
算定値
Measurement
実測値
Water supply volume [L/h]
Figure 12-Hourly water demands based on the simulation and measurement
as an example of 881 people
8000
6000
4000
2000
0
0
2
4
6
8 10 12 14 16 18 20 22 24
2,556 people
1,278 people
881 people
[h] Figure 13-Hourly water demands based on the difference of the number of
workers
117 3.2.4 Instantaneous maximum flow rates
Figure 14 shows the instantaneous maximum flow rates of the measurement and
simulation which were carried out on the basis of 881 people as the number of workers.
The every values calculated by the simulation are shown in each failure factor as the
water supply demands per one minute. The maximum value of measurement in each
hourly zone approximates to the value of simulation within the range from 0.2 % to 5%
of failure factor. As for the planning of capacity for building facilities, it is very
important to estimate the instantaneous maximum flow rate on the basis of suitable
failure factor.
Simulation max.
300
Simulation 0.1%
250
Simulation 0.2%
200
Simulation 2%
給水量(
L/min)
Water supply volume [L/min]
350
Simulation 5%
100
80
60
40 Simulation 10%
20
0
Simulation
25%
10
12
14
16
020
18
222
424
68
150
100
50
Measurement max.
0
0
2
4
6
8
10
12
14
16
18
20
22
24
[h]
Figure 14-Instantaneous flow rates based on the simulation and measurement
4. Analysis of the booster pump system
4.1 Setting up for the suitable capacity of booster pump
When the booster pump system is applied for the water supply system in buildings, it
seems to increase the electric energy consumption of the pump because the duration of
running will be longer than the other systems such as the gravity tank system etc.
In case of the booster pump system, the water supply pressure at the highest end of
piping system has to be kept with the nearly constant value that is designed. Therefore,
the ratio of the actual head to the total head for pumping up water in high-rise building
is larger than that in low-rise building. It means that the ratio of the energy loss due to
the fluid resistance in the pipe lines is small according to be high-rise building.
Therefore, to set up the pump of suitable capacity for energy consumption is required
118 from the viewpoint of energy saving based on the estimation of high accuracy for the
water supply demands.
In this Paragraph 4.1, we set up the water supply systems based on the two kinds of
methods to estimate the water supply demands. And we compare the effect of saving
energy in each system.
(1) Pump capacity by the existing method
The instantaneous maximum flow rate that is calculated by existing method ; the
Hunter’s method used the fixture loading unit, is shown in Table 7. The maximum flow
rate on the basis of the total number of fixture loading unit can be calculated by the
following equation that is shown as the Hiroshima Waterworks Standard.
y  10(0.673  log x  0.859)
(2)
in which x = total of the fixture loading unit ; and y = instantaneous maximum flow
rate (L/min).
On the basis of the instantaneous maximum flow rate ; 733 L/min and the total head to
pump up water ; 800 kPa that was set up as the designed value, the booster pump
system operated by the rotation control is set up with the following specifications,
which is selected from the catalog of a pump maker.
950 L / min  800 kPa  7.5 kW ( INV )  3 pumps
Table 7-Calculation result by the existing method
( the Hunter's method used the fixture loading unit )
Fixture unit
Number of fixture
10
38
5
38
10
38
10
1
Total of the loading unit [ x ]
Instantaneous maximum flow rate [ L/min ] [ y ]
Male Water closet
Urinal
Female Water closet
Hundicapped Water closet
Loading unit
380
190
380
10
960
733
(2) Pump capacity by the simulation method
As for the non-drinking water supply system, Table 8 shows the instantaneous
maximum flow rates in each failure factor calculated by the Monte Carlo Simulation
method. The measurement maximum value is closely to the simulation value chosen by
the failure factor 0.2% in case of 881 people to an enrollment. In case of 2,256 people
that is supposed to be the number of workers decided at the actual design process, the
119 instantaneous maximum flow rate chosen by the failure factor 0.2% shows the value of
259 L/min, which is about the value of 35% to the actual designed value ; 733 L/min.
On the basis of the simulation results, the booster pump system is set up by the value of
failure factor 5% as the standard capacity of a pump because of saving energy
consumption not to be over specifications for water supply demands. When the more
large demands are required to the designed capacity of a pump, one more pump
composed of two pumps system set up for the estimated maximum load will be
operated. The requirement specifications of the pumps for each number of workers are
shown in Table 9. However, when we select the suitable capacity of a pump from a
maker’s catalog, the specifications for each number of workers are same because the
power of a pump is required to keep the head for pumping up water. Therefore, the
booster pump system is set up with the following specifications.
460 L / min  750 kPa  5.5 kW ( INV )  2 pumps
Table 8-Instantaneous maximum flow rates based on the difference
of the number of workers
Number of workers
to an enrollment
Non-drinking
system
[ L/min ]
881 people
1,278 people
2,556 people
Max.
295
320
395
Simulation
Failure factor [ % ]
0.1
0.2
1
161
146
114
190
176
139
279
259
214
2
98
122
192
5
77
98
162
Measurement
Failure factor [ % ]
Max.
0.1
0.2
149
113
109
-
Table 9-Requirement specifications of the pump for each number of workers
Number of workers
to an enrollment
881 people
1,278 people
2,556 people
Capacity per one pump
(F. factor : 5%) [L/min]
80
100
160
Max. flow rate
(F.factor : 0.2%) [L/min]
150
180
260
4.2 Comparison of the electric energy consumption
As for the daily water supply demands, the average values were calculated by the every
one minute interval in each number of workers.
(1) In case of the existing booster pump system
The electric energy consumption is calculated by the equation (1). When the water
supply volume is larger than the value of 400 L/min, two pumps are operated. In this
case, the electric energy consumption of two pumps is calculated by the 8.74 kWh for
120 running of one pump plus the value calculated with the equation (1) using the water
supply volume except the value of 400 L/min.
(2)In case of setting up the new booster pump system
As for the new booster pump system mentioned in Paragraph 4.1, the relationship
between the water supply volume and the electric energy consumption is shown with
the following equation, which is set up from the performance curve that was exhibited
by a pump maker.
y  32.78x 2  7.70x  1.66
(3)
in which x = water supply volume [m3/min]
x>0: and y = electric energy consumption [kW/h].
When the volume of water supply demands exceeds the value of 280 L/min, two pumps
are operated. In this case, the electric energy consumption for two pumps is calculated
with the same method above-mentioned.
Table 10 shows the electric energy consumption calculated by the simulation for each
capacity of the booster pump system. In spite of the same water supply demands for
each booster pump system, the electric energy consumption in the new booster pump
system is smaller than that of the existing booster pump system. Therefore, the new
system reduces the volume of energy consumption about 14-19% to that of the existing
system. And the energy consumption in 881 people as the number of workers shows
about 30% reduction to that in 2,556 people though the number of workers ; 881 people
Table 10-Electric energy consumption for each capacity
of the booster pump system
Water supply
system
Existing system
7.5kW × 3
pumps
New system
5.5kW × 2
pumps
Number of
workers
[people]
881
1,278
2,556
881
1,278
2,556
Water supply
volume
[L/d]
22,803
32,431
63,708
22,803
32,431
63,708
Electric energy
consumption
[kWh/d]
30.91
33.61
41.68
25.15
27.67
35.96
is reduced about 35% against 2,556 people.
121 Ratio of the electric energy consumption
by the same people
by the different people
in the different systems
in the same system
1
0.74
1
0.81
1
1
0.81
0.70
0.82
0.77
0.86
1
Table 11 compares the electric energy consumption with the new booster pump system
and the gravity tank system. The booster pump system shows about 1.5~2.9 times of
energy consumption in the gravity tank system.
Number of
workers
to an enrollment
[people]
881
1,278
2,556
Water supply Booster pump
volume
system
[L/d]
[kWh/d]
22,803
32,431
63,708
25.15
27.67
35.96
Gravity tank system
Duration of runnig
[min]
81.4
115.8
227.5
Energy consumption
[kWh/d]
8.67
12.34
24.23
Ratio of energy
consumption
Booster system
to Gravity system
2.90
2.24
1.48
Table 11-Comparison of the electric energy consumption
between the booster pump system and the gravity tank system
5. Conclusion
In this paper, we clarified the water supply demands and the running conditions of the
booster pump system in an office building.
In case of the water supply system in high rise buildings, the ratio of energy loss due to
the fluid resistance in the pipe lines is small against the total head to pump up water.
Therefore, it cannot be expected to have high effect for saving energy consumption in
the booster pump system equipped with inverter control of frequency. Especially, as for
the low flow rates in piping systems, there is not large difference for the electric energy
consumption of the booster pump between large and small demands.
On the basis of the estimation of water supply demands calculated by the Monte Carlo
Simulation technique which has been developed by us, we showed an example of
setting up the booster pump system.
As for the instantaneous maximum flow rates, the calculation value by the simulation in
the case of 2,556 people showed about 35% to the actual designed value that was
calculated by the existing method; the Hunter’s fixture loading unit.
The electric energy consumption of the new booster pumps that was set up with the
calculation water supply demands showed the reduction volume about 14-19% to that
calculated by the existing specifications for the booster pump system.
The new suggested booster pump system showed about 1.5~2.9 times of energy
consumption in comparison with the gravity tank system.
In this study, we selected the specifications of the booster pump from the existing
equipment that is on the market. However, if we hope to have more high efficiency for
energy saving in the booster pump system, it will be expected to develop the pump
122 system with the performance curve of high head to pump up water and small discharge
flow rate.
Furthermore, it is better to set up the control system to maintain the water supply
pressure etc. for the booster pump system from the view point of energy saving.
Lastly, we point out that it is an essential condition to have high accuracy estimation of
water supply loads as for planning of energy saving in the water supply system.
Acknowledgment
The authors wish to express their gratitude for the subsidy of science research from the
Ministry of Education Culture, Sports, Science and Technology (Project number :
19560595, Research subject : Calculation of cold-hot water supply demands and
instrument’s capacity in buildings aimed for the international standardization, Head
investigator : Saburo Murakawa).
6. References
1. S. Murakawa and H. Takata: Development of the Calculating Method for the Loads
of Cold and Hot Water Consumption in the Apartment Houses; Proceedings of the
CIB/W62 International Symposium on Water Supply and Drainage for Buildings
(Ankara), pp.281-295 (2003)
2. S. Murakawa, H. Takata and D. Nishina: Development of the Calculating Method for
the Loads of Water Consumption in Restaurant; Proceedings of the CIB/W62
International Symposium on Water Supply and Drainage for Buildings (Paris), A-7,
pp.1-14 (2004)
3. S. Murakawa, D. Nishina, H. Takata and A. Tanaka: An Analysis on the Loads of
Hot Water Consumption in the Restaurants; Proceedings of the CIB/W62
International Symposium on Water Supply and Drainage for Buildings (Brussels), F3, pp.1-10 (2005)
4. S. Murakawa, H. Takata and K. Sakamoto: Calculation Method for the Loads of Cold
and Hot Water Consumption in Office Buildings based on the Simulation Technique;
Proceedings of the CIB/W62 International Symposium on Water Supply and
Drainage for Buildings (Taipei), G-2, pp.1-14 (2006)
5. S. Murakawa, Y. Koshikawa, H. Takata and A. Tanaka: Calculation for the Cold and
Hot Water Demands in the Guest Rooms of City Hotel; Proceedings of the CIB/W62
123 Internatiional Sympposium on Water
W
Suppply and Draainage for Buildings
B
(B
Brno), B-1,
pp.73-855 (2007)
6. K.Sakam
moto, S.Murrakawa, Y.K
Koshikawa and H.Takaata: Evaluattion for Boooster Pump
System in
i office Buuildings ; J.E
Environ.Enng., AIJ, No.622, pp.89-94 (2007) (Japanese)
7. Presen
ntation off Author
Saburo Murakawa
M
i the Em
is
meritus Proffessor at Hiroshima
H
University. His speciaal fields aree building and
a city en
nvironment
engineering, plumbbing engiineering and enviironmental
c
psychologyy. He is noow studyingg to developp the new calculation
technique for
f cold andd hot water demands inn various bu
uildings to
establish thhe standard method.
12
24 II.4
Consideration on the water environment
performance of the architecture
H. Kose
hkose@toyonet.toyo.ac.jp
Faculty of Information Sciences and Arts, Toyo University, Japan
Abstract
From the research situation of water supply and drainage system and water environment
field in recent Japan, this paper showed the approach of the water environment
performance in the architecture. Moreover, the issues that should cooperate with other
research field were shown.
Water environment steering committee in Architectural Institute of Japan (AIJ)
announced "Approach on the formation of the water environment in the building and
circumference" in December 2008. In the script, five items on "security of the safety
and maintenance of the health", "symbiosis with the nature", "resource conservation and
energy conservation", "formation of the society assets", "succession to the future
generation" have been advocated. This content is introduced.
Moreover, working group for water environment performance on architecture of AIJ
selected eight evaluations of the water related equipments of the housing. These are
"safety and sanitary", "functionality", "durability", "maintenance", "environmental
loading reduction", "suitability to the architecture", "amenity", "correspondence of
elderly persons". The result of trying this evaluation in actual housing is shown.
In addition, the outline is shown on development situation of "The BEST Program"
which is the Integrated Energy Simulation Tool for Buildings and MEP Systems.
Moreover, the simulation result on energy consumption and water consumption on the
rainwater utilization is shown.
125 Finally, the issues to be tackled by the field of water supply and drainage system and
water environment cooperating with other architecture field in future, is shown.
Keywords
Water environment performance; Sanitation; Infrastructure; Hot water supply;
Architectural planning.
1. Introduction
Water environment field in the architecture of Japan is handled as one of the building
equipment, and the positioning as an environmental engineering is insufficient. The
researcher of this field in Japan is very few. In Japan, the field of the building
engineering is distinguished from the civil engineering field, and research and education
field, which are identical with architecture design and building construction field have
been formed. However, the existence in the architecture has not been very much
recognized in present state in the water, since the infrastructure of water supply and
sewerage exists.
It is not possible to use present building, when supply and drainage of the water are not
possible. And, they are necessary as a lifeline of the building. In addition, the water
environment problem is a global problem. Quantitatively and qualitatively sufficient
water is necessary to safe and sanitation of human and region.
It is a purpose of this paper to recognize the positioning of the water environment in the
architecture to other field. Mainly on the architecture including city and earth, the
challenge of the research in Japan is outlined on the water environment performance.
And, the matter with other field to should be cooperative will be discussed in future.
Still, this paper reconstituted the content lectured in the research council of
environmental engineering department at Annual meeting on Architectural Institute of
Japan (AIJ) 2008.
2. Activity on water environment steering committee of AIJ
Water environment steering committee in AIJ is active two fields concerning water
supply and drainage system and water environment. It is active, while facility and
principle and design field cooperate in the wide frame of "water". In fiscal 2009, next
five subcommittee and special committee are active including the committee.
126 1. The formation examination subcommittee of the sound water environment.
Popularization and enlightenment pamphlet for the architect on "Approach on the
formation of the water environment in the building and circumference" is being made.
2. The rainwater architecture standardization subcommittee.
The work on the decision of Architectural Institute of Japan Environmental Standard
(AIJES) on the rainwater utilization followed for DIN 1989 of Germany has been done.
In fiscal 2009, it became a responsible organization of AIJ engineering department
design competition "The architecture which enjoys rain and controls water in the city".
3. City and water-familiarization subcommittee.
The research on the action of the human and organism in the waterfront, and the
community planning of the citizen subject mainly on the waterfront are carried out.
4. Effective advantage utilization special research committee in the woody biomass
resources.
Concerning the maintenance of global environment by the reduction of CO2, pickup
and the effective advantage utilization are examined from the viewpoint of architecture
field in respect of the woody biomass.
5. Architecture and urban environment future model special research committee
It is a committee for the purpose of making up environment model of architecture and
city by each particular field cooperating. At present, the parameter to be introduced into
architecture and urban environment model in the existent city is examined.
Still, in this steering committee, AIJES-W001-2009: "Guideline for architectural
planning and waste treatment system which contributes to recycling promotion of nonindustrial wastes, operations management of large-scale office and commercial facility
and multiple dwelling house" it was published in February, 2009. It is first AIJES as an
water environment field.
127 3. The water environment performance of architecture, city and earth
3.1 The water environment performance in the architecture
1. "Approach on the formation of the water environment in the building and
circumference"
In "Approach on the formation of the water environment in the building and
circumference" announced in December, 2008, the five matters of "security of the safety
and maintenance of the health", "symbiosis with the nature", "resource conservation and
energy conservation", "formation of the society assets", "succession to the future
generation" has been raised as the matter in which all the humans who participate in the
architecture should try (Table-1).
Table-1 Contents of "Approach on the formation of the water environment in the
building and circumference" 1)
2. Environment performance of the kitchen and sanitary space
In the Working group for water environment performance on architecture, which was
active over fiscal 2001-2004, kitchen and sanitary space in which the improvement
needs in the housing was high were noticed, and the method for comprehensibly
128 displaying these kitchen and sanitary space and performance of plumbing system was
examined.
In this WG, eight evaluations of "safety and sanitary", "functionality", "durability",
"maintenance", "environmental loading reduction", "suitability to the architecture",
"amenity", "correspondence of elderly persons" were set as environment performance of
kitchen and sanitary space.
It was tried that it established evaluation point of the 5 stages on multiple evaluation
items involved for these and displays the water environment performance as a space
according to the radar chart (Figure-1).
Figure-1 Radar chart of water environment performance in the kitchen 2)
3.2 Water environment performance to be examined, while the architecture
considers urban and global environment.
1. The evaluation on rainwater utilization and wastewater reuse and these runoff
controls.
Rainwater utilization and wastewater reuse have been examined until now from the
viewpoint of quantity example for effective utilization of water resources, reduction of
tap water and the water quality.
At present, in "The BEST Program" which is the Integrated Energy Simulation Tool for
Buildings and MEP Systems (the following BEST) promoted by Institute of Building
Environment and Energy Conservation (IBEC), it can be calculated the energy
consumption on rainwater utilization and wastewater reuse with air conditioning
equipment and electric equipment (Figure-2).
129 The energy consumption increases on rainwater utilization and wastewater reuse, since
energy consumption for double piping and water purification increases, when it is
observed in the building of the simple substance. However, it is connected for the
reduction in the energy consumption concerning water supply in the city scale. And, the
countermeasure to city flood alleviation for the local heavy rain is also important
environment performance required for the architecture. It is necessary to appropriately
evaluate effect of decreasing of the load for the city infrastructure.
Figure-2 Total energy of rainwater utilization in the apartment house 3)
2. The evaluation in the water usage of the architecture from the viewpoint of the virtual
water
The water supply use in Japan decreases by popularization of the water conservation
type equipment, etc. In the other, the use of the bottles water increases. And the
proportion of the virtual water (the water which the foreign country with the import of
cereal and meat consumes) increases.
From this fact, it is necessary to evaluate water usage in the human and architecture in
addition to the water conservation in the plumbing system from the viewpoint of these.
This fact is important in order to catch the appropriate water supply load in the building.
3. The problem of the water environment in global environmental problem.
As a global environmental problem, there are 1. Global warming (rise in the sea water
temperature), 2. acid rain, 3. Marine pollution, 4. Shortages in the fresh water resources,
etc. as a problem of water itself. And, there are 1. Diminishing tropical forests, 2.
130 Environmental pollution of developing country, 3. Desertification, 4. Decreases in the
biodiversity, etc. as relating problem.
The water circles sky, surface and underground in three forms of gas, liquid and solid.
By architecture and infrastructure, which the human produced, the new water cycle
system has been constructed.
All architecture activity with estate development such as consumption of the energy by
the use of fossil fuel and wood use are related on the global environmental problem on
hydrologic cycle.
All architecture activity with estate development such as consumption of the energy by
the use of fossil fuel and wood use are related on the global environmental problem on
water cycle. It is necessary to catch the water problem as not only therefore water
environment field but also problem of the whole architecture.
4. The problem, which should wrestle by cooperating with other field
1. Grasp of various water usages in the architecture and the effect
In the architecture, the water is large used for the cooling of the air-conditioning system.
And the role of the water which is related to air conditioning and heat of the architecture
and city of humidifying, water sprinkling and fog cooling system, etc. increases.
In until now water environment field, investigation and prediction of the water
consumption on the application, which the human actively uses, are made to be the
research object. In the future, it is necessary to grasp various water consumption
included except for these application. And, it is necessary to synthetically clarify effect
on water cycle, water quality and energy consumption of the building.
2. The optimization of hot water supply heat source and the effective utilization of hot
water supply waste heat
In the home of Japan, the proportion of CO2 discharge that the hot water supply
occupies is 13.8% 4). Reduction of the hot water use and improvement in the energy
efficiency are effective for the reduction of the CO2 discharge.
There are the case in which it is handled in the water environment field and case in
which it is handled in thermal environment and building equipment field on the hot
water supply. However, load pattern that the human actively uses is important for the
hot water supply. It is possible that this point utilizes the storage of the research in the
water environment field.
131 Utilization of thermal energy got in cogeneration and solar heat panels, etc. and waste
heat management of hot water supply, which becomes an unused heat must be
examined, while it will cooperate with other field in future too.
3. The architectural planning of the kitchen and sanitary space
To decide the arrangement in the building is the kitchen and sanitary space especially
drainage and vent system. Therefore, the relation with the architectural planning field is
very big.
The examination is required on the water usage in the building on the design of facility
and space from the psychological and physiological side, because it occurs by the active
and human activity many.
Then, it seems to be the necessity that the research is tackled by the expansion of
research object of the psychophysiology field in the water environment field by the field
of the mutuality cooperating.
The water usage in the building produces many by the human action. Therefore, the
examination from the psychological side is important for the design of facility and
space.
It is necessary that the field of the mutuality cooperatively tackle the research by the
expansion of research object of the psychophysiology field in the aqueous environment
field.
5. The problem of water environment field concerning other field
In the future, next 3 points are raised within the problem in which the water
environment field in water environment steering committee or relating institute should
wrestle as a result of relating to other field.
1. The examination concerning the water environment on the problem on the security of
the sanitary environment of the building
For example, it is related to the problem of air quality (generation of Legionnaire's
disease) and thermal environment (cooling effect) in the environment where the water
becomes a condition of the aerosol. In the future, the necessity of academically
determining the standard on application and water quality of the rainwater utilization,
which is suitable for it, has been indicated.
In addition to this, the following must be examined concerning air environment and
thermal environment: Available application, effects and risk, etc.
132 2. Reduction of the energy consumption in the plumbing system and examination from
the wide and long-range view
The part as above-mentioned way and contact between other building equipment and
city facility is abounding for the examination on the energy saving.
In the future, the linkage system between air conditioning equipment and electricity
equipment in the BEST must be made to be the thing of which in addition, the accuracy
is high. And, evaluating scale is also expanded in the city infrastructure, and by having
long-range view by the LCA, evaluating things, etc. must examine ideal way of water
environment system in the low carbon society.
3. How rainwater utilization and improvement of the water-familiarization space are
concerned in the relaxation of thermal environment of city (heat island) or building, and
examination on the appropriate form
By cooperating the role which water such as water surface, water sprinkling, wall
cooling which are effective in the heat island relaxation fulfils with fields such as urban
environment and city facility, it must be pursued.
6. Conclusions
The science field is segmented, and building and system as a whole city become
difficult to be seen. In the meantime, it is important to grasp the whole system, because
the water circulates in respect of the earth. Therefore, the activity in which many fields
had the aqueous environment in mind is important. It is important to send information in
the form in which each fields are comprehensible in other field. The proclamation of
"Approach on the formation of the water environment in the building and
circumference" by water environment steering committee is the concrete activity, which
had this fact in mind.
7. References
1. Hiroyuki KOSE (2007). Current knowledge of research situation of the water
environment field and future view of research subject, Annual meeting of AIJ
(Kyushu) environmental engineering department research round-table conference
"Current research and future view of the environmental engineering field, traced
way and way to be advanced" (pp.37-40).
2. Kanako SHINMYO et al. (2003): Proposal for a method of performance estimation
of water environment on house (Part 3 The bathroom), Annual meeting of AIJ (D-1,
40211, pp.437-438)
133 3. Hiroyuki KOSE et al. (2008): Outline of Rainwater Utilization Program,
Development of an Integrated Energy Simulation Tool for Buildings and MEP
Systems, the BEST (Part 38), Annual meeting of SHASE (II, OS-26, pp.1153-1156).
4. The national prevention from global warming activity promotion center: The carbon
dioxide emission from the home per household, breakdown according to the
application (2007). http://www.jccca.org/content/view/1048/789/
8. Presentation of Author
Hiroyuki Kose is the Associate Professor at Toyo University,
Faculty of Information sciences and arts from 2009. Special
fields of study are plumbing engineering, water environment
and environment enhancement. At present, I am a leader of
water environment steering committee of AIJ (Architectural
Institute of Japan).
134 Session III: Hot Water Systems
III.1
Study on the System Efficiency of The Latest
Hot Water Supply System under Practical
Use Condition
Masaharu Itagaki
m-itagak@tokyo-gas.co.jp
Tokyo Gas Co., Ltd, Japan
Chiho Mito
Tokyo Gas Co., Ltd, Japan
Saburo Murakawa,
Graduate School of Engineering, Hiroshima Univ. Hiroshima, Japan
Abstract
The use of hot water in Japanese residence is unique in the traditional bath use. An
average daily use of hot water is 450 liters, the total of filling the bath tub for soaking
with 200 liters and the shower use with 100 ~ 150 liters.
The paper presents the experimental study on the actual efficiency of the latest hot water
supply systems, the condensing boiler with unit heating efficiency of 95% and the latest
Heat-Pump Hot Water System with CO2 refrigerant. The latter uses the heat-pump
technology with the unit COP of 5.20 ~ 3.00, typically equipped with the storage tank of
370 liters. The degrading of the efficiency should be resulted from three factors. The
first factor is the heating loss from the storage tank. The second factor is the low heatpump unit efficiency effected by the residual warm water in the storage tank. The third
factor is the heating loss from piping system. The laboratory testing shows that these
three factors compromise the heat-pump hot water system efficiency by 40%. This
135 indicates the importance of system operation learning function in order to minimize the
residual hot water in the tank at the end of the day.
The laboratory testing result was also compared with the field testing in actual use. The
field testing shows 10% lower efficiency compared with the laboratory testing. This
degrading is believed to be produced from the modeled hot water demand in the
laboratory testing and the actual use. An improvement is required in the hot water
modelling to reflect the typical bath re-heating use in the actual use.
Keywords
Energy consumption; Hot Water Demand; Bath Re-heating Use; Practical Use
Condition
1. Introduction
1.1 Energy Demand Growth in Japan
Despite the historical effort in energy efficiency in industrial sectors, the overall energy
demand in Japan has been constantly growing, mainly due to the demand growth in
residential and commercial sectors.
The increase in the electricity use in the residential sector is the primary driver of such
growth. Fig.1 shows that both the growth in the energy use per household and the
increase of number of houses have resulted that the total energy use in residential sector
in 2005 has increased by 230% compared with 1973.
FY1973=100
Personal consumption
Energy consumption
Number of houses
1973
2000
2005 year Fig. 1 Trend of Energy Consumption in Residential Sector in Japan1)
136 Fig.2 shows the residential energy consumption in various countries by end use. It is
noted that the hot water energy consumption is relatively large in Japan, accounting
approximately 30%. Reducing hot water energy use is important in improving Japanese
residential energy efficiency.
Heating
Hot Water
Cooking
Lighting,etc
Cooling
5,945
5,267
US
66,270
France
64,637
18,470
7,710
4,467
8,365
8,474
Germany
18,290
66,491
1,647
7,927
3,871
UK
47,524
19,437
4,027
Japan
11,813
0
15,882
20,000
8,781
1,340
11,830
40,000
(MJ/household・year)
60,000
80,000
100,000
120,000
Fig. 2 Energy Consumption in Residential Sector in Various Countries 2)
1.2 Japanese Building/Housing Energy Efficiency Regulation 2009
The regulation of energy efficiency in the residential sector has been developed in US,
Europe(EU) and Japan. Table 1 shows the regulations/codes for energy efficiency
improvement, and it can be noted that new regulations/codes have been recently
implemented in Building/Housing, with the measures of total energy consumption
regulation both in US and EU.
137 Table 1 Energy Efficiency Improvement Regulation
Appliance
Building/Housing
International
Energy  Minimum Energy Performance
Standards(1987-)
Conservation Code), MEC(Model Energy
Code1995), ASHRAE(American Society of  ENER Guide Labeling(1987-)
Heating, Refrigerating and Air-Conditioning
Engineers, Inc.)
 IECC(2006
US
 ENERGY STAR
-total energy in house
 Energy Performance of Buildings  Energy efficiency regulation(EU
Directive(EU directive) (2003.1-)
directive) (1992-)
 Energie Pass(2006.1-)
 Energy Labeling(EU directive)
EU
-total energy,CO2 emission in house
(1992-)
-Energy Star Program
9grades[A~I]:standard:250kWh/m2/Y
-EU eco label
 Top-Runner
System:
 Energy Saving Act(1979- )
Appliances(1998-)
-insulation level standard(1980-)
System:
Appliances
- total energy limit (2009.4- * Applicable  Labeling
Japan
(2000-)
to Detached House only)
 Labeling System(2009.6-)
After the recognition of the need of direct regulation on building/housing energy
efficiency improvement, the Ministry of Land, Infrastructure, Transport and Tourism
(MLIT) has implemented the regulation of energy efficiency level for a new detached
house in Spring 2009.
The new regulation does not simply specify the level of insulation level, but it regulates
the modeled energy consumption reduced by 10% compared with a typical new
detached house, as the same measures of US and EU. The modeled energy consumption
is the total of space-heating, cooling, ventilation, lighting and hot water supply.
In order to achieve 10% reduction of the total energy, high efficient appliances have to
be implemented, such as latest energy efficiency air-condition, compact fluorescent
lights and condensing type boilers. Regarding hot water supply, a new CO2 refrigerant
heat pump system has become popular in Japan, followed by new residential CHP
(combined heat and power) systems using a compact engine or a fuel cell technology
have attracted the market. Solar Energy use is also recommended, such as Photovoltaics
(PV) in generating electricity and solar thermal system in hot water supply.
2. Hot Water Use in Japanese Residences
2.1 Hot Water Usage Model in Actual Use Condition
Upon the validation of the observance of new regulation, the appliances energy ratings
are the primary information. It is widely acknowledged that the energy ratings by
manufactures are usually validated at test conditions with relatively continuous use,
138 while the actual in-use energy ratings will be degraded. The new MLIT regulation uses
energy ratings under modeled usage load variations, not the manufactures energy ratings
at constant use condition.
2.2 Hot Water Usage Survey Study
In order to develop a hot water usage model, hot water usage in Japanese residences
should be thoroughly studied.
Fig.3 shows the typical hot water usages and their proportions.
Bath Soaking
140
0
50
100
Shower
138
150
Kitchen
106
200
250
300
Lavatory
58
350
400
450
500 (Liter) Fig.3 Average Daily Load of Hot Water Demand in Japan
In order to validate the energy ratings under actual in-use condition, the actual hot water
demand has been modeled from site surveys.
Fig.4 shows a sample date of the survey of actual hot water demands for an average of
surveyed detached houses. Table 2 shows the hot water model (Model M1) based on the
actual hot water demand survey. The usage model (Model M1) consists from 11 days
for weekday Large Demand, 11 days for weekday Small Demand, 2 days for weekend
Large Demand, 3 days for weekend Small Demand and 2 days for weekend-going out
Demand. Such models have been developed for Winter Model, Summer Model and
Spring/Fall Model.
湯量40℃換算
[L/day] 1400
Hot water
湯量40℃換算
消費量[L/ 日・
戸]
1200
1000
Ave. in 月平均
Ave. + stand.Dev.
月平均-月標準偏差
Ave. ‐ stand.Dev. 月平均+月標準偏差
[L/] 800
600
400
200
0
Nov. 98/
11
Jan. 99/
01
Mar
99/
03
May
99/
05
Jul.07
99/
Sep.
99/
09
Nov. 99/
11
Jan. 00/
01
Fig.4 Variation of Hot Water Demand(Liter-40 /Day) 4)
139 Mar
00/
03
Table2 Hot Water Demand Load: Model M1
Per
month
Kitchen
Bath
Shower
Lavatory
Total
Weekday(Large)
11days
120
150
140
60
470
Weekday(Small)
11days
100
150
80
50
380
Weekend (Large)
2days
160
150
140
100
550
Weekend (Small)
4days
200
150
200
100
650
Weekend going-out(Large)
1day
10
-
200
30
240
Weekend going-out(Small)
1day
10
150
200
20
380
[unit. Liter-40 ]
2.3 Bath Soaking and Re-heating
It can be noted that “Bath Soaking” usage is unique in Japan, and it accounts
approximately 32% of hot water usage as seen in Fig.3. “Bath Soaking” has been a
unique traditional bath culture in Japan. After washing and shower use, people soak
themselves in a bath tub filled with hot water (Fig.5). This “Bath Soaking” usage is
regarded as the most essential part of a bath use in Japan. This unique culture requires
three characteristics. It requires a wide and deep bath tab to allow fully soaking in the
hot water, enough hot water, estimated as mush as 150 liter to 200 liter and “reheating”.
The “re-heating” is required in order to heat up the hot water in the bath tub during the
“Bath Soaking” time. The heat loss from the bath surface and lower body temperature
will result in the decrease of water temperature in the bath tub. The initial hot water
temperature in the bath tab is approximately 40 degrees Celsius (varying with people’s
preferences) and the temperature will decrease to 38 degrees during the bath use. The
decrease of the temperature may result in the discomfort in the bath soaking, requiring a
re-heating. Fig.6 shows the survey of the use of re-heating during the bath soaking.
More than 90% people use “re-heating”.
140 Fig. 5 Bath Soaking
No Use
4%
Unknown Yes with manual op. Yes with Auto 95% (in winter) Fig. 6 Survey on the Use of Re-Heating
3)
3. The Study on Energy Ratings in the Latest Hot Water Supply
Systems in Actual Use
3.1 Condensing Type Boilers
3.1.1 Condensing Type Boiler Technology
The MLIT new regulation does not require the use of condensing type boilers, but the
tight efficient regulation may not be observed without using condensing type boilers.
141 Fig.7 shows the typical condensing type boiler’s schematic. It recovers heat from
discharged high temperatures exhaust gas in the pre-heating process, allowing the less
use of energy in supplying the same amount of hot water. The energy rating under the
continuous usage becomes 95%, compared with 80% by a conventional type boiler.
By recovering the temperature from exhaust gas, the temperature of discharged gas will
decreased to 50-80 degrees Celsius from 200 degrees. Upon the recovery of such
temperature, condensate water is produced during operation and the pH neutralizer and
the drain pipe is required to expel condensate into sewage. On the other hand, there are
benefits of allowing the use of low-temperature exhaust piping without insulation or
chimney requirements.
exhaust gas(50~80℃)
neutralizer
air
water
gas hot drain Fig.7 Schematic of Condensing Type Boiler
3.1.2 Condensing Type Boiler Energy Ratings under Actual Use Condition
In order to validate the energy ratings under actual in-use condition, the modeled hot
water load (Model M1) was used.
Fig.8 shows the comparison of the Energy Ratings under constant use by manufactures
and the Energy Ratings under the modeled demands. The Energy Ratings under the
modeled demands are 7-9% less than that of constant condition. There is an obvious
explanation of the energy rating under the modeled demands. They are the remained hot
water in the piping system and the heating loss absorbed by the boiler body. When the
duration time of the hot water is short by turning on and off continuously, the heating
loss becomes larger.
142 Fig.9 shows the efficiency variation along the hot water usage time-duration and timeinterval, and the graph indicates that the frequent turning on/off use will result in the
less efficiency. Continuous hot water use (5 min.) shows better efficiency than short
time use (10 seconds) with frequent on/off. The interval time also affects the efficiency.
The longer interval may result in the heat loss from remained piping system, thus the
efficiency will be degraded. Although the testing shows such degrading due to frequent
on/off actions, turning on and off of hot water use can be regularly seen in the use of
kitchen and lavatories. Fig.10 shows the single-lever basin faucet, which controls the
water volume and the temperature by a single lever. A seasonal optimization of the
temperature may help the efficiency improvement, by altering the neutral position to use
cold water only in Summer.
Efficiency(%) in actual use condition
100
95
90
Condensing
Boiler
85
80
75
Conventional
Boiler
70
70
75
80
85
90
Efficiency(%) in constant use
95
100
Fig.8 Efficiency of Boilers under Practical Use Condition
143 Efficiency[%]
100
Unit Efficiency=95%
90
80
70
60
50
Interval of hot water demand
40
1mi
30
Hot water
Supply Duration
10S
3mi
1min
10mi
5min
5L/min
Fig.9 Efficiency change by hot water supply time-duration and frequency5)
Fig.10 Example of Single-Lever Faucet
3.2 CO2 Refrigerant Heat Pump Hot Water Supply System
3.2.1 CO2 Refrigerant Heat Pump Hot Water Supply System Technology
A new hot water system using CO2 refrigerant heat pump (Fig.11) has been widely
utilized in many Japanese houses. The new heat pump hot water system uses the heatpump technology to produce hot water as warm as 60-85 degree Celsius. Although the
efficiency of heat-pump technology is relatively high, the instantaneous power of
producing hot water is very limited, requiring a storage tank to store hot water
production during the night. The efficiency of heat pump becomes higher when the
system inlet water temperature is low. Inside the storage tank, the produced hot water is
usually stored in the upper layer of the tank and the colder water in the lower layer of
144 the tank is usually heated up by the system. Despite the higher efficiency under the
constant use condition, there are a few drawbacks in the heat pump systems.
1) Lack of instantaneous heating power: It is as small as 2.6 liter/min, only one-fourth
of a regular shower supply volume of 10 liter-40 degrees per minute. This typically
causes a problem that the lack of the stored hot water results in immediate short of hot
water, preventing the use of shower or bathing until the hot water storage becomes
sufficient to use.
2) Degrading of the efficiency: During the use of hot water, the upper layer of hot water
and the lower cold water will create a mixed layer and the heat-pump efficiency would
be degraded.
In both cases, the heat pump hot water supply system requires an intelligent learning
function in order to optimize the storage volume, no less than the level of shortage in
use and not leaving too much unused hot water in the storage tank.
hot water
60~85℃
Storage
Heat
t
k
Bath re-
exchanger
Heat
f
pump
heating
b th
water Fig.11 Schematic Drawing of Heat pump Hot Water System
3.2.2 CO2 Heat Pump System Energy Ratings under Actual Use Condition
Fig.12 shows a schematic drawing of three degrading of the efficiency in the heat pump
system. They are:
- Heat Loss from piping system
- Heat Loss from Storage Tank
- Degrading of heat pump efficiency by residual warm water in the tank
The last factor is inevitable in the fundamental technology of heat pump. Heat Pump
technology uses the ambient thermal recovery by the refrigerant and the efficiency
145 becomes higher when the temperature difference becomes larger. In the heat pump
system, the system efficiency becomes higher when the temperature difference between
the system outlet hot water (60-85 degrees Celsius) and the inlet water becomes larger.
In case the unused warm water is left in the tank, the efficiency will be degraded.
As indicated in the previous chapter, the heat pump system has a dilemma of two
opposites:
(A) Better having enough hot water in the tank in order to provide enough hot water
under unexpected use, because the system has very limited instantaneous supply
power.
(B) Energy Ratings may be degraded if there is too much residual hot water in the tank,
because of the technology of heat pump.
heat loss
the low heat-pump unit
efficiency effected by
the residual warm
water in the storage
heat loss
residual
warm
water
hot water
water
Fig.12 Three Factors Resulting the Degrading of Heat-Pump System Efficiency
146 Fig.13 shows the system puts priority on the efficiency of the system, rather than the
convenience of availability of hot water, by Manufacture C. In the Fig.13 of Day 7, the
minimum level of remained water is almost zero, which indicates the possible
insufficient hot water supply. The maximum hot water storage is controlled as much as
80MJ. On the other hand, Fig.14 shows the different approach by Manufacture D. It
puts priority on the convenience of hot water availability rather than the system
efficiency. The maximum hot water storage level is controlled as much as 100MJ, and
the minimum storage level is controlled as less as 20MJ, leaving enough volume in the
unexpected event.
[MJ]
Daily MAX Storage Based on
water temp.
th
th
th
Based on 40℃
th
Based on
50℃
th
Based on 60℃ Daily MIN Storage th
Fig.13 Variation of Stored Volume (Manufacture C)6)
[MJ
Daily MAX Storage Based on
water temp.
th
th
th
Based on
40℃
th
Based on
50℃ th
Based on 60℃ Daily MIN Storage
th
Fig.14 Variation of Stored Volume (Manufacture D)6)
3.2.3 Degrading of Efficiency by “Re-Heating”
The degrading of the efficiency by warm water inlet may be regarded as the primary
challenge for heat pump hot water supply system. The second problem should be
carefully studied in the efficiency degrading by “re-heating” function. As indicated in
147 the Chapter 2-3, “re-heating” of hot water in the bath tub is one of the essential
functions in the Japanese bath system. In order to provide enough thermal energy for reheating, the heat pump hot water needs to produce higher temperature storage water in
the tank, as high as 80 degrees Celsius, compared with 60 degrees in normal hot water
supply. 80 degrees hot water is used for heat exchange in the re-heating process.
However, this higher temperature hot water production will sacrifice the efficiency.
More heat loss is expected from storage tank, due to higher temperature and more
unused hot water. Unused hot water may also degrade the heat pump efficiency. Fig. 15
shows the re-heating function measurement in the laboratory testing. During Day 1 to
Day 4, re-heating load is applied to the test and the higher temperature hot water is
produced. The system learns that the site needs warmer temperature for re-heating as
high as 85 degrees Celsius during the first 4 days. After Day5, the system will keep
supplying 85 degrees Celsius hot water until Day18, even if the test gives no re-heating
loads after Day 5. Thus, there are more heat loss from remained 85 degrees hot water
and the efficiency will be largely affected by re-heating function and learning logic.
In order to reflect the practical efficiency in actual use in the validation of
Building/Housing Efficiency Regulation, it will be necessary to implement the reheating load in the modeled heat load.
Temp. from HP‐unit
Temp. into HP‐unit
7th 14th
21th System COP
bath re‐heating use 7th 14th
21th Fig.15 Laboratory Testing on Re-Heating in Heat Pump System8)
148 3. Conclusion
- The study shows the Energy Ratings under the modeled demands (Model M1) are 79% less than those in constant use condition.
- The efficiency of instantaneous boilers is affected by the hot water supply duration
time and frequency of turning on/off. The unused hot water in the piping system and the
heating loss absorbed by the boiler body may result in the degrading of the efficiency.
- There are three factors causing the degrading of efficiency in the heat pump hot water
supply system. They are “Heat Loss from piping system”, “Heat Loss from Storage
Tank” and “Degrading of heat pump efficiency by residual warm water in the tank”.
- The heat pump system efficiency in-actual use may be degraded than that of laboratory
testing under the modeled hot water load (Model M1). This is believed to be associated
with further degrading by “re-heating” in the actual use.
- Because Re-heating is the traditional and common bath use in Japan, implementing
“re-heating” load into modeled hot water load should be considered to reflect the
practical efficiency.
4. Reference
1) White Paper in Energy 2007
2) 4th Meeting of Energy Saving (2001),“Data1 The present situation of energysaving measures in various countries”
3) Mitsubishi Research Institute,Inc. ,“Report of Investigation for Practical Use of
Gas and Kerosene Appliances(2008-9)”
4) Masayuki Mae, et al. (2009),“Practical Use and Efficiently Usage of Boilers”,
Blue & Green Project Seminar 2009
5) Masayuki Mae, et al., “Study on the Standardization of Methods to Measure
Efficiency of Domestic Gas and Oil Hot Water Heaters by the Using Mode (Part1)
Average and seasonal, daily, time fluctuation of hot water consumption.”; Technical
Papers of Annual Meeting The Society of Heating, Air Conditioning and Sanitary
Engineers of Japan,(2006.9)
6) Yasunori Wagatsuma, et al., “Study on Performance Evaluation of Domestic Hot
Water Supply System using Storage Tank(Part 6) ”; Proceeding of the 28th Annual
Meeting of Japan Society of Energy and Resources,(2009.6)
7) Hiroki Kitayama, et al., “A study on the evaluation of running condition for hot
water storage tank systems in houses (Part11) Performance of CO2 heat pump water
heater in winter”; Proceeding of the 28th Annual Meeting of Japan Society of Energy
and Resources,(2009.6)
8) Yasuhiro Hamada, et al., “Study on Performance Evaluation of Domestic Hot
Water Supply System using Storage Tank(Part 7) ”; Proceeding of the 28th Annual
Meeting of Japan Society of Energy and Resources,(2009.6)
149 5. Presentation of Author
Masaharu Itagaki has been working on the gas appliances
efficiency nearly two decades since he joined Tokyo Gas in 1990.
He is now deeply involved in the development of national
building/housing efficiency codes in the fields of hot water supply,
air conditioning and residential CHP (Combined Heat and Power)
application.
150 III.2
Calculation Method for Loads of Hot Water
Demand with the Hot Water Storage Tank
System in Houses
( Part 2 ) Modeling of the loads of reheating
bathwater and experimental evaluation of CO2 heat
pump water heater
Hiroshi Takata (1), Saburo Murakawa (2), Akiko Takaaze (3)
Hiroki Kitayama (4), Yasuhiro Hamada (5), Minako Nabeshima (6)
(1) takatah@hiroshima-u.ac.jp
(2) muraka@hiroshima-u.ac.jp
(3) takaaze1127@azusasekkei.co.jp (4) kitayama@ip.kyusan-u.ac.jp
(5) hamada@eng.hokudai.ac.jp
(6) nabeshima@urban.eng.osaka-cu.ac.jp
(1) Graduate School of Education, Hiroshima Univ.
(2) Graduate School of Engineering, Hiroshima Univ.
(3) Azusa Sekkei Co., Ltd.
(4) Faculty of Engineering, Kyushu Sangyo Univ.
(5) Graduate School of Engineering, Hokkaido Univ.
(6) Graduate School of Engineering, Osaka City Univ.
Abstract
The purpose of this study is to suggest the calculation method for loads of hot water
consumption on the basis of the hot water usage with the hot water storage tank system
in houses. The authors presented a paper to the 34th International Symposium of
CIB/W62, Hong Kong (China). In the previous paper, the calculation models which are
divided into some levels for hot water usage were prepared.
151 In this paper, based on the results of the measurements, the behavior of hot water usage
and the loads of hot water consumption were analyzed. Especially, as for the loads of
reheating bathwater, the relationships between loads and intervals of reheating were
shown. And, the authors made a calculation method with calculation process of
reheating bathwater. In addition, the loads of hot water consumption for family size of
four people were calculated under the condition of “Medium level” which uses a
medium amount of hot water. According to the calculation results, two modes of hot
water loads for fourteen days were suggested. Finally, the authors carried out
experimental evaluation for CO2 heat pump water heater by using the mode of hot water
loads.
Keywords
Hot Water Supply Demands, Storage Tank, CO2 Heat Pump,
Simulation, Experiment, Houses
1. Introduction
As for the CO2 heat pump water heater, the measurement of hot water consumption and
actual condition of this system were carried out in 14 houses. We analyzed the
fluctuation and tendency for the loads of hot water consumption in each house.
Additionally, we analyzed the dweller’s behavior for hot water usage on the basis of the
measurement data and the questionnaire data. We suggested the calculation method for
the loads of hot water consumption in households which had various styles on hot water
usage [1].
In this paper, we focused the function of reheating bathwater and developed a new
calculation model with the reheating bathwater. In addition, according to the results of
calculation, two modes of hot water loads for fourteen days were suggested, as “Mode
of manual keeping warm” and “Mode of adding hot water” for bathwater. Finally, we
carried out experimental evaluation to clarify the effect of water usage on the
performance of CO2 heat pump water heater by using the two modes of hot water loads.
2. Outline of the investigation
The measurements of hot water consumption in houses with the hot water storage tank
were carried out in 14 houses located at the regions of different climates in Japan.
Figure 1 shows the distribution diagram of hot water supply system and the
measurement points. We measured mainly cold and hot water temperatures, flow rates
and consumption of electricity at each point in the system. These points were the same
in all houses except the surface temperatures of hot water storage tank depended on the
tank capacity. These values were recorded automatically every two or three seconds
through a year. Hot water was supplied to a bathtub and other uses by each pipeline. In
addition, this system had the function to reheat water in the bathtub by the circuit of
152 pipe putting the heat exchanger inside the storage tank. The measurement in each house
was carried out from December 2005 to February 2007. The period of analysis was
February 2006 as winter season, May and November 2006 as middle season, and July
and August 2006 as summer season.
Hot water storage tank
T7 Air temperature
T8 ( Top )
T9 ( 50L )
T10 ( 100L )
H1 Air humidity
T18
T19
T20
Hot water
supply to
bathtub
T11 ( 150L )
T12 ( 200L )
T13 ( 250L )
T14 ( 300L )
T6
Heat Pump
Hot water supply
F3 T3
Bathtub
T16 ( 350L )
T17 ( 400L )
T15 ( Bottom )
F4 T5
F2 T2
Circulation to
reheat
T4
F1 T1
Cold water supply
Electricity supply
E2
E1
Humidity
Temperature
Flow
Electricity
Figure 1 – Distribution diagram of hot water supply system and measurement
3. Classification of heat loads for reheating bathwater
Hot water
consumption
[L/min]
30
20
1500
Hot water usage
Filling a bathtub
10
0
16:00
1000
500
18:00
20:00
a) Automatic keeping warm
153 Heat loads for reheating
0
22:00
Heat loads for
reheating [kJ/min]
Heat loads for reheating bathwater were classified into three types, “Automatic keeping
warm”, “Manual keeping warm” and “Reheating”. Figure 2 shows the classification of
heat loads for reheating bathwater. Heat loads for reheating bathwater at regular
intervals, like Figure 2 a), were regarded as “Automatic keeping warm”. These loads
occurred after filling a bathtub with hot water by the function of CO2 heat pump water
heater. Heat loads for reheating bathwater which occurred in same time-zone with hot
water usage, like Figure 2 b), were regarded as “Manual keeping warm”. These loads
occurred by the dweller’s operation for keeping warm of bathwater. Additionally, a
relatively-large heat load for reheating bathwater with no loads of filling a bathtub, like
Figure 2 c), was regarded as “Reheating”.
1500
Hot water usage
20
Filling a bathtub
1000
Heat loads for reheating
10
500
0
18:00
20:00
0
0:00
22:00
Hot water
consumption
[L/min]
30
20
1500
Hot water usage
Filling a bathtub
Heat loads for reheating
1000
10
0
16:00
c) Reheating
500
18:00
20:00
0
22:00
Heat loads for
reheating [kJ/min]
Hot water
consumption
[L/min]
30
Heat loads for
reheating [kJ/min]
b) Manual keeping warm
Figure 2 - Classification of heat loads for reheating bathwater
Table 1 shows the statistics of the classified heat loads for reheating bathwater. As a
general trend, according to the outdoor air temperature rose from winter to summer, the
loads and duration times decreased. Because bathwater’s temperature did not decrease
in summer than in winter, bathwater’s temperatures were near to the preset temperatures
of automatic keeping warm or temperature of usage in summer. The loads of
“Automatic keeping warm” and “Manual keeping warm” were obviously smaller than
that of “Reheating”. And, the loads of “Manual keeping warm” were twice larger than
the loads of “Automatic keeping warm”. Duration times of “Reheating” were from 18
minute to 26 minute, the seasonal difference was large.
Table 1 - Statistics of the classified heat loads for reheating bathwater
Average
S. D.
Maximum
Average
Duration time per frequency
S. D.
[min/frequency]
Maximum
Average
Loads per 1 minute
S. D.
[kJ/min]
Maximum
Loads per frequency
[MJ/frequency]
Automatic keeping warm
Winter
Middle Summer
0.98
0.73
0.61
0.62
0.48
0.45
3.17
3.56
2.45
4.4
3.8
3.2
3.0
2.6
2.0
13
18
10
233.3
206.5
188.2
112.6
94.7
85.6
612.0
526.5
455.5
154 Manual keeping warm
Winter
Middle Summer
1.85
1.63
1.25
1.32
0.93
0.67
6.95
6.11
2.91
6.1
6.1
4.8
4.8
3.6
2.5
37
23
15
309.9
274.0
256.4
105.3
84.2
64.2
682.8
720.9
431.0
Winter
12.20
4.29
31.04
26.7
18.2
131
524.7
143.4
864.1
Reheating
Middle Summer
8.54
4.69
3.12
1.88
24.81
14.02
24.0
18.7
13.7
9.1
98
58
394.8
275.0
99.4
83.9
598.2
495.2
4. Modeling of heat loads for reheating bathwater
Decrease of bathwater’s temperature, which is occurred by the change of outdoor air
temperature and the difference of bathroom’s structure and condition, affected the
amount of heat loads for reheating bathwater. In this chapter, the calculation models of
heat loads for reheating bathwater, as for “Manual keeping warm” and “Reheating”, are
made from the relationship between bathwater’s temperature and heat loads for
reheating bathwater.
In order to study the decrease of bathwater’s temperature, “Reheating interval” was
defined as interval time between the end of reheating bathwater and the start of the next
occurrence for reheating bathwater. Figure 3 shows the relationship between “Reheating
interval” and decreased water temperature in each season. The measurement data were
some varied, because the climate characteristic and the heat insulation capacity of
bathtub were different in each measurement house.
Decreased water temperature
[ºC]
0
0
-1
-2
-3
-4
-5
-6
-7
-8
-9
-10
60
Winter
Middle
Summer
Reheating interval [min]
120
180
240
300
360
Winter:y=-0.1291x0.6271
Middle:y=-0.1028x0.6267
Summer:y=-0.0264x0.8113
Figure 3 - Relationship between “Reheating interval” and decreased water
temperature
Figure 4 shows the relationship between increased water temperature by “Manual
keeping warm” and heat loads for reheating bathwater in each season. There was a
proportional relationship between the two. Slopes of linear approximate equations in
each season were nearly equal. Additionally, increased water temperature and loads
became small from winter to summer.
155 Loads per frequency
[kJ/frequency]
8000
7000
6000
5000
4000
3000
2000
1000
0
Winter
Middle
Summer
Winter:y=1038.5x R2=0.81
Middle:y=1058.8x R2=0.59
Summer:y=1022.6x R2=0.66
0
1
2
3
4
5
6
7
8
Increased water temperature [ºC]
9
10
Figure 4 - Relationship between increased water temperature by “Manual keeping
warm” and heat loads for reheating bathwater
8000
7000
6000
5000
4000
3000
2000
1000
0
Calculation value (+0.5ºC)
0.6271
y=134.1x
+519.3
Loads per frequency
[kJ/frequency]
Loads per frequency
[kJ/frequency]
The relational expressions between “Reheating interval” and heat loads for reheating
bathwater were calculated by using the linear approximate equations of Figure 3 and
Figure 4. Figure 5 shows, with measurement data, the relationship between “Reheating
interval” and heat loads for reheating bathwater in each season. In this Figure, “(±0 ºC)”
means relational expression when the water temperature is increased by reheating as
much as decreased water temperature in “Reheating interval”. “(+0.5 ºC)” and “(+1 ºC)”
mean relational expressions when the water temperature are increased respectively
“Decreased water temperature +0.5 ºC” and “Decreased water temperature +1 ºC”. In
winter, the relational expression of (+0.5 ºC) was closely to the measurement values. In
this paper, heat loads of reheating in winter were calculated by using the relational
expression of (+0.5 ºC).
(+1ºC)
(+0.5ºC)
(±0ºC)
0
60
120
180
240
300
Reheating interval [min]
8000
7000
6000
5000
4000
3000
2000
1000
0
Calculation value (+0.5ºC)
0.6267
+529.4
y=108.8x
0
360
60
Figure 5 - Relationship between
“Reheating interval” and
heat loads for reheating bathwater
120
180
240
300
Reheating interval [min]
8000
7000
6000
5000
4000
3000
2000
1000
0
Calculation value (+0.5ºC)
0.8118
+511.3
y=27.0x
0
60
156 (+1ºC)
(+0.5ºC)
(±0ºC)
120
180
240
300
Reheating interval [min]
c) Summer
360
b) Middle
Loads per frequency
[kJ/frequency]
a) Winter
(+1ºC)
(+0.5ºC)
(±0ºC)
360
As for the heat loads for “Reheating”, which is shown in Figure 2 c), heat loads were
decayed with the duration time. In each occurrence of “Reheating”, the ratios of loads to
the instantaneous maximum loads were calculated. Figure 6 shows the average decay
curve of the ratios of loads to the instantaneous maximum loads in each season. This
decay was occurred by the difference between bathwater’s temperature and hot water
temperature in storage tank, the trends of decay were closely to each other.
Ratio of loads to
the instantaneous
maximum load
100%
Winter
Middle
Summer
80%
60%
40%
20%
0%
0
10
20
30
40
Duration time [min]
50
60
Figure 6 - Average decay curve of heat loads for “Reheating”
5. Calculation of hot water consumption loads
We aim to calculate the hot water consumption for simulation of the running condition
and the efficiency of the CO2 heat pump water heater. As a time series simulation, it is
possible to apply the calculation method with the Monte Carlo Simulation technique,
which had presented in previous paper [1, 2, 3].
We generate the pseudo-random numbers by using personal computer. The generated
random numbers are applied to calculate the occurrence time interval of water usage,
the duration time and the flow rate. The simulation is carried out with repetition until we
can get the stability for the calculating results. In this paper, the number of trials is one
hundred, and the time interval of simulation is one minutes.
Table 2 shows the calculation model with new model of heat loads for reheating
bathwater. Hot water consumption were considered as 40 ºC hot water consumption on
each usage, but “Reheating” and “Manual keeping warm” were models converted to the
value of heat loads.
The calculation of hot water consumption loads in winter was carried out. Table 3
shows the model patterns of bathing styles from the high possibility of occurrence on
the basis of the measurement data. Case-1 has the filling a bathtub everyday. Case-2 has
alternating styles of filling a bathtub and reheating every two days. Case-3 has
alternating styles of filling a bathtub and taking a shower every two days. Case-4 has
filling a bathtub in four days and not taking a bath in three days per week.
Table 4 shows the levels of hot water demands in the household, in a family size of four
people. The levels were set up by assuming three levels, “Large model”, “Medium
model” and “Small model”, depending on the amount of hot water consumption.
“Medium model” was a standard household which consumed about 450 [L/d] (40 ºC hot
water consumption), “Large model” and “Small model” were assumed the hot water
consumption per day as 600 [L/d] and 300 [L/d] respectively.
157 Figure 7 shows the ratio of frequency of hot water usage in winter. “Manual keeping
warm” occurs at the same time-zone of “Usage for bathing”.
Table 2 - Calculation model for houses
Usage for bathing
Level
Frequency of usage per day
[frequency / house/d]
Winter
Hot water consumption
per frequency
[L/frequency]
Middle
Summer
Winter
Flow rate
[L/min]
Middle
Summer
1
2
Shower
3
4
5
1
2
Number of persons per household
Average
Distribution*1
Average
Distribution
Average
Distribution
Average
Distribution
Average
Distribution
Average
Distribution
35.6
Erl.K=02
28.6
Erl.K=03
30.6
Erl.K=03
54.3
Erl.K=03
49.6
Erl.K=02
56.4
Erl.K=10
80.8
Erl.K=03
73.4
Erl.K=05
81.2
Erl.K=08
5.6
Erl.K=04
Filling a bathtub
3
Number of persons per household
115.7
Erl.K=05
121.6
Erl.K=10
113.1
Erl.K=05
177.3
Erl.K=30
27.1
Erl.K=09
36.1
Erl.K=04
44.2
Erl.K=05
43.6
Erl.K=06
5.0
Erl.K=05
5.5
Erl.K=05
57.1
Erl.K=05
65.2
Erl.K=03
65.5
Erl.K=15
7.4
Erl.K=10
7.5
Erl.K=10
1
2
1
*2
Erl.K=100
*2
Erl.K=100
*2
Erl.K=100
15.0
Erl.K=50
13.7
Erl.K=50
13.2
Erl.K=50
107.3
Erl.K=05
101.1
Erl.K=05
108.3
Erl.K=05
4.8
Erl.K=05
Reheating
Kitchen etc.
Level
-
4
3
Frequency of usage per day
[frequency / house/d]
4
130.4
Erl.K=30
Average
21.9
45.9
89.8
Distribution
Erl.K=02
Erl.K=09
Erl.K=20
Frequency of usage
Average
18.9
32.6
67.7
Middle
per day
Distribution
Erl.K=02
Erl.K=03
Erl.K=04
[frequency / house/d]
Average
9.2
18.8
50.4
Summer
Distribution
Hyp.K=100
Erl.K=03
Erl.K=04
1.7
Average
Winter
Hyp.K=02
Distribution
1.6
Flow rate
Average
Middle
[L/min]
Hyp.K=02
Distribution
1.7
Average
Summer
Hyp.K=02
Distribution
1
Duration time per frequency [min/frequency]
Note : *1 Erl. : Erlang distribution, Exp. : Exponential distribution, Hyp. : Hyperexponential distribution, K : Phase
*2 Volume of bathtub's water consumption are determined in accordance with the capacity of bathtub.
*3 "x" means the reheating interval [min].
Winter
Winter
Heat load
per frequency
[kJ/frequency]
Middle
Summer
Winter
Loads per 1 minute
[kJ/min]
Middle
Instantaneous
maximum load *[kJ/min]
Summer
1
Average
Distribution
Average
Distribution
Average
Distribution
Average
Distribution
Average
Distribution
Average
Distribution
12743.4
Erl.K=10
8541.9
Erl.K=08
4648.9
Erl.K=06
*
901.5
Erl.K=30
*
697.5
Erl.K=20
*
442.3
Erl.K=15
Manual keeping warm
Number of persons
per household
Calculation value (+0.5ºC)
0.6271
+519.3 *3
y=134.1x
Calculation value (+0.5ºC)
0.6267
+529.4
y=108.8x
Calculation value (+0.5ºC)
0.8118
y=27.0x
+511.3
309.9
Erl.K=08
274.0
Erl.K=08
256.4
Erl.K=08
Table 3 - Model patterns of bathing styles
Patterns of bathing styles
case-1
"Filling a bathtub + Usage for bathing" occur every day.
case-2
"Filling a bathtub + Usage for bathing"
and "Reheating + Usage for bathing" occur alternately once every 2 days.
case-3
"Filling a bathtub + Usage for bathing"
and "Shower" occur alternately once every 2 days.
case-4
"Filling a bathtub + Usage for bathing"
and "Not taking a bath" occur in the proportion of 4 to 3.
Table 4 - Levels of hot water demands in a family of four persons
Usage for bathing
Level
1
case-1
Large
2
2
3
1
1
case-2
2
2
case-3
2
2
case-4
Medium
Shower
2
3
1
2
3
2
1
1
case-2
1
1
1
1
case-3
1
1
1
1
1
2
1
case-1
4
case-2
4
case-3
4
case-4
4
4
Filling a bathtub
〇
200
[L/frequency]
〇
2
2
〇
〇
〇
180
[L/frequency]
〇
1
2
1
〇
〇
〇
160
[L/frequency]
〇
3
1
〇
〇
Note: Numbers in the table mean the number of persons.
158 Kitchen etc.
4
4
case-1
case-4
Small
4
Ratio of frequency [%]
40
Kitchen etc.
30
Filling a bathtub, Reheating
Usage for bathing, Shower
20
10
0
4
6
8
10
12
14
16
18
20
22
0
2
[hour]
Figure 7 - Ratio of hourly frequency of hot water usage
As the simulation results, Figure 8 shows the fluctuation of hourly loads of hot water
consumption as a representative case. The simulation values could recreate the
fluctuation of hourly loads without contradiction of time-zone for hot water usage, as
dwellers took a bath after filling a bathtub with hot water or reheating water. It was
clarified that “Manual keeping warm” occurred at the same time-zone of “Usage for
bathing”, and the decay of heat loads for “Reheating” was correctly reconstructed. As
the results of estimate of hot water consumption per day and heat loads per day, the
average volumes of hot water consumption per day in “Medium model” were 450 [L/d]
in each case. The standard deviations of hot water consumption per day were from 70 to
250[L] by setting the deferent model patterns of bathing styles.
2000
Kitchen etc.
Usage for bathing
10
1500
Filling a bathtub
Manual keeping warm
1000
5
500
0
Heat loads for reheating
[kJ/min]
Hot water consumption
[L/min]
15
0
16
18
20
22
0
[hour]
a) “Filling a bathtub”, “Manual keeping warm”
2000
Kitchen etc.
Usage for bathing
10
1500
Reheating
Manual keeping warm
1000
5
500
0
0
16
18
20
22
0
[hour]
b) “Reheating”, “Manual keeping warm”
Figure 8 - Fluctuation of hourly loads of hot water consumption
159 Heat loads for reheating
[kJ/min]
Hot water consumption
[L/min]
15
7. Mode of hot water consumption loads for experimental evaluation
Heat loads per day
[MJ/d]
As for the mode of hot water consumption loads for experimental evaluation, “Mode of
manual keeping warm” and “Mode of adding hot water” for bathwater were made on
the basis of calculation results.
“Mode of manual keeping warm” was made by extracting contiguous fourteen days
from calculation results of “Medium model, case-1”. On manual keeping warm, the
temperature of bathwater, which volume was 180L, were increased from initial
temperature to 42 ºC depending on the volume of heat loads. On “Mode of adding hot
water”, the temperature of bathwater were increased by adding 60 ºC hot water from a
faucet. Figure 9 shows the heat loads per day on “Mode of manual keeping warm” and
“Mode of adding hot water”, and Table 5 shows the statistics of the mode. The cold
water temperature, achieving temperature of bathwater and volume of bathwater were
set in constant values.
120
100
80
60
40
20
0
120
100
80
60
40
20
0
1
3
5
7
9 11 13
1
Manual keeping warm
Kitchen etc.
Usage for bathing
Filling a bathtub
3
5
7
9 11 13 [day]
Adding hot water
Kitchen etc.
Usage for bathing
Filling a bathtub
a) Mode of manual keeping warm b) Mode of adding hot water
Figure 9 - Heat loads per day on mode of hot water consumption loads
Table 5 – Statistics of mode of hot water consumption loads
Average
S.D.
Maxmum
Minimum
Hot water consumption
per day [L/d]
Mode of manual Mode of adding
keeping warm
hot water
454.1
547.4
79.7
85.5
621.8
729.8
346.4
431.9
160 Heat loads per day
[MJ/d]
Mode of manual Mode of adding
keeping warm
hot water
73.1
84.9
11.9
12.8
98.4
112.0
57.2
67.5
7. Experimental evaluation
7.1 Outline of the experiment
As for the specification of CO2 heat pump water heater, we set up the new system made
by C Company in experimental laboratory. The storage tank capacity is 370L.
Coefficient of performance (COP) of heat pump unit is 4.81 in rated condition and 3.00
in winter season.
The experiment was carried out on the assumption that the climate condition in
experimental laboratory was winter. Accordingly, outdoor air temperature and cold
water temperature were 5 ºC and 8 ºC respectively, and relative humidity was 57%.
The experimental method was simplified on the condition that heat pump COP was not
affected by the difference of method.
7.2 Results of the experiment
As for the experimental results in each mode of hot water consumption loads, the first 7
days were the period of learning for equipment and the last 7 days were the period of
analysis. Table 6 shows the statistics of experimental results in each mode. Figure 10
shows the amount of hot water loads and electric power, and COP as the actual
condition of hot water supply system.
Table 6 - Statistics of experimental results
Mode of manual
keeping warm
Number of days
[day]
Air temperature in laboratory
[ºC]
5.4
5.5
Humidity in laboratory
[%]
56.5
56.5
Cold water temperature
[ºC]
8.8
8.8
Hot water loads
[MJ/d]
74.8
87.3
Amount of heat from heat pump unit
[MJ/d]
102.2
117.5
Amount of heat from heat pump unit in midnight power time-zone
[MJ/d]
93.7
86.5
[%]
90.2
73.6
Amount of electric power
[kWh/d]
11.2
12.3
Amount of electric power for heat pump unit
[kWh/d]
10.8
12.0
Amount of electric power in midnight power time-zone
[kWh/d]
10.0
8.8
[%]
90.2
73.6
Ratio of heat from heat pump unit in midnight power time-zone
Ratio of electric power in midnight power time-zone
Amount of heat storage and heat discharge
7
[MJ/d]
27.1
27.1
Ratio of heat storage and heat discharge
[%]
28.1
23.9
Water supply temperature to heat pump unit
[ºC]
18.6
12.5
Hot water supply temperature from heat pump unit
[ºC]
89.1
88.8
Average volume of heat storage in hot water storage tank
[MJ]
80.3
80.9
Maximum volume of heat storage in hot water storage tank
[MJ]
105.7
103.2
Minimum volume of heat storage in hot water storage tank
[MJ]
27.4
25.4
Heat pump COP
[-]
2.61
2.72
System COP
[-]
1.92
2.17
161 7
Mode of adding
hot water
4
100
3
50
2
0
1
-50
0
Heat pump COP, System
COP
Heat loads, Electric power
[MJ/d]
150
Amount of heat storage and heat discharge
Hot water loads
Amount of electric power for heat pump unit
Amount of electric power for else use
Heat pump COP
System COP
Mode of manual Mode of adding
keeping warm
hot water
Figure 10 - Actual condition of hot water supply system as the experimental results
In “Mode of adding hot water for bathwater”, hot water loads were large. Accordingly,
the heat pump operated beyond the midnight electric power time-zone; when the charge
of electricity is lower than that of daytime. The ratio of electric power in the midnight
power time-zone was low. Ratios of heat storage and heat discharge to amount of heat
from heat pump unit were 28.1% in “Mode of manual keeping warm” and 23.9% in
“Mode of adding hot water for bathwater”. The values of heat pump COP had little
difference between the two modes, but the values of system COP were 1.92 in “Mode of
manual keeping warm” and 2.17 in “Mode of adding hot water for bathwater”. System
COP in “Mode of adding hot water for bathwater”, which has lower ratio of heat
discharge, was higher than that in “Mode of manual keeping warm”. Considering the
experimental results, usage of the function for reheating bathwater has large effect on
the performance of CO2 heat pump water heater.
8. Conclusion
In this paper, we focused on the function of reheating bathwater. From the results of
measurement data, heat loads for reheating bathwater were classified into three types,
“Automatic keeping warm”, “Manual keeping warm” and “Reheating”. Heat loads for
reheating bathwater were modeled by analyzing the relationship between the reheating
interval and heat loads for reheating bathwater. Furthermore, we proposed the new
calculation model with the reheating bathwater, the levels of hot water demands and the
model patterns of bathing styles in the household to apply the calculation of various hot
water usage styles.
In addition, according to the calculation results of hot water consumption loads, two
modes of hot water loads for fourteen days were suggested, as “Mode of manual
keeping warm” and “Mode of adding hot water for bathwater”.
Finally, we carried out experimental evaluation for CO2 heat pump water heater by
using the two modes of hot water loads. As the results, usage of the function for
reheating bathwater has large effect on the performance of CO2 heat pump water heater.
Acknowledgments
This study has been conducted by Center for Better Living. We wish to express
gratitude for the great cooperation of the participants.
162 References
1. H.Takata et al.: Calculation Method for Loads of Hot Water Demand with the Hot
Water Storage Tank System in Houses, Proceedings of the CIB-W62 International
Symposium on Water Supply and Drainage for Buildings (Hong Kong), pp.23-35,
(2008)
2. N.Yamamoto et al.: A study on the loads of hot water consumption in houses with
the hot water storage tank system (Part 1) An analysis of the hot water usage and the
loads of hot water consumption, Proceedings of the CIB-W62 International
Symposium on Water Supply and Drainage for Buildings (Brno), pp.39-47, (2007)
3. H.Takata et al.: A study on the loads of hot water consumption in houses with the
hot water storage tank system (Part 2) Calculation for the loads of hot water
consumption, Proceedings of the CIB-W62 International Symposium on Water
Supply
and
Drainage
for
Buildings
(Brno),
pp.49-60,
(2007)
Presentation of Author
Hiroshi Takata is the Lecturer at Graduate school of education,
Hiroshima University. His special field is the planning of cold and
hot water supply systems.
163 III.3
Hotspot free design, building & installation of
drinking water installations
Ing. O.W.W. Nuijten (1)
o.nuijten@isso.nl
ISSO, Sanitary and Gas installations, The Netherlands,
Rotterdam
Ir. M. Kuijpers (2)
mku@biqstad.nl
biq-stadsontwerp, The Netherlands, Rotterdam
Ir. J. van Wolferen (3)
hans.vanwolferen@mep.tno.nl
TNO-Bouw en ondergrond, The Netherlands, Apeldoorn
Abstract
In The Netherlands we don’t have the possibility to chlorinate the potable water. So we
try to keep the water quality high by other means.
Finest quality of the water delivered by the watercompany at the watermeter
Regular use and/or flushing of the potable water installalation in the building
Control of the cold water temperature
The Buiding Code in The Netherlands requires that the drinking water temperature has
tot stay below 25 oC. This temperature is chosen because bacteria and Legionella grow
faster above this temperature. Another reason is that the user wants fresh and cold water
to come out of a the “cold” water taps.
This requirement has led to rules for:
places where potable water installations can be located (cool places)
minimum distances between watersupply pipes and pipes for central heating.
insulation requirements
ventilation requirements.
164 ISSO and SBR “the knowledge centers for installations and buildings in The
Netherlands” have developed guidelines and concepts for prevention of undesired
heating up of drinking water in new buildings.
This resulted in a technical ISSO-SBR-guideline 811 “Hotspot free design, building and
installation” (the title is translated from dutch)
The practice up to now is, that the installer has to cope with this problem on his own. He
is faced with the following situations:
there is no cool place for drinking water pipes in the floor with floorheating
there are no solutions to prevent crossings with pipes for floor- or
radiatorheating in the floor
there is only one shaft for both drinking water pipes and pipes for central
heating
the room with the waterheater gets to high in temperature
the watermeter cupboard with a unit for districtheating gets to warm
the heat cannot escape form the pipeshaft in a highrisebuilding.
TNO has calculated the minimum distances in the floor between drinking water pipes
and central heating pipes for different temperatures and situations. TNO also developed
software with which one can calculate the temperatures that can exist in a pipeshaft with
cental heating pipes. In ISSO-guideline 30.5 and in the ISSO-SBR-guideline 811 and
are given the solutions based on research of Researchinstitute TNO.
The solutions cannot be solved only bij the installer of the drinking water installations.
In the concept stadium of a project the architect has to design a floorplan which it makes
possible to keep the pipes for central heating and drinking water apart. The architect has
to create cool and warm zones in his building. In a later stage the installer of the central
heating and the installer of the drinking water pipes have to coordinate their pipe layouts to prevent crossings and coming to nearby.
Keywords
Potable water, legionella, hot spots, cold water, unintentional warming, cool zones,
organisation, central heating, floor heating, pipe shaft, lowered ceiling, concrete floor,
finishing floor architect, coordination
1 Introduction
The investigation has three phases:
1. Phase of preliminary study
2. Phase of analysis
3. Phase of writing the report (ISSO-SBR-directive 811
165 2 The research
2.1 The organization of the research
2.1.1 Preliminary study
In the first phase Dutch Standard NEN 1006 “General requirements for water supply
installations” and the guidelines in ISSO-30.5 were studied. Both documents where
taken as the starting point.
2.1.2 Phase of analysis
In this phase the following actions were taken:
-
Interviews with various various installers of potable water systems en central
heauting systems and appartement buildings;
Formulating of thesises (questions) ;
Presenting these thesises to a forum of experts;
Make a report of the reactions of the forum of experts.
2.1.3 Phase of writing ISSO-SBR-directive 811 “Hotspotfree design, building and
installation”
In the third phase a method is worked out to design and install the water supply system
of a family house taking in consideration the possible unintentional warming up of the
potable water by pipes of the central heating system. In addition examples are
developed with practical solutions of how to design en install a potabele water
installation
without
so
called
“hotspots”.
2.2 Results
2.2.1 Preliminary study
Dutch Standard NEN 1006 “General requirements for water supply installations” (2)
gives the following requirements to prevent the unintentional warming of cold potable
water:
-
The temperature of cold water in the distribution pipes has to stay below 25 ºC.
Hot potable water distribution pipes must cool below 25 ºC after they are used.
Distance between water distribution pipes and central heating pipes must be
large enough to prevent unintentional warming of water
166 In ISSO 30.5 “LegionellaCode for installations in familyhouses” (3) solutions are given
for the following rooms and places in a familyhouse to keep the watertemperature
below 25 ºC:
1. rooms with water distribution pipes;
2. meter cupboard;
3. technical rooms;
4. pipe shafts;
5. lowered ceilings;
6. the neighbourhood of warm objects;
7. hidden in floors and walls.
8.
In fig. 2.1 some situations are shown.
Watermeter in cupboard together with Technical room
heatexchanger
Pipeshaft
Hidden in finishing floor
Fig. 2.1 Some situations with possible unintentional warming of potable water
167 Solutions in ISSO 30.5 for the different situations:
1. In rooms with water distribution pipes;
-
Good ventilation
Sun obscuration
No direct radiation on the pipes
Right
place
of
water
supply
pipes
at
a
ceiling
2. In watermeter cupboard with heatexchanger for district heauting
-
Place watermeter low in the cupboard;
Ventilation of the cupboard through low and high opening in the door;
No watersupply pipes above the heatexchanger in the cupboard
3. In technical rooms
-
Keep temperature below 25 ºC bij good insulation of central heating pipes and
sufficient
ventilation
4. In pipeshafts
-
Create separate shafts foir central heating and potable water
If not possible keep temperature below 25 ºC bij very good insulation of central
heating pipes. To calculate the temperature TNO developed for ISSO a
calculating model “Hotspotsim”.
Fig. 2.2 Calculation of temperature in pipe shaft with Hotspotsim
168 5. In lowered ceilings
-
Keep the potable waterlines at the lowest level above a lowered ceiling.
Good insulation of the central heating pipes
Fig. 2.3 Lowered ceiling
6. In the neighbourhood of warm objects
-
Keep the potable waterlines away from war objects like radiators and
condensors at the lowest level above a lowered ceiling.
7. In hidden in floors and walls
In the Netherlands the pipes for water supply central heating and elekticity are most of
the installed within the floors and walls to keep those out of sight.
Problems with unintended warming up of potable water (hot spots) arise when the pipes
are to close from each other, or are crossing each other.
169 Fig. 2.4 Problems with “hot spots” in floors and walls
An additional problem is the vertical temperature gradient in a room with radiator
heating (2 – 3 ºC) . When floor heating is used than the vertical temperature gradient is
almost zero.
Fig. 2.5 Verical temperature gradient in rooms with heating
170 TNO (3) has calculated the minimum distances in floors between drinking water pipes
and central heating pipes for different temperatures and situations. For that purpose
TNO used the universal modelling programm Comsol Physics. In figure 2.6 you see a
visual presentation of this programm of one of the situations that were calculated.
Fig. 2.6 Temperature gradient in a finishing floor with floor heating pipes at a
water temperature of 40 oC and a room temperature of 20 oC.
171 Table 2.1: Results of calculations with Comsol Physics for radiator heating (above) and
flor heating (below)
Pipe insulation
temperature [°C]
minimum
horizontal
room
room
distance
under
above
[mm]
floor
floor
1
2
3
4
5
Situation with radiator heating pipes in concrete floor
o
insulation 10 mm
80/60 C
20
23
450
55/40 oC
20
22
200
40/30 oC
20
22
100
jacket pipe
80/60 oC
20
23
750
55/40 oC
20
22
350
40/30 oC
20
22
150
o
insulation 10 mm
80/60 C
22
25
Not allowed
o
55/40 C
22
24
450
40/30 oC
22
24
200
jacket pipe
80/60 oC
22
25
Not allowed
55/40 oC
22
24
650
40/30 oC
22
24
350
Situation with hot water circulation pipes in concrete floor (70 °C)
o
insulation 10 mm
70 C
20
23
400
jacket pipe
20
23
650
o
insulation 10 mm
70 C
22
25
Not allowed
Jacket pipe
22
25
Not allowed
Values in this table are calculated by
Scientific institute TNO, Apeldoorn
Heating system in
the room under
flooring
feed /
return
temperatuur [°C]
floor
room room
heating
above under
water
floor floor
minimum
Iinsula- horizontal
tion in distance from
between floor heating
[mm]
Radiator heating
light carpet
5 0°C
20
23
300
22
25
400
Floor heating
50 °C
20
20
250
22
22
350
24
24
500
20
20
no
250
22
22
350
24
24
700
40 °C
20
20
yes
200
22
22
250
24
24
400
20
20
no
150
22
22
250
24
24
500
30 °C
20
20
yes
100
22
22
150
24
24
250
20
20
no
50
22
22
100
24
24
250
*) mhd = minimum horizontal distance from outer pipe of floor heating
Values in this table are calculated by Scientific institute TNO, Apeldoorn
172 Crossing of
watersupply
pipe allowed?
6
no
yes
no
no
Water supply pipe in constructio
allowed ?
yes, mhd 250 mm
no
yes
yes, mhd 150 mm
yes, mhd 500 mm
no
yes
yes, mhd 400 mm
no
yes
no
Conclusions from the tables
173 See for example the lay out in fig. 2.7.
Fig. 2.7 Lay out with hots spots due to crossings and small distances between cental
heauting and potable water pipes
2.2.2 Phase of analysis
Interviews with various various installers of potable water systems en central heauting
systems and appartement buildings were helt. In the interviews mainly problems with
organisational origine were mentioned. Dutch standard NEN 1006 (2) and ISSO 30.5
(3) are well known under consultants and engineers, but the implementation in the
construction phase fails. The central heating system is often layed by another contractor
than the plumber of the potable water system. There is lack of coordination.
The installer of the central heating system is not interested in prevention of hot spots.
174 Another very important problem is that the architect or designer of the family house are
not aware of this matter and he often creates a lay out in which it is almost impossible to
prevent crossings of central heating pipes and potable water pipes.
That’s why mr. Kuipers of biq stadsontwerp (2) choose top develop a method of
integrated design of family houses in which the architect or designer has to create
separate cool zones for the potable water system in an very early stage of the buiding
process. See for example the lay out in fig. 2.7.
Fig. 2.7 Lay out with hots spots due to crossings and small distances between cental
heating and potable water pipes
175 Fig. 2.8 Central heating and potable water pipestogether at a door passage
2.2.3 ISSO-SBR-directive 811 “Hotspotfree design, building and installation”
The method of preventing hot spots in family houses is published in ISSO-SBRdirective 811 (4).
The method contains resumed the following:
1. Create uninterrupted waterzones like shown below. We see three types of zones.
Fig. 2.8 Central
Fig. 2.9 Three types of uninterrupted waterzones
176 2. Connect piled waterzones in a multy floor family house with separate shafts for
water supply (cool shaft) and central heating (warm shaft)
Fig. 2.10 Connecting of the different zones with separate shafts
The water zone (cool zone) consists of:
-
Potable water and hot water supply pipes;
Draw off points, faucets, etc.;
Delivery point of potable water (watermeter);
Water supply shaft;
Room /cupboard for waterheater.
The heating zone (warm zone) consists of:
-
Pipes for radiator- and floorheating;
Radiators;
Distributors;
Heating shaft.
Guidelines for zoning:
For all three types of zoning:
-
no walls for mounting radiators in water zone;
no heating shaft in the water zone.
For the “corner” type:
177 -
Water supply shaft at tail end of water zone.
For the “knot” and the “island” type:
-
Reserve enough room for heating pipes;
Connect water zones with the water supply shaft;
Connect heating zones with warm heating shaft.
Guidelines for piling of zones:
-
At each floor the guidelines for its zone type counts.
Example of zoning:
Example of creating zones in a family house :
178 179 What to do when zoning is not possible?
One has to choose out of two options:
1. Use of floor heating T  40 oC with insulation or low temperature heating
with radiators T  40/30 oC.
2. Do not install the water supply pipes in the finishing floor at places where:water
supply pipes cross de central heating pipeswater supply pipes are to close to
central
heating
pipes
Option 1 is only possible in low energy house. Option 2 is difficult to work out the
building practice in The Netherlands.
How to organize zoning?
-
The architect / designer has to create separate cool and warm zones and seperate
shafts voor water supply and heating (He has to “think ahead”)
The plumber of the water supply and the installer of central heating have to
coordinate their lay outs
The plumber must check the distances between the water supply pipes and the
central heating pipes before the finishing floor is poured.
Summary:
1. To prevent unintentional warming of potable water all parties in the building
process have to practise hot spot free design, building and installation.
2. Create separate cool zones for the potable water installation, or if not possible
use central heating with a low temperature regime
3. Don’t forget to organize and coordinate
180 3 References
1.
TNO
Technical
report
2008-A-R0664/B, Ir. J. van Wolferen
Recommendations for prevention of inintentional
warming of potable water pipes in floors with
pipes for floor heating, radiatorheuating and
hotwatercirculation.
2008
2.
NEN 1006, “General requirements for potable NEN, Delft
water supply installations” (see www.nen.nl)
2007
3.
ISSO-publication 30.5, LegionellaCode for ISSO
potable water installations in family houses (see
www.isso.nl)
2008
4
ISSO-SBR-publication 811, “Hotspotfree design, ISSO and SBR
building and installation” (see www.isso.nl)
2005
5
ISSO-publication
55.2,
Guidelines
for ISSO
Legionellaprevention in collective potable water
systems (see www.isso.nl)
2005
4 Presentation of Author
Oscar Nuijten is currently working as a project coordinator
with ISSO (knowledge centre for building services in
Rotterdam, The Netherlands), where he is responsible for
the realisation of technical publications in the field of
sanitary and gas installations and guidelines for
legionellaprevention.
181 III.4
An Investigation into Factors Affecting the
Design Techniques Used to Control
Legionella in Water Systems
Paul Angus
BEng (Hons) EngTech, FCIPHE, ACIBSE RP
Senior Public Health Engineer, WSP Ltd.
Steven Ingle
–
DSc, MSc, I Eng, FCIPHE, ACIBSE, FSoPHE, LCGI, RP
Director, Ingle Project Design Ltd.
Derek King
MPhil BEng (Hons) Cert Ed CEng MCIBSE
Senior Lecturer, Liverpool John Moores University
John Turner
MSc, CEnv, I Eng, MCIWEM, FIHEEM, FCIPHE,
ACIBSE.
Director, Public Health Engineering Direct Ltd.
Abstract
Over the past ten years there has been much development in various and novel
means of controlling Legionella in water systems, whilst trying to maintain
compliance with the current water and health & safety regulations.
However, as most large projects now are value engineering driven, in the current
economic climate there is often an insistence on re-interpreting the Codes of
Practice to provide designs with cost advantages. In many instances this saving
may not provide the best overall cost and low risk solution for a particular
installation.
182 Decision makers in the construction process, driven by financial considerations,
appear too often to insist on the most cost effective design solutions, as long as the
spirit of the design code indicates that such a solution is acceptable.
This paper suggests that this philosophy encourages undesirable compromises. An
environment is fostered where decisions may often be made on a value engineering
and profit basis, while leaving real public health risks and non-sustainable, nonmaintainable systems in their wake.
Introduction
It is intended that this paper will critically analyse some methods of control of legionella
bacteria in hot and cold water systems.
The paper will address, in particular, but not be confined to, the impact of using
temperature as a sole means of Legionella control in domestic hot water systems based
on value engineering design. Under examination also are the issues of management and
operational costs of utilising this practice, and, in the context of the constant quest for
CO2 reduction and sustainability, whether such a technique can be considered
environmentally friendly.
Furthermore, the paper will discuss the use of trace heating without secondary
circulation pipework, and the associated risks and advantages of this application, along
with the benefits or otherwise of alternative treatments.
When considering cold water systems, the present Codes of Practice accept the widely
held view that at temperatures below 20oC, Legionella bacteria remain dormant2.
Furthermore, it is commonly accepted that cold water pipework will remain at “safe”
temperatures without the need for any particular practical step to guarantee this (other
than insulating pipework, which has long been standard procedure). However, global
warming is beginning to cause appreciably higher air and ground temperatures. In
addition, changed rainfall patterns may mean that water supplies in the UK must come
from alternative sources, even from desalination processes (as in the Middle East).
Thus cold water may well be supplied to buildings at temperatures closer to the higher
end of what may be considered safe.
22 HSE. (2000) Legionnaires Disease: The control of legionella bacteria in water systems. Approved
Code of Practice and Guidance L8, Health and Safety Executive.
183 Bacterial growth in hydrodynamic systems
Research has long shown that the use of heat and chlorine treatments to control bacterial
growth in water pipework has given rise to certain bacteria developing higher tolerances
to both heat3 and chlorine4. Legionella bacteria have been isolated from natural
environmental locations where temperatures have been as low as 5oC and as high 63oC3.
Isolation in systems at temperatures maintained at 30oC – 54oC is common, however
legionella bacteria are not found at temperatures above 71oC5. In addition, it has been
shown that thermophilous bacteria, Pseudomonas aeruginosa and other similar bacteria
are becoming increasingly common and may “partner” the Legionella bacterium. These
bacteria are believed to cause an increase in health problems, particularly those related
to the skin6. Free-living amoebae are often found alongside Legionellae and other
pathogenic bacteria, amoebae of the genera Naegleria and Acanthamoeba can be
pathogenic to humans. These can cause rapid fatal meningoencephalititis, when
introduced through the nasal mucosae, as well as brain, eye, pulmonary and kidney
infections7.
It is an accepted control technique in hydrodynamic systems that a temperature should
be maintained at which bacterial life cannot be sustained. Such factors as “kill time”
(the time taken for temperature to kill bacteria) and “residence time” (the time that
water stays in the system before being drawn off), however, are also critical. If the
residence time in a system is shorter than the kill time (and this can easily be the case at
60oC) then this is clearly unsatisfactory. At 60oC it takes 32 minutes to kill Legionella
bacteria, while at 55oC this period is increased to 5-6 hours3.
3Wadowsky RM et al.(1985) Effect of Temperature, pH ,and Oxygen levels on the multiplication of naturally occurring Legionella pneumophila in Potable Water. Applied & Environmental Microbiology May 1985 p 1197‐1205 4Ridgway HF. Olson BH. (1982) Chlorine resistance patterns of bacteria from two drinking water distribution systems. Applied & Environmental Microbiology 44(4)972‐987. 5Wadowsky RM et al. (1982) Hot water systems as sources of Legionella pneumophila in hospital and non hospital plumbing fixtures. Applied & Environmental Microbiology 43;1104‐
1110 6Ovesen K, Schmidt‐Jorgensen F, & Bagh L. (1993) Bacteria growth in hot water systems. Danish National Building Research Institute 7Tyndall RL, Domingue EL. (1982) Cocultivation of Legionella pneumophila and free living Amoebae. Applied & Environmental Microbiology44;954‐959 184 It is also known that dead bacterial cells can provide a food source for living bacteria5,
and that bacteria can survive in anaerobic conditions (although they do not grow or
multiply readily in such conditions3). Mutant strains have also been shown to develop,
depending upon the quality of the source supply, and materials utilised in the system
installation, galvanised steel, copper and plastics being the most common. Corrosion in
galvanised steel in particular is accelerated at temperatures above 60oC, as is the build
up of debris from the killed bacteria in all materials at these elevated temperatures. The
uneven surfaces of corroded pipework and settled detritus then provide nutrition and
breeding grounds for bacteria. Plastics have been shown to actively increase bacterial
growth rates across all temperature ranges6.
In many cases, legionella bacteria are found in the sediment in cooling towers,
calorifiers or pipework systems – even at 65oC high colony counts have been found in
sludge in calorifiers8. In cooling towers, where water is treated using biocides, it is not
unusual for the biocides to have to be alternated to ensure that the bacteria do not
become tolerant.
Standing water is known to be undesirable in hydrodynamic systems: longer residence
times lead to stagnation and this is known to increase the growth rate of bacteria as
more nutrients become available. Heat transfer between hot and cold water services
pipework in close proximity is also inevitable if water is left standing in the pipework
for long periods, no matter how well pipes are insulated, and this is even more
unhelpful. Stagnant water in a system is considered to represent a fluid category 3 risk
under the current water regulations9.
If engineers could interpret such micro-biological research language and relate this to
practical and effective engineering design reality, they could possibly make design
recommendations with confidence and rebut the constant arguments in favour of value
engineered solutions and cost cutting, which may well compromise public health and
sustainability.
Influences on design techniques
Hot water supply accounts for as much as 30% of all energy consumption in buildings
such as hotels, hospitals and residential homes.
It is estimated that 65-70% of hydrodynamic systems in public buildings utilise heat as
a means of “preferred control” of contamination10 and this approach is seen to comply
8Schmidt‐Jorgensen F.(1991) Danish investigations on bacterial contamination in hot water systems. Danish National Building Research Institute. Building Services Division. 9
HMSO. (1999) The Water Supply (Water Fittings) Regulations 1999
10 Arrowsmith, M. (2006) Legionnaire’s Disease Risk Minimisation, IHEEM Journal, August 2006.
185 with all design guidance and legislation. In practice, however, particularly in large
systems, it is difficult to ensure that balancing achieves the required temperatures at all
draw off locations, or allows the whole system to be properly pasteurised in the heat
cycle. It is difficult to be certain that frequent stagnation does not occur at locations
with low water usage (e.g. washbasins in hospitals). In addition, designers seldom fully
address the importance of temperature maintenance of the cold water distribution
network, and the consequences should cold supply temperature be allowed deviate.
Over recent years there have been significant changes to how we store, heat and
distribute water around buildings. These changes are mainly influenced by the ongoing
pressures to reduce space taken up by mechanical services equipment and “value
engineered” solutions to reduce costs. Thus centralised plant rooms have tended to
become the norm, coupled with boosted or pressurised unvented hot and cold water
networks with extensive distribution systems. Off-site prefabrication of pipework has
also been introduced, often with little consideration of possible heat transfer between
services, coupled with the move away from copper and other metallic pipework
materials (common in the UK) to plastics. Strangely enough, there are many examples
of where strainers or filters are added to a system to protect equipment, but these are not
considered as a necessity to help reduce system contamination and protect public health
on incoming water mains.
Selection of pipework sizes is another factor which should be examined when looking at
established design practices: Current practice is to use the loading units method
described in BS6700, and in the CIPHE and CIBSE Design Guides. In this method
pipework sizes are selected based on the inter-relationships between several dynamic
properties of water and the pipework material, though flow velocity and its relationship
to pipework generated noise is used as the main limiting factor. Normal practice when
using this method is to maintain velocities below 1 m/s, or exceptionally 1.5 m/s. It is,
however, some time since any study was carried out as to the extent that noise generated
in pipework is actually likely to be problematic in modern buildings, given that
installation practice is constantly evolving. For instance, pipes are more likely to be
located centrally and in ducts, remote from working areas of buildings, and all pipework
is routinely thermally insulated. Furthermore, at larger pipe sizes, there is evidence that
the loading unit method becomes unreliable and over-sizing is routine. This, as well as
being non-sustainable, is likely to increase the chance of stagnation and precipitation of
debris. It may thus be possible to relax the requirement to keep noise at low levels so
that flow velocities can be increased, and this would promote a useful scouring action
within pipes. Consequently there would be less likelihood of debris and detritus settling
and providing nutrition or breeding places for bacteria. The welcome by-product of
such a technique would reduce pipework sizes and thus contribute further to
sustainability.
186 More recently, the pressure to reduce carbon footprints and make savings of energy and
water have moved designers into new areas of risk and managing that risk. For example
in the UK, BREEAM points can be earned by utilising solenoid valves controlled by
Passive infrared detectors (PIRs) to isolate infrequently used hot and cold water supplies
to sanitary fittings11. There is no guidance to insist that the greatest of care should be
taken in location of these valves while there is quite clearly a risk that pipework dead
legs and possible stagnation may inadvertently be the result of using this method. It
could certainly be argued here that public health is potentially compromised for the sake
of sustainability!
There is also a constant paradoxical question for engineers as to how it can be possible
to reduce the fuel costs of producing hot water, when increasingly elevated temperatures
are advised (in preference to biocides to which bacteria grow immune) in an attempt to
control bacterial contamination of supplies. How much benefit is there then, from the
emergence of condensing boilers, which perform at their highest efficiency with return
temperatures of 40-45oC?
When questioned, many practicing design engineers and facilities or maintenance
managers will not know exactly the quality of the water entering their building, though
most will affirm that it “complies with the Regulatory Drinking Water Standards”12.
Given the current financial climate and drive towards sustainability, the chances of
reducing resultant levels of settled suspended matter, scale formation and corrosion, all
contributors to fouling and microbiological activity, seem minimal.
From a water conservation perspective it is noted that rainwater harvesting and re-use of
grey and even black water is gathering pace, certainly as again it attracts BREEAM
points. From a public health viewpoint, however, how sensible is it that water is
brought into buildings that could potentially contain high numbers of pathogenic
bacteria for possible aerosol dispersion within confined spaces such as toilet cubicles?
The future risk of erroneous cross-connection between re-cycled water pipework and
potable water pipework is also often not considered in the design of such systems.
Legionella in cold water systems
Legionella is often considered as being a problem associated only with hot water
installations. As previously stated however, it is necessary to ensure that temperatures
in cold water systems are maintained ideally below 20oC2, although the current UK
water regulations permit utilities companies to supply at temperatures up to 25oC9.
11 BREEAM. (2008) The Code for Sustainable Homes
12 Proceedings of Legionella Seminar. (2009) Old Trafford Cricket Ground, Old Trafford,
Manchester, 05/11/2009
187 Cold water piped installation systems are required to be insulated and kept as far away
as possible, from sources of heat such as hot water pipe-work, warm air ductwork and
electrical services. Practically, this is difficult to achieve since service risers and ducts
are often of restricted size and are usually expected to accommodate all services, thus
much may depend upon the skill and ingenuity of the installer. Such concerns may
easily be overlooked on site, especially where cold water pipework is installed within
ceiling voids or plant rooms.
It is of upmost importance to consider carefully the location of the cold water storage
cistern. The installation itself should comply with the Water Supply (Water fittings)
Regulations 19999 and Water Byelaws 2000 in Scotland13. It is the responsibility of the
design engineer to ensure the cistern is located in a cool, dry place, being protected from
extremes of temperature by thermal insulation. The cold water storage requirement for
everyday purposes such as drinking, washing and cooking, is normally determined on
the basis of compensating for a 24 hour interruption of supply, as indicated by the
appropriate British and European Standards14. Where a larger storage volume is needed
for particular requirements such as catering, this may require additional treatment such
as regular or continuous chlorine dosing, chlorine dioxide dosing or other means of
disinfection such as ionisation.
Figure 1: Cold Water Storage and system distribution temperatures (Source: Copper
Development Association)
13 HMSO. (2000) Water Bye Laws (Scotland).
14 BSI (2006) BS6700 Design, installation, testing and maintenance of services supplying water for
domestic use within buildings and their curtilages.
188 In countries where high ambient temperatures are usual, such as in the Middle East, the
incoming cold water mains supply to the building can often be characteristically in
excess of 20°C. As discussed earlier within this paper, such temperatures provide ideal
conditions for Legionella bacteria to breed within pipework and components like tanks.
In order to combat this problem, the water must routinely be disinfected, usually
chemically or by ionisation treatment, before distribution. In addition storage tanks are
often located in the basements of buildings, taking advantage of constant cooler
temperatures underground. As global warming takes effect, such techniques are likely
to be required in parts of the world where the problem of cold supply temperatures were
previously seldom considered, such as in northern Europe and North America.
Legionella in hot water systems
In practice it is known that around half of the reported cases of Legionnaires disease are
associated with contaminated hot water, though these outbreaks are not normally widely
reported because the number of cases involved is relatively small. It is helpful to
examine the situation in homes and commercial buildings separately.

Housing
There have been only occasional links between the hot water services in housing and
Legionnaires’ disease, despite surveys of domestic systems showing there are indeed
traces of legionella. The main reasons for this are:
1. It is difficult to link persons to any one source: households are unlikely to contain
more than one susceptible occupant, and since an epidemic would not be recognized
unless two or more cases arose in the same household, General Practioner’s and
epidemiologists do not search for a common link.
2. There is rapid turnover of hot water in domestic systems, with residence time in the
pipework and components tending to be very short. In addition there is an ever
increasing use of instantaneous water heaters or combination boilers. The modest
size of a five person British house (around 80 m²) also means that there tend not to
be long dead legs of piping.
3. Domestic hot water systems tend to comprise of predominantly copper pipework
and storage vessels (though plastics have been introduced increasingly in recent
years). It has been shown that micro-organisms find it difficult to colonize copper
pipework and components: microbiological colonization in pipework always starts
at the wall surfaces, and copper a natural biocide when new initially inhibits this
189 initially until the surface becomes oxidised15,16. (It is, however, worth making the
point that in hard water areas, this protection quickly disappears when mineral
deposits coat the wall surfaces.)
Once such a storage cylinder is heated and at steady state, stratification becomes evident
and a temperature gradient exists, which is shown in Figure 217.
Figure 2: Temperature profiles in an indirectly heated water cylinder (Source:
Whiteside, 1990)
The potential problem area in such hot water cylinders lies in the base. Debris from
killed bacteria will drop to the base during time of non draw off and will collect there
over time, thus providing nutrition for bacteria, and, due to the cold feed water being
supplied in this region of the cylinder, temperatures at the base can be lukewarm
(around 30oC) for long periods of time. If the base of the cylinder does heat up to an
acceptable temperature, such a temperature may not be maintained for a sufficiently
long period to kill the bacteria. The UK Health & Safety Executive (HSE) recommends
15 Rideal S. and Baines E. (1904) The suggested use of copper drinking vessels as a prophylactic
against water borne typhoid. Royal Sanitation lnstitute Journal (London), 25 591-595
16 Place, F.E. (1905) Water cleansing by copper. Journal of Preventive Medicine, (London), 13, 379
17 Whiteside, D. (1990) DHW: heat loss from tanks. Building Services Journal 12 (l), 57
190 that in larger storage vessels a de-stratification pump be fitted to circulate water from
the top of the cylinder to the bottom via the cold feed2.
In a typical domestic hot water distribution system thermal insulation is usually
sufficient to minimize heat losses and maximize outlet temperatures at the taps. Selfregulating trace heating, which is a relatively new technique, claimed by many to be
cost effective, may also be considered in conjunction with the thermal insulation,
particularly if dead legs exceed regulations. This technique offers instant hot water at
the outlets, at some energy cost, but with less wastage of water since when users operate
the taps they do not need to let the tap run while waiting for hot water.
Instantaneous electric water heaters are used for domestic showers. These are
connected directly to the mains cold water supply and supply water is directed over a
powerful (up to 12 kW) immersion heater, emerging hot and at pressure. Such units are
also available as spray taps for hand washing.
It has been shown and is well known that lime deposition rates in hard water areas
increase dramatically above 60°C, whereas in soft water areas there is no deposition
problem. However, corrosion is accelerated at the higher temperatures18.

Commercial buildings
The hot water system in commercial buildings starts with a large storage vessel known
as a calorifier. In large buildings this contains a heating coil through which hot water
from the boiler plant flows. Some plant contains both LPHW and steam.
Four factors are desirable in the design of a calorifier to minimize the risk of
contamination with legionella:
1. The bulk storage temperature should be reasonably uniform throughout the vessel
when fully heated. In older systems it is common for the lowest 100 mm in the
vicinity of the cold inlet feed to be lukewarm. This potentially lukewarm zone in
the base, with the debris mentioned above, can provide a most suitable environment
for the multiplication of legionella. A technique of preventing this is to include a
de-stratification pump (as recommended by the HSE in Approved Code of Practice
L82) which continually circulates water from the top to the bottom of the calorifier.
Alternatively, the secondary return pipe of the hot water system can be connected
into the cold feed pipe at the bottom of the calorifier instead of its normal
connection point two-thirds of the distance from the bottom.
18 ASHRAE (1987) HVAC Handbook American Society of Heating Refrigeration and Air
Conditioning Engineers, Atlanta, USA
191 2. The calorifier and entire distribution system should be capable of being heated to
70°C for pasteurisation purposes, and must be capable of maintaining this
temperature for some hours if necessary.
3. Calorifiers must have easy access for draining, dismantling and cleaning2. All
calorifiers which are heated by pressurised high temperature water must be
inspected by law to ensure that the heating coil is able to withstand the pressure and
so must be fitted with an access hatch. This hatch can also be used for cleaning
purposes when the calorifier is drained down ready for inspection. Some engineers,
however, specify no hatch where there is no heating coil, which is technically
permissible, but is seen as bad practice.
4. Volumetric sizing is closely linked with the heat recovery time of the heat
exchanger in the calorifier and therefore calorifier should be adequately sized for its
duty. A two hour recovery period is usual but the storage volume can be reduced if
a shorter recovery time is desired, dependant on the energy input available. If the
unit is under-sized in volume, then at times of high hot water flow demand, the
outlet temperature will fall and water at temperatures lower than the design
temperature will pass through the system, carrying with it initially dormant bacteria
from the mains supply. Subsequently if the water in the base of the calorifier is
contaminated then the contamination will be spread downstream through the
distribution network. This passing of live bacteria into the distribution network will
of course allow the bacteria to colonise in dead legs and areas of low velocity. In
practice the current design guides are over generous in their sizing and there is little
risk of under-sizing in new installations19. It has been found that the highest
discharge temperature during draw off is achieved by use of vessels with a height
four times their width, these dimensions providing optimum stratification
characteristics during discharge20.
5. On pressurised unvented hot water systems care must be taken to allow maintenance
of the expansion vessel. The expansion vessel should maintain a temperature below
20oC and be constructed of a material that does not support bacterial growth.
For buildings in continuous use (such as hospitals, hotels etc.) it is usual to provide two
calorifiers to allow for maintenance: a service and standby calorifier both sized at 100%
capacity may be provided, or alternatively, two calorifiers sized at two-thirds capacity.
This latter option is preferable since the standby unit will come online more frequently
19 Corless DK. (1990) Hot water services for a modern hospital ward unit. Building Services
Engineering Research and Technology 11 (2), 57-63
20 Cole, RL. and Bellinger F.O. (1982) Thermally stratified tanks. ASHRAE Transactions, 88 (2)
1005-1017
192 at times of high demand and this will ensure that water is not allowed to stagnate. In
systems where the standby unit is offline for considerable periods, it will remain full of
water and may be heated slightly by leakage at the valves controlling flow to the heating
coil. Thus, such standby calorifiers must be considered contaminated and should be
pasteurised before being brought back into service. Pasteurisation is achieved by
bringing the water temperature up to a minimum of 60°C and maintaining this
temperature for at least 2 hours. During pasteurisation hot water must be circulated
from the calorifier to the various outlet points around the building and back to the
calorifier via a return pipework loop. The loop takes the water from the hottest part of
the calorifier, the top, and returns it to the calorifier cold feed pipe. A small water pump
is provided to circulate the water gently but sufficiently fast for the heat losses to permit
the water return temperature to be 50°C or more when the water leaves the calorifier at
70°C.
Commercial kitchens also deserve special mention because they require water at 82°C
for dish washing rinsing. This can be achieved by controlling the calorifier at 82°C for
the dish washer and blending cold water to provide temperatures around 40°C at normal
outlets to prevent scalding, or by maintaining the calorifier at 60°C and providing
specialist local heating for the dish washers. Both systems are likely to be legionellafree at these temperatures.
A particular problem with high temperature calorifier operation is the increase in water
volume with temperature, and this relationship is illustrated in Figure 321. This increase
in volume may force the water at the bottom of the calorifier back into the cold feed
supply as well as through the open vent pipe. Since the water at the base of the
calorifier is likely to be the most contaminated then this expansion can contaminate the
feed pipe. This could subsequently lead to rapid recontamination after calorifier
cleaning. Non-return valves (NRVs) are recommended by the HSE (Approved Code of
Practice L8) at the cold feed inlet to calorifiers2, however, careful thought is needed
when positioning these, since an incorrect location could result in wastage of heat
energy in the water and water itself. In pressurised systems the expansion vessel will
contain water at ideal temperatures for bacterial colonisation, thus careful thought is
needed to incorporate this piece of equipment such that bacteria cannot enter the
distribution pipework.
21 Oughton D. (2008) Faber & Kell's Heating & Air-conditioning of Buildings, Butterworth
Heinemann.
193 Figure 3: Expansion volume of heated water (inlet 4oC) (Source: Oughton, 2008)
In public buildings temperature checks under HSE L8 requirements2 are normally
carried out at as part of the maintenance regime at tap outlets to ensure that hot water is
delivered within a reasonable time, and that the delivery temperature is not below 46°C.
In hospitals special guidance defines reasonable time as within 1 minute of opening the
tap the water is expected to reach 50°C, although if a blending device is used to lower
the outlet temperature to prevent scalding of vulnerable users, then 50°C at the hot inlet
to the blending valve is recommended22.
Maintenance for public buildings must include regular inspection of shower heads to
check that they are clean and, if dirty, they should be cleaned immediately to avoid the
colonisation of the debris with micro-organisms. Taps or showers which are rarely used
should be removed along with the associated pipework to minimize the danger of
lukewarm water remaining stagnant for long periods.
It is seen as good design practice that the tap or outlet most frequently used should be at
the end of any dead leg to ensure regular flow through the length of the supply pipe. It
has long been standard practice to insulate both hot and cold pipework and ensure that
the hot pipe is located above the cold so that convection heat loss from the hot pipe will
not readily affect the cold supply.
Early guidelines were concerned with the small volume of warm water left in shower
hoses after use. There were proposals that self-draining showers were inherently safer
than conventional showers. Experiments on showers with a self-draining valve and
conventional showers without such a valve showed little difference to re-colonisation by
22 DHSS (1990) Control of Legionellae and safe hot water temperatures. Health Notice HM
(90)12, HN (EP) (90)
194 bacteria after sterilization. After 10 days the re-colonisation was stable and very similar
for both types of shower23. It has been found, however, that removal of dead legs,
regular changing of the mixing valve components, and regular flushing of shower hoses
and heads produces significant reductions in legionella24.
Secondary return systems and trace heating
The conventional approach to the design of hot water services in commercial buildings
recirculates hot water through a secondary return system as depicted in figure 4. This
approach ensures that water flows constantly through the pipework and does not
stagnate even if a tap is not used for some time. It ensures that water is maintained at
design temperature around the network, and it allows the designer to keep supply dead
legs to outlets as short as possible. These systems prevent the inconvenience of waiting
for hot water to arrive as well as preventing bacterial contamination in lengths of pipe
containing tepid water. Continuous circulation is normally achieved by pumping,
although in some cases this is programmed to switch off at night.
Open vent pipe
Secondary flow pipe
Draw off points Hot storage vessel Control Secondary return Circulating pump Figure 4: Representative sketch of secondary return type hot water distribution
An alternative method of maintaining temperature around the network is to provide selfregulating electrical trace heating along a single delivery pipe. The operating
temperature of the pipe is determined from the knowledge of the heating tape
characteristics with respect to temperature, and knowledge of the heat loss
characteristics of the pipe material. The power output of the tape falls with higher
23 Humphrey TJ. (1989) Microbial contamination of hospital showers and shower water: the effect
of an automatic drain valve. Journal of Hospital Infection, 13, 55-61
24 Makin T. and Hart CA. (1991) The effect of a self-regulating trace heating element on legionella
within a shower. Journal of Applied Bacteriology, 70, 258-264
195 temperatures and the heat loss from the thermally insulated pipe increases with higher
pipe temperatures. These two characteristic curves can be plotted and the intersection
between these determines the equilibrium temperature of the trace heated pipe. This
procedure is illustrated in Figure 5.
Figure 5: Determination of operating temperature of trace heated pipework
The installed piping scheme is much simpler than the conventional one and, on the face
of it, appears to be more economically appealing. However, the energy costs of
electrical pumping and pipe heat losses of the conventional system must be compared
with the energy costs of running the single pipe trace heated system. The capital costs
of single pipe with trace heating and the re-circulating system should also be compared.
The economic case is often strongest when the hot water service is a long and narrow
building where the return pipe simply returns the water. It is less attractive in a square
building where the hot water circuit may be one pipe running in a loop around the
building.
There are two other applications of trace heating for hot water pipes. The first is to
remedy deficiencies in existing equipment or systems. If there are long dead legs in the
distribution system then electrical trace heating can be incorporated simply along this
particular length of dead leg. Experiments using the technique to maintain the
temperature in the dead legs in a shower to 50°C were very successful provided that the
circulating hot water was greater than 45°C. The temperature was maintained very
closely (± 1.5°C) and legionella were eradicated or severely reduced by the technique25.
25 Makin T. and Hart C.A. (1990) The efficiency of control measures for eradicating Legionellae in
showers. Journal of Hospital Infection, 16, 1-7
196 The second is for plain trace heating used for pasteurisation. There is a growing
practice to recommend regular elevation of hot water pipes to 77°C for 15 minutes and
to draw this water from each tap or shower head to disinfect the components18,26. Care
must be taken if this practice is followed to ensure that the deposition of calcium
carbonate will not be excessive or that the corrosion rate will not become high in soft
water areas. In some natural waters where zinc is used as a natural protective coating
on steel a reversal of polarity occurs around 60-65 °C. The zinc then becomes cathodic
to steel stimulating attack which results in localised pitting corrosion27,28.
A suggestion
This paper suggests that general design good practice for public buildings is often being
compromised by too much focus on project economics. Furthermore the client often
does not seem to benefit from any cost savings realised, and the end user is left to pick
up the costs for the ongoing failure or poor performance of systems after handover
(although it must be said that in speculative design and build projects, clients are more
likely to see some of the savings). Currently, project managers are heeding many
manufacturers’ claims of more inexpensive installation costs and thus are constantly
requesting designers to carry out value engineering of projects. Capital costs may
subsequently be reduced but life cycle costs seem to be ignored altogether.
One such aspect which raises concern in the context of this paper is the recent tendency
to introduce trace heating as a means of temperature control, and this technique is now
being promoted as an economic alternative to traditional secondary return type hot
water distribution systems. There are a number of issues that are raised by the
introduction of this type of system:




Maintenance of design temperatures throughout distribution networks;
Scale build up, stagnation and allowing suspended matter to settle out thus
creating corrosion spots and bacterial breeding grounds;
Energy use and life cycle costs;
Carbon footprint and sustainability.
26 Fliermans CB. and Nygren JA. (1987) Maintaining industrial cooling systems free of Legionella
pneumophila. ASHRAE Transactions, 93 (2) 1405-1415
27 BRE. (1968) Durability of metals in natural waters. Building Research Station Digest 98 Second
Series
28 Institute of Plumbing (1988) Plumbing Engineering Services Design Guide, Institute of
Plumbing, Hornchurch
197 The installation of central water heating plant with distribution pipework running
throughout the building is generally considered by engineers as the most cost effective,
efficient and reliable solution to providing hot water supplies in public buildings. The
cost effectiveness and reliability of this tried and tested system is borne out by the long
life (usually 30 or more years) of low stress components such as traditional calorifiers,
tube bundles, boilers and so on. The question of cost effective and reliable methods of
maintaining temperatures in the distribution pipe networks at safe levels is where there
may be some disagreement.
It is claimed that temperature maintenance tape will cut both installation and running
costs in comparison to a re-circulating hot water distribution system, whilst maintaining
temperature within dead legs, and will thus prevent Legionella growth and allow full
compliance with current legislation. The main cost argument is clearly the removal of
the traditional insulated domestic hot water secondary return line, with its associated
balancing valves and circulator pump. The function of the return pipe is to maintain the
temperature within the system such that water is available at the point of use quickly,
avoiding wastage of large quantities of water whilst waiting for hot water to arrive –
trace heated systems could make the same claim. To ensure maintenance of safe
temperatures throughout systems, the same quantity of heat must be delivered as that
which is lost due to natural cooling to the ambient surroundings. In the traditional
system standing losses in primary energy are claimed to be considerably higher than in
systems using trace heating; in addition, installation costs for trace heating systems are
lower and thus the overall capital costs for a traditional system are usually far higher.
Engineers have a duty to their client and to the environment and should consider wider
aspects in relation to the installation. In fulfilling these responsibilities it is only proper
that detailed comparisons between the two systems of temperature maintenance be
made.




The environmental impact and sustainability of the system should be fully
considered: unless the trace heating installation is to be powered by electricity
generated from renewable sources, then heating water using electricity instead of
gas is inherently non-sustainable and energy inefficient.
The choice of primary fuel is an important deciding factor: within the UK
purchasing costs of fuels generally place the use of natural gas in front of others
in an urban environment. If an alternative fuel must be used, or electricity is
generated on-site, this may radically change the dynamic of the decision making
process.
Ongoing operational costs must be determined by carrying out a series of
detailed calculations and forecasts, detailing exactly how the operating costs of
the options compare;
Detailed knowledge of the relative life cycle costs for all components is required
to enable the engineer to make design decisions with confidence – for example
198 
the likely ongoing maintenance, repair and replacement requirements and costs
must be considered;
In an increasingly litigious world the engineer needs to be confident when
specifying either system, that safe temperatures are maintained and the
requirements of legislation and design codes are met.
Conclusions
Trace heating installations on the larger scale have little recorded history. It is known
that they can be effective in treating relatively short dead legs such as shower runs and
remote appliances; there is, however, scant evidence to suggest this technique can be
effectively scaled up to meet all criteria required for public health. For instance, it is not
known exactly what happens in a trace heated pipe when positioned vertically when
water is not moving: does stratification and its associated problems occur as it does in
storage cylinders?
Further research is essential if engineers are to have confidence in utilising the trace
heating technique as a reliable solution. There are at present too many unknown factors
relating to the points made earlier: maintaining design temperatures throughout
distribution networks, scale build up, stagnation, settlement of suspended matter, energy
use, life cycle costs, carbon footprint and sustainability.
In addition to all the likely technical problems engineers face when employing the trace
heating tape technique in large scale centralised systems, other matters must also be the
borne in mind:
Any outbreak of Legionnaires disease is normally sensationalised by the media, and in
today’s society’s culture of blame, engineers could well find themselves culpable
(corporate manslaughter is a constant fear).
The public are generally unaware that Legionella is a common inhabitant of water
distribution systems. The incorrect, generally held assumption is that Legionella is an
unwelcome invader of poorly maintained water systems and that negligence plays a role
in its presence29. This is of course a half truth, borne out by the fact that most outbreaks
of Legionnaire’s disease are indeed related to occasions where basic maintenance and
health and safety procedures have been neglected.
This is ironic, because too few public buildings have any diagnostic testing regime or
results available to alert staff that there is a problem. Diagnostic tests, because of their
high cost and alleged complexity, are not recommended in the present Health and Safety
29 Stout JE, Yu VL. (2001) Legionella in the hospital water supply; A plea for decision making
based on Evidence-Based Medicine. Infection Control Hospital Epidmiol. (22) 670-72
199 advice, unless a particular problem is suspected. Many would argue that in many
instances Legionellosis goes undiagnosed, and mortality is incorrectly attributed to
other causes29.
The key issue here is the question of whether the presence and colonisation of legionella
leads to Legionellosis.
Should engineers perhaps be concentrating their efforts on the source of the problem?
Should they seek to understand the true status of the incoming water supply, and what
implications this has on the particulars of any installation and how it is to be operated?
The “fit and forget” approach to water supplies, which many might argue is the norm
today, has immeasurable consequences. Therefore, should engineers, as standard
practice, attempt to reduce the amount of suspended matter, organic loadings and the
potential for scaling through additional on-site filtration and associated treatment
regimes? Should maintenance regimes encourage the checking and draining down of
sludge from the bottom of calorifiers more regularly than at six month intervals? Rather
than relying on heat treatment in isolation to offer protection at a cost, should engineers
alternatively adopt an “end of pipe solution” approach, where efforts are concentrated
on control and reduction, localised to areas perceived as high risk?
The use of UV, ultra-filtration, copper-silver ionisation, silver-hydrogen peroxide and
so on offer many capital and operational cost advantages, yet are least often utilised as a
preferred method.
There may well be an argument for localised hot water generation, with the acceptance
of plate heat exchangers (or similar) delivering almost instantaneous hot water. These
take up little space, reduce storage problems and can greatly reduce and simplify the
requirement for secondary pipework installation and maintenance. The cost of the
primary pipework is also lower since cheaper materials may be used, and the
distribution temperature could be maintained network wide with more confidence. In
addition turnover of water supply would be guaranteed, thus reducing stagnation and
associated problems. The supply could utilise a single mains (pressurised) cold water
supply, which could be routed externally, therefore limiting heat gains and maintaining
the quality again by guaranteeing the turnover.
Finally, until environmental cultures are performed routinely, hospitals in particular will
continue to experience Legionnaires’ disease with its attendant high mortality. The
disease will remain under-diagnosed and undetected unless diagnostic testing is carried
out more frequently. Only if it is known with confidence that colonisation levels have
increased can truly appropriate and cost effective, reliable disinfection measures be
recommended29.
200 III.5
A new hygiene system for
cold and warm drinking water installations
U. Petzolt1
(1) upetzolt@kemper-olpe.de
Gebr. Kemper GmbH + Co. KG, Postfach 1520, D 57445 Olpe, Germany
(2) mete.demiriz@fh-gelsenkirchen.de
Fachhochschule
Gelsenkirchen
University
of
Applied
Sciences,
D 45977 Gelsenkirchen, Germany
Abstract
It is well known that the circulation of hot water at a relatively high temperature keeps
the hot water installations hygienic. According to the European and German directives
on the quality of water intended for human consumption the operators of the drinking
water system have to observe the requirements for cold drinking water as well. The
drinking water should be wholesome and clean if it is free from any micro-organisms
and parasites and from any substances which, in numbers or concentrations, constitute a
potential danger to human health. The stagnation of water in the domestic installations
results in an unhygienic and unhealthy system. Therefore a circulation distribution unit
for cold and warm water sytems has been developed for a forced flow in every pipe of
the installations. This new unit is called “venturi distributor- dynamic” because it is
based on the principle of the Venturi nozzle in combination with a dynamic cartridge
technology. The innovative solutions of the hygiene system avoid stagnation and
combine hygienic subjects with energy saving for the whole drinking water system in a
building.
Keywords
Hygienic installations, cold drinking water, warm drinking water, quality of water,
water hygiene, stagnation, circulation, Venturi distributor - dynamic .
201 1. Introduction
Within the scope of the present work a new hygiene system for cold drinking water
installations was developed and tested. Flow measurements were carried out after the
first installation of the new hygiene system in a building including shops, doctor’s
surgeries and apartments for people needing special care. With the help of a newly
developed circulation distribution fitting based on the Venturi principle a forced flow
circulation was realized in sub plumbing systems when not being used as intended.
The described solution for drinking water cold was taking place in 2008. In 2009 a new
invention joined the solutions for cold and warm drinking water systems. The dynamic
distribution unit was born. This invention can be used for the distribution of cold and
warm water.
It is well known that temperatures within 30-45 °C lead to germination and to a biofilm
in the installations pipes. Therefore directives and guidelines recommend a minimum
temperature of 60 °C in a hot water storage system, 55 °C in hot water circulation pipes
(1,2) and a maximum of 25 °C in cold water pipes. Regarding the cold water pipes the
aim is to have a temperature below 20 °C (3).
Reserve for future attic extension: (x years of stagnation)
Heating filler pipe (x years of stagnation) Garden watering system pipe Figure 1 – Usual installations with stagnation in rarely used pipes in a system for
cold drinking water
202 Until now in houses and the public sector buildings like hotels, hospitals and doctor
surgeries etc. stagnated water in the pipes can be treated only by a manual or time
controled flushing. House owners or users do not have any idea about the risks of their
water system. The usual installations of a detached house (Figure 1) with the familiar
flaws which lead to stagnation in the rarely used pipes with all the risks particularly for
babies and elderly people.
A generally used method to optimize the installations is to loop through the rarely used
pipes. This method also has still flaws. Depending on the length of the pipe larger
diameters must be used when planning loops in case of different flow rates for e. g.
caused by pressure flushing units (Figure 2).
Reserve for future attic extension: (x years of stagnation) Loop to rarely used taps Larger diameters because of long tap connecting Larger diameters caused by long tap connecting pipes and different flow rates
Figure 2 – Improved installations with without any stagnation in rarely used pipes in a
system for cold drinking water
With the help of the innovative circulation distribution fitting hygienically safe
installations can be realized (Figure 3). Here every consumption in upper flats causes a
circulation of cold water in the rarely used pipes even if there is no consumption. The
lightly warmed and stagnated water is replaced with cold and fresh water.
203 A generally used method to optimize the installations is to loop through the rarely used
pipes. This method also has still flaws. Depending on the length of the pipe larger
diameters must be used when planning loops in case of different flow rates for e. g.
caused by pressure flushing units (Figure 2).
2. Experimental issues, cold drinking water system
Several components have been developed and combined to create a hygienic installations
system for cold drinking water distribution. The most innovative component of the system is the
circulation distribution unit, which works static or dynamic. The distribution unit-static- in
combination with other components were tested in the lab as well as in the first realized
installations. The first monitored drinking water sytem was in the building Martinus-Höfe in
Olpe (West Germany) (Figure 8). There the distribution units –static- were mounted.
2.1 Venturi circulation distribution units -static- / –dynamicThis static unit has been designed and dimensioned so that approximately 10% of a main flow
branches off to a circulation loop pipe (Figure 4). The action principle of the unit is based on the
Venturi nozzle technology. The minimal pressure difference between the main flow and the
return pipe of the branched loop causes a forced flow to all taps, for instance, in bathrooms. This
forced flow refreshes the water in the connection pipe and keeps the water temperature low so
that provision and maintance of drinking water quality in the drinking water system up to the
tapping point is certainly provided.
204 Venturi distribution unit Figure 3 – Installations optimized by the Venturi circulation distribution unit –
staticPressure difference ∆p
Venturi nozzle
Figure 4 – Circulation distribution unit- static-
205 Pressure difference ∆p Venturi nozzle with integrated cartridge Figure 5 – Circulation distribution unit- dynamic-in a low flow situation
The static distribution unit has been developed to a dynamic working unit (Figure 5).
It is nearly closed at low flow situation (5% main flow) and it opens at a specific flow
pressure when the main flow is increasing. The effect is that 95% of the water is
running through the branched loop, although there is low flow situation in the drinking
water system. So the new distribution unit –dynamic- has a better efffciency to avoid
stagnation.
Figure 6 Circulation distribution unit- dynamic- with a high flow situation
206 The distribution unit –dynamic- (Figure 6) is nearly full opened at high flow situation
(95% main flow). The dynamic pressure opens a cartridge and the effect is that 95% of
the water is running through the main branch and ca. 5% through the branched loop.
The 5% of huge flow is enough to avoid stagnation in the loop.
2.2 Other components of the hygienic installation system
The system includes the following components:








The circulation distribution unit combined with stop valves
Hygienic flushing unit with control valves
Ball valve with spring-reset servo drive
Timer unit
Tee temperature sensor valve Pt 1000
Vortex flow sensor
Drain with overflow monitor
Logic Control System incl. software and control modules for the sensors
By combining the sensors and the flow control units and valves intelligent installations
can be realized controlled by temperature, volume or time. For example, if the
temperature rises in pipes detected by the temperature sensor the control system
activates a hygienic flushing unit or a valve for a forced flow. Though it is simpler to
control the flow by the timer.
2.3 Measurements under lab conditions
Experiments done by Rickmann have shown that a pipe insulated according to the
regulations warm up in less than 3 hours depending on the environment temperature (4).
As shown in figure 7 the cold water temperature rises from approx. 13 °C up to 25 °C in
approx. 3 hours which makes a flushing to refresh the water in the pipe every 3 hours
necessary.
207 Temp.
Flow-
[°C]
rate
[l/min]
Time
Upper curve :
Middle curve:
Lower curve:
Lab environment air temperature
Cold water temperature in the pipe
Initiated flush water flow rate
Figure 7 – Time controlled flushing of a pipe (4)
2.4 Measurements at the first installations
For the first time this hygiene system was installed in a newly constructed building
called “Martinus Höfe” in Olpe, Germany. It is an extension of the St. Martinus
Hospital in the middle of the city. On the ground floor there are only retail trade shops.
Doctors cooperating with the hospital have their surgeries on the first floor. On the
second floor there are 21 apartments with nursing service. Elderly people who do not
need special care can hire apartments on the third floor. It was very important for this
building to have a hygiene system because the most of the users are elderly people
and/or have immunodeficiency. On the other hand the doctors’ surgeries and shops are
closed at weekends, thus the sanitary facilities are not used for 1-2 days. That would
inevitably lead to stagnation in the pipes. Another dangerous situation would be caused
by the apartments and shops which are not hired for weeks and months (Figure 8).
208 Figure 8 – Martinus-Höfe-Building
Flow measurements were done on the second floor in the bathroom of apartment
No. 7.38 after the valve No. 4 by a Fluxus ultrasonic flowmeter. The flowmeter was
connected to the circulation pipe as shown in figure 9 and 10.
Figure 9 – Principle of installations as in Martinus-Höfe-Building
209 Figure 10 – Ultrasonic flow measurement
The PEX circulation pipe size was 16 x 2,2 mm and it was aprox. 7 m long. Therefore
the water volume in the pipe was aprox. 0,75 liter. As shown in figure 10 every forced
flow by a pipe flushing causes a flow with a volume of 2,9 liter in the observed
circulation pipe. That means the water in the circulation pipe has nearly been refreshed
4 times. The water consumption in other apartments also causes forced flows. The
forced circulating water volume (without circulation water volume caused by the pipe
flushing at 6.37) between the two pipe flushes at 5.08 am and 8.37 am was 0,6 liters
(Figure 10).
210 1
0,9
Pipe flushing 70,3 liter
Flowrate at 14,54 l/min
Starttime: 5:08
Circulation Flow
Pipe Flushing
0,8
Pipe flushing 69,16 liter
Flowrate at 14,82 l/min
Starttime: 6:37
Flowrate l/min
0,7
0,6
0,5
Pipe flushing 60,11 liter
Flowrate at 14,26 l/min
Starttime: 8:33
0,4
0,3
0,2
0,1
0
5:00 5:10 5:20 5:30 5:40 5:50 6:00 6:10 6:20 6:30 6:40 6:50 7:00 7:10 7:20 7:30 7:40 7:50 8:00 8:10 8:20 8:30 8:40 8:50 9:00
Time
Figure 10 – Pipe flushing and circulation flowrates 07.06.2008
3. Drinking water systems –warmWarm water as a source of infections!
The warm water system takes on special significance, especially when operated as a low
temperature system (< 55 °C) for scalding-protection or energy savings reasons. The
slight entrance of germs from the public water supply or from other sources (e.g.,
construction, installation of devices, repairs) is unavoidable. At temperatures between
30 °C and 48 °C, an explosive growth can occur in the plumbing system within a few
days. Above all, lines in large buildings with often kilometre-long, stagnating lines and
high volumes of accumulated warm water are affected. More than 70% of these
buildings can be populated with legionelles.
Based on the described circumstances, the warm drinking water systems, operated
centrally through drinking water heating using drinking water circulation, need to be
hydraulically equalised. The circulation system is hydraulically equalised using
regulation valves. Supported by the KEMPER Dendrit CAD 5.4 calculation and
simulation software, the piping systems can be calculated and subsequently simulated in
order to obtain maximum planning security.
211 3.1 Drinking water -warm- systems with Venturi circulation
distribution unit -dynamicIn many big buildings like hospitals, hotels, schools etc. with central heating system for
the warm drinking water a distribution unit –dynamic- can be used to realize the
circulation system for drinking water warm. Instead of regulating valves in every single
room the distribution unit-dynamic- is setted. The function is the same as described in
figures 5 and 6, the temperature in the system is 60°C, the present temperature at the tap
and at the shower is 55°C.
Figure 10 Circulation distribution unit- dynamic- in warm water in a low flow
situation
Figure 11– Circulation distribution unit dynamic in warm water in a high flow
situation
212 3.2 Optimised drinking water circulation with energy and
economical benefits
The temperature presence is achieved solely through the drinking water warm
consumption lines, which are connected to a dynamic distribution unit. The functional
lines for the drinking water circulation are omitted in the area of distribution for the
drinking water circulation. Consuming drinking water warm during provision provides
the required volume flow (in the ring) via the dynamic distribution unit. If there is no
consumption, the circulation starts, driven by the circulation pump. The reduced
pipeline installation for the drinking water circulation and the surface reduction in the
pipework of the drinking water warm area provides up to 20 % energy savings in the
area of drinking water circulation standing losses.
Figure 12- section of a drinking water warm system, high temperatured in the
loop with distribution unit –dynamic- in warm water (Fig. 11, 12)
213 Figure 12 Figure 13 - Drinking water warm system and drinking water warm circulation
with minimised circulation lines.
An exemplary representation of a horizontal distribution system for a circulation
pipe system in a hotel, home for senior citizens or similar
214 3.3 Focusing consumption lines - Minimising circulation lines:
The goal of our drinking water warm systems should be to exclusively need piping
systems with functional lines for consumption (Figure 13). Additional lines in the
piping system, installed only to cover the “circulation” function, cause high energy
losses in the drinking water circulation area, meaning also investment and operating
costs that are higher than necessary. That allows saving up to 20 % of the standing
losses for the drinking water circulation.
The piping system consists up to 90 % of consumption lines (red); only 10 % are
circulation lines (magenta), which are needed solely for the “circulation” function.
By using ring-line conduction and the distribution unit -dynamic- the “consumption”
function and “circulation” function are merged, so pipelines for drinking water
circulation are omitted in the floors. That reduces the pipe area and energy requirements
for the drinking water circulation overall. The required circulation system is kept as
small as possible, converged shortly and, using suitable control technology,
hydraulically equalised. The circulation pump circulates the water through the supply
lines of the drinking water warm system with stable circulation.
4. Conclusions
The peresentated solutions for drinking water cold an warm systems are the core of the
KHS (KEMPER Hygiene System) technology.
The KHS can make a decisive contribution to maintaining drinking water hygiene in
new and existing buildings for both cold and warm drinking water. Each building, based
on its usage, is a “prototype”, so it can never be compared with another building of the
same design – it is always an object that needs to be individually considered. The use,
and accordingly the underlying operational use as intended, needs to be defined in detail
during planning, construction and operation. KHS technology demonstrates new
“innovative” ways for sanitary drinking water installations in the three areas of drinking
water hygiene, economy and ecology. Consistently implementing KHS means achieving
an additional milestone in the area of “health” and makes an important contribution to
responsibly dealing with our planet Earth.
215 DRINKING WATER HYGIENE
Maintenance of the parameters for drinking water according to the Water Quality
Regulation 2001
Use as intended operational water exchange in the drinking water system through ring
line systems and KHS valve technology
For all tapping points in the cold and warm drinking water system
KHS solutions enable temperature maintenance through consumption and circulation
processes
Establishment of use as intended operation in the drinking water systems
Automation processes enable establishment of “use as intended” operation at all times
ECONOMY
Controlled consumption of water in the building
Intelligent technology provides control, which saves energy and personnel expenditures.
Automatically monitor and document flushing measures
A “flush log” is created for flushing measures.
Reduce costs for water and flushing measures
KHS greatly reduces the “actually required water volume” to implement measures for
the water exchange as compared with the manually implemented flushing measures that
are conventional today. The manually implemented flushing measures on every tapping
point are replaced by the KHS-Distribution Unit and by automated flushing processes.
Due to the constant movement in the piping system, protective layers can build up that
prevent the formation of corrosion products and corrosion damage. The costs for repairs
and remedial measures in the piping system can be reduced.
216 ECOLOGY
Sustainably consume the resource of water and prevent unnecessary water
consumption
During operational use as intended, the KHS technology prevents stagnation through
sole consumption at the tapping points. No additional flushing measures are necessary.
Drinking water consumption becomes “sustainable”.
Provide drinking water “naturally” at the tapping points
The constant movement means drinking water hygiene can be realised without
additional water treatment throughout the drinking water system.
Energy savings in the area of drinking water circulation by lowering the standing
losses
When using inliner circulation in connection with KHS valve technology, up to 40 % of
the energy from standing losses can be saved as compared with “conventional plumbing
technologies”.
CO2 reduction and environmental protection
KHS significantly contributes to environmental protection: The energy savings potential
of KHS lies in the area of “sustainable water usage” and reduction of the expenditures
for drinking water heating. In addition, energy in the area of water provision and
personnel is saved. The energy savings mean saving fossil fuel, leading to a reduction of
CO2 emissions for the sanitary system.
A circulation distribution unit static or dynamic as well a hygiene system has been
developed for the purpose of a forced flow in the unused pipes of cold and warm water
installations. The system was successfully tested under real conditions.
The new system allows a simple provision and maintaining of the drinking water quality
in the drinking water installations up to the tapping point. Further benefits of the system
are






Prevention of stagnation
Use of installantions as directed
Reduced water consumption
Lower operating costs because of more efficient flushing
Lower maintenance cost
Less microbiologically induced corrosion in copper pipes
217 4 References
5. van der Schee, W. G. (2005) Regulation on legionella prevention in collective water
systems. Proceedings – 31st International Symposium CIB W062 Water Supply and
Drainage for Buildings. Brussels, Belgium: BBRI
6. VDI 6023, (2006) Hygiene for drinking water supply systems, Berlin: Beuth-Verlag
7. Umweltbundesamt (Federal Environment Agency of Germany)(2006) Trink was –
Trinkwasser aus dem Hahn, Dessau, Germany: UBA
8. Rickmann, B. (2007) Unpublished experimental data, FH-Münster University of
Applied Sciences, Steinfurt, Germany
5 Presentation of Authors
Ulrich Petzolt is the head of the product management department
for plumbing fittings at Gebr. Kemper GmbH + Co. KG, Olpe and
is a former student of Fachhochschule Gelsenkirchen University of
Applied Sciences, Department of Utility Technologies. He is
specialised in water saving and hygiene, water hydraulics.
Mete Demiriz is a professor at Fachhochschule Gelsenkirchen
University of Applied Sciences in the Department Utility
Technologies where he is the head of the research and
development lab of sanitary technologies. He is specialised in
water saving and hygiene, water and waste water hydraulics and
water support of special buildings.
218 Session IV: Water conservation
IV.1
WATER CONSERVATION &
QUANDARIES
Dr. Lawrence s. Galowin
lgalowin@nist.gov, larrygales@earthlink.net
I. ABSTRACT
Newest United States endeavors for achieving diverse water conservation thrusts with
many elements have advanced rapidly. Efforts extend to many diverse concerns for
irrigation, building grey water-reuse, recovery/utilizations. Advances extend to
evolutionary plumbing technologies/elements also into environmental aspects, climate
change, “greening the environment” all with impacts on energy conservation. New
perspectives relate to U.S./worldwide climate change portents/perspectives. Economic
stimulus measures have relevant water conserving activities, as priority needs and an
EPA report draft focuses on climate/weather issues and concerns. Water-Sense/EPA
thrusts include water-conserving labeling on water saving products (Energy Star model)
for consumer information. Emergence of the Alliance for Water Efficiency organization
brings many organizational interests for national water saving into reality. Water-Sense
thrusts extend to select standards in usages for reduced consumption. Some required
elements have not achieved expectations fully in implementations. Potentials may have
negative impacts for devices/fixtures due to inadequate research or testing that possibly
sets “faultily set standards”. Need for in-depth and comprehensive research appears
necessary - but unfulfilled - to assure elimination of problem potentials (e.g., waterless
urinals usages). Provisions for indicated specifics have disappointments in certain
applications that are illustrated in this study for water savings and needs for essentials.
219 Keywords
Water conservation, water closet tests, plumbing and test methods
1 Introduction
The strengths/achievements for
extensive efforts in the United
States (US) advancing water conservation developed from Environmental Protection
Agency (EPA) with evolution of the Alliance for Water Efficiency. Many aspects in
realities have significance for projected endeavors; example (modified) follows:
WaterSense Product CertificationEPA requires all products bearing the WaterSense
label to be independently certified. This certification provides consumers with
confidence in both the water efficiency and performance of WaterSense labeled
products.
To ensure that WaterSense labeled products meet specific efficiency and performance
criteria, EPA has released its final WaterSense product certification system (PDF) (14
pp, 91K). This certification system outlines the process EPA requires all products
bearing the WaterSense label to be independently certified. This certification provides
through April 1, 2010. Certifying bodies 2009, although EPA will be transitioning To
ensure that WaterSense labeled products meet specific efficiency and performance
criteria, EPA has released and procedures for the product certification and will
supersede the interim certification and will supersede the interim certification process
outlined in Appendix A of the program guidelines (PDF) (42 pp, 283K).
The WaterSense product certification system becomes effective on April 1, To ensure
that WaterSense labeled products meet specific efficiency and
A of the program guidelines (PDF) (42 pp, 283K).
through April 1, 2010. Certifying bodies not currently licensed by EPA should be
accredited directly to these requirements2009, although EPA will be transitioning from
the interim certification process performance criteria, EPA has released its final
WaterSense product certification system (PDF) (14 pp, 91K). This certification system
outlines the process and procedures for the product certification process outlined in
Appendix its final WaterSense product certification system (PDF) (14 pp, 91K). This
certification system outlines the process from the interim certification process not
currently licensed by EPA should be accredited directly to these requirements
consumers with confidence in both the water efficiency and performance of WaterSense
220 labeled productsand procedures for the product certification and will supersede the
interim certification process outlined in Appendix A of the program guidelines (PDF)
(42 pp, 283K).
The WaterSense product certification system becomes effective on April 1, 2009,
although EPA will be transitioning from the interim certification process through April
1, 2010. Certifying bodies not currently licensed by EPA should be accredited directly
to these requirements.
EPA requires all products bearing the WaterSense label to be independently certified.
This certification provides consumers with confidence in both the water efficiency and
performance of WaterSense labeled products.
What does this transition mean for you?

Consumer, Manufacturer, Certifying Body , Accreditation Body
Consumer
Consumers should see no noticeable change. Products will still be certified by an
independent third-party certifying body, only now EPA can:



Ensure consistent application of its minimum product certification requirements.
Establish uniformity in the certifying body accreditation process, while making
the process open to all qualified accreditation organizations.
Provide fully transparent criteria for product certification and the accreditation
of product certifying bodies.
Manufacturer Beginning April 1, 2010 manufacturers will have to obtain all
WaterSense
related product certifications from an EPA licensed certifying body that is accredited to
provide certification services for WaterSense. EPA will post an updated list of its
certifying bodies as they are licensed under the final certification system.
EPA anticipates that all of its current licensed certifying bodies will obtain this
accreditation for WaterSense. If, for any unforeseen reason, a licensed certifying body
chooses not to seek accreditation, EPA will work with the affected manufacturers to
transition their certifications to another licensed certifying body.
For manufacturers with existing WaterSense product certifications, this transition
should require very little operational change.
221 EPA does not require notice from its manufacturer partners for this transition unless the
manufacturer switches its product certifications to another licensed certifying body. If
this is the case, please notify the WaterSense Helpline and fill out and
submit a new certified product notification form with the updated certification
information
222 Certifying Body
Effective April 1, 2009, certifying bodies not currently licensed by EPA should be
accredited directly to the final WaterSense product certification system (PDF) (14 pp,
91K) requirements. EPA will post a list of accreditation bodies to begin this process.
EPA has built in some transition time for those licensed certifying bodies currently
offering certification services for WaterSense. They may continue to operate according
to the interim certification process, but between now and April 1, 2010, will have to
transition to and obtain accreditation in accordance with the final certification system.
To be licensed, for a particular product specification, certifying bodies must extend their
scope of accreditation for each WaterSense product specification.
Upon accreditation, please contact the WaterSense Helpline for application procedures
and to obtain a copy of the licensing agreement.
Accreditation Body
EPA has opened up the application process for all accreditation bodies interested in
providing services for WaterSense. Any accreditation body meeting the final application
and approval criteria may apply at any time.
To qualify under the transitional criteria, accreditation bodies must submit an
application to EPA no later than April 31, 2009, and show continual progress toward
meeting the final application and approval criteria. All accreditation bodies must meet
the final application and approval criteria by April 1, 2011.
To apply either under the final or transitional application and approval process, please
submit an application letter to the WaterSense Helpline, as described in the product
certification system (PDF) (14 pp, 91K).
Technical Information
Are you a manufacturer or other party interested in finding out more about the
development of the final WaterSense product certification system? To learn more about
the certification process, including the proposed draft certification scheme, public
response to the draft requirements, and EPA's response to the public comments, please
see:
Final WaterSense Product Certification System (PDF) (14 pp, 91K)
223 





Response to Public Comments Received on May 2007 Draft WaterSense
Certification Scheme (PDF) (33 pp, 161K)
Comments Received on the Draft WaterSense Certification Scheme (PDF) (33
pp, 151K)
Presentation From June 20, 2007 Public Meeting (PDF) (29 pp, 196K)
Summary from June 20, 2007 Public Meeting (11 pp, 66K)
Draft WaterSense Certification Scheme Cover Letter (3 pp, 100K)
Draft WaterSense Certification Scheme (PDF) (22 pp, 124K)
2 Open Ends and Issues
Projects have achieved useful results although insufficient time has elapsed for
impacts/end results that have anticipated benefits and satisfactory outcomes.
Manufacturers speedily offered lower desired water flows but possibly without in-depth
substantial trials/testing assurance. Reduced faucet flow rate (1.5 gpm instead of 2.5
gpm national requirement) leaves unreported openly available report findings for
fulfilling user needs. Pursuit of waterless urinals installations remains questionable but
flaws in adopted requirements hangs over that fixture (no surface wash – odors, residual
and potential contaminant/health hazards from sputum or vomit deposits, permits odors
and surface drying into room). Details for utilizations are required but not supplied for
engineered methods, retrofit/design guides, varied connected elements that remain
inadequate. Confusion exists from three WC standards (ASME, CSA/ASME
Harmonized, EPA WaterSense) with differing and uncertainties for adequacy of test
extraction materials and drain transport. However, from endeavors of anticipated
successes there remains uncertainties since nowhere are means for long-term results
evaluations planned. Examples follow (modified format) of announced “too late”
changes (lacks technical practice guidance - request to authors was unfulfilled).
High‐Efficiency Toilets in Non‐Residential Applications
A Caution for Water‐Efficiency Practitioners, Design Professionals, and Facilities
Managers
October 2008
By Bill Gauley and John Koeller
There are two primary concerns related to the proper operation of toilet fixtures ‐
flushing performance and drainline carry. While it is fairly easy to define an adequate
level of flushing performance (we want the toilet to remove all of the waste from the
bowl in a single flush), it is far more difficult to define how far a toilet flush should
transport this waste through the building drain piping to the sewer. There are many
224 things to consider when attempting to assess drainline carry performance, e.g., how
much waste is being flushed, what type of waste is being flushed, what is the diameter
and slope of the drainline, what is the age and physical
condition of the drain, etc.
The current ASME/CSA drainline carry testing protocol not only uses non‐realistic test
media (¾‐inch plastic balls), but it also uses the same testing procedure regardless of
whether toilet fixtures are residential or commercial models. Therefore, we believe that
this protocol fails to adequately address the concern of real waste movement in the drain
system. This is somewhat disconcerting since we know (a) that there is a clear
difference between ¾‐inch balls and “real” waste and (b) that there are significant
differences between residential and commercial toilet installations. For example:
• Commercial fixtures are often installed on 4‐inch diameter drain pipes set at a
1‐percent slope whereas residential fixtures are typically installed on 3‐inch diameter
pipes set at a 2‐percent slope
• Commercial toilets, which are often required to flush paper toilet seat covers, paper
towels, large amounts of toilet paper, etc., are typically subjected to a much
greaterwaste loading than residential toilets
• The lengths of drain runs are often much longer in commercial installations, and
• Supplemental flows are often much less in commercial installations (supplemental
flows from bathing, clothes washing, etc., help transport waste through drainlines).
A significant number of testing programs have shown a direct correlation between
drainline carry distance and flush volume. As such, we know that if we continue to
reduce toilet flush volumes without making commensurate changes to the
plumbing/drain system, we will eventually begin to experience blockages in drains. This
is a fact!
While we are relatively confident that high‐efficiency toilets (HETs) are suitable for use
in residential applications because of the smaller diameter drain piping, steeper slope,
and availability of supplemental flows, we are not nearly as sure about the use of HETs
in all
commercial applications. We know by experience that 1.6 gallons (6 litres) appear to be
sufficient to satisfactorily transport waste in commercial sites, and we know by
experimental data that there is some flush volume less than 1.6 gallons where we will
begin to experience problems. The problem is that we don’t know what that lower
volume is. What’s more, we want to avoid a situation where the flush volumes of
commercial toilets are reduced to the point where drainline problems begin to occur
with frequency only to have to ‘take a step back’ and increase the flush volume back to
225 1.6 gallons (6 litres). There are significant differences between residential and
commercial toilets – both in their physical construction and operation – there is no
inherent reason that they should be required to pass the same flush performance /
drainline testing requirements or that they should necessarily flush at the same volume.
While we are not saying that installing HETs in commercial installations will
necessarily lead to drainline plugging and problems, we are suggesting that caution be
used when choosing whether or not to install HETs in existing commercial sites – especially if the physical conditions of the existing drainlines may be suspect or if there
are little or no supplemental flows available at the site to augment the carry of waste. In
new construction, we recommend that building designers seriously consider (a) the
placement of water consuming fixtures (e.g., lavatories and flushing urinals) upstream
of the HETs and (b) providing for drainline slopes of greater than 1‐percent.
Significant elements for engineered systems were not provided for design/installation
and is necessary for all . For replacements of existing fixtures in renovation, projects
existing drains may be of pitch and diameter that could be undesirable and not fitting as
the description indicates. No engineered advisory appears and does not even suggest
concerns that would deter failures!
2.1 Commentaries – other select sources
Personal communications on shared concerns on water saving fixtures inadequacies
(excerpted here) illustrates the need for establishing performance criteria that are
sufficient and practical. Intensive testing/research and timely trial experiences prior to
advocacy for general installations/purposes is necessary.
“ ….. comparison created in 2006 to compare Falcon and Duravit urinals. ..
The conclusions stated at the end … particularly important in that at present all of
the waterless urinals contribute to contamination of the existing discharge piping.
…have observed this contamination to varying degrees depending on the urinal
manufacturer.
We began installing waterless urinals about 4 years ago .. rebate funds were
available from our local utility. … tried every manufacturer of urinal available at
the time .. concluded that it was more economical to retrofit with 0.5 gpf water type
urinals.
Retrofitting from existing 3+ gpf urinals to 0.5 gpf urinals over a five year period is
more economical than the cost of replacing waterless cartridges, sealing fluid,
226 cleaning, and maintenance of the waterless urinals…. had to replace cartridges
after 1200 to 2000 uses not the 7000 predicted by some manufactures.
Also the dual flush toilets we have in one building are the Caroma "Walvit". Due to
some concern, the construction of the bowl provides rather steep sides with a small
water spot. This tends to allow excrement to cling to the sides requiring additional
flushes for a sanitary appearance …
John Leaden, Energy Management Coordinator, University of Washington, Seattle,
WA
Another:
Cost projected comparisons (abbreviated) for three urinal types for 2004 to 2008
showed water free fixtures required considerably greater expenses - Falcon
Waterfree, 0.5 gpf, 3.5 gpf urinals:
Year Item
Cost
Falcon 0.5 gpf 3.5 gpf
2004 Cartridge & sealing fluid $589.66 $0.00 $0.00
$0.00 $0.00
water / sewer @ $7.00/CCF $0.94 $89.85 $628.92
$628.92
2008 Cartridge & sealing fluid $589.66 $0.00 $0.00
$0.00 $0.00
water / sewer @ $11.00/CCF $1.47 $141.19
$141.19
5 year totals
Cost
Unclog Labor
Cost
$0.00
$590.60 $89.85
Unclog Labor $0.00
$988.31$591.13
$2,954.32 $577.59 $4,043.08
Estimate monthly cartridge & sealing fluid replacement @ $39.95 + stores markup
and inventory costs > $589.66/yr
10 Gal water for flushing per application x 12 months @ annual rate. Unclog labor
costs for Falcon urinals have been infrequent or minimal to date. Water usage costs
estimated for 1600 flushes per month
227 Diminishing acceptance grows and rejections for water conserving action:
Larry,
Many thanks … same day the magazine came out we received an addendum to a
project we were bidding on at Fort Stewart Georgia. …. stated that Ft. Stewart no
longer accepts water free urinals on base. Existing waterless units were to be
removed due to "high maintenance costs and offensive odor problems”
Here in Virginia, the City of Roanoke is removing all the waterless urinals in the
city limits… intend on contacting and interviewing … principals involved with these
decisions. .. that kind anecdotal first-hand knowledge could go a long way.
…remember … info about Washington State University … removed all their units as
well.
.. water saving mistakes are going to wear out their own welcome before too long.
Stay Tuned, Mike
Another waterless urinals study showed extensive depositions in drains (1). Attempt to
overcome such formations appears in Australian
standard (2); that requires upstream basin fixture
so that waste flow water into drain (apparently)
dilutes and/or carries away such chemical
formations. An important aspect for newly adopted
plumbing fixtures into localities, that usually
require (by reference) existing codes and
standards, coupled with almost no depths for engineered plumbing systems applications,
challenges remain high for acceptance(s). Standards without adequate depths of
technical bases, or substantiated correlations from incidents/data, or derived
fundamentals from detailed studies are perilous! Research input and/or portents of
impacts are not necessarily required considerations in adoptions and implementing
requisites or convincing evidence without performance validations for many endeavors
in retrofit actions.
Water Sense requirements for WCs extraction/drain transport with bean paste
cylindrical solids tests indicates excellent extraction/transport (perhaps it breaks apart).
The ASME national plumbing standard (3) allows partial waste simulants extractions
of mixed sponges and tissues (lesser total volume compared to bean paste solids and
pass criteria allows partial extractions. Conclusions may result in unperceptive
circumstances - really little merit!
228 3 Technical challenges – highlights, issues
Solid wastes extraction standards test requirements do not simulate installed water
closet installation practice with a turning fitting into a pitched horizontal drainpipe. A
basis for Water Sense from research basis (5) indicated “.. flushing systems better at
clearing media from the bowl {no drainline} – MaP testing – are not necessarily same
systems that carry waste furthest in the drainline. … apparent that … no strict
correlation …” has not been challenged. Note – that has apparently not influenced
testing acceptances or practices established in the programs for water savings. Such
procedures for test materials and applications in conflict or neglected/ignored and
incongruent/unrealistic requirements remain acceptable despite such contrary findings.
Currently, the tests require a mounted WC vertically assisted gravity acceleration
of solids from the WC outlet into open-air discharge. That condition assists
clearances but neglects actuality that very different chaotic mixing into/from entry
turning fitting (4) occurs.
Knowledge of partial solid waste extractions allowed in the ASME A112.19.2 water
closet standard applies even for HET fixtures (since both tests are required by
procedures).Whether plumbing professionals or built systems practitioners are aware
and/or concerned is another uncertain aspect of water conserving advocacy. An
additional need is for label information of test results (simple manner) that would
indicate performance levels; it would provide useful ratings from testing (similar to
energy type labeling for energy consumption).
4.1 Critical transport – selected
details
Newest thrusts for high efficiency toilets
(1.28 gal) have achieved significant attention but data and issues needs are unfulfilled
on testing and functional adequacy for transport in drains. (Interests have extended to
discussions of ‘dry drains” - whatever that may mean!). Concern for such issues led to
select study/comparisons. Initially, request was to Dr. M. Gormley for DRAINET
computer program calculations (familiar to CIB W 62). The assumed WC profile set
three solids locations in the rise, peak and fall locations shown. Results for computed
transport in plastic drains from the assumed profile indicate distances in repeated
flushes. First flush had solids in the profile while subsequent flushes were without
solids. Shared data sent to Dr. Cummings for feedback comments provided that
although different solids and pipe slope conditions were applied in his tests the
229 determined transport was “in the ballpark” - indications of agreement.
transport data determined many different conditions, sample results are:
Computed
SUBSEQUENT FLUSHES HET TYPE WC
SG > 1 & SG = 0.98 Carry Distances
100 mm pipe 1%
slope 4 subs
flush
16
14
12
Distance Carry (m)
(a) Solids lighter then water and neutral
buoyancy in repeated flush conditions (100
mm diameter, 1% slope) indicates buoyancy
increases transport distance and extends into
each repeat run for both densities.
18
10
100 mm pipe
Slope 1% 4 subs
flushes Solid SG
0.96
8
6
4
2
HET TRANSPORT 75 mm Diam. 2% Slope - 4 SUBSEQUENT FLUSHES
75
10
0
12
0
14
0
16
0
18
0
20
0
22
0
23
0
24
0
35
43
65
25
29
12
6
16
20
tim
e
0.
8
1.
6
2.
6
3.
6
4.
6
75 mm
2% Slope
4 sub
flushes
8
10
0
50
Time (sec)
Transport Distances (m)
40
(b) Solids with neutral buoyancy indicate
similar results (75 mm diameter, 2 % slope)
that repeated flush conditions extend the
transport distances for each added flush.
30
20
10
43
65
75
10
0
12
0
14
0
16
0
18
0
20
0
22
0
23
0
24
0
25
29
35
12
16
20
6
8
10
1.
6
2.
6
3.
6
4.
6
8
0.
tim
e
0
TIME (sec)
4.2 Waste transport phenomena
Recently, Dr. Steve Cunnings/Australia (6) provided test data for waste transport with
five differing solid waste material loadings, and controlled or regulated WC discharges
(partial reproduction chart follows). Averages from repeated tests are displayed; no
indicated ‘error bars’ shown. Controlled WC discharges (gal/sec) varied flow from the
WCs (no comments on profiles). Indicated solids identifications are on the strip and
colors show the first and second flush results (difficult to observe). Unknown influences
on flush profiles (limitations) do not appear.
Procedures followed Australian standard requires open to air vertical extraction tests
(gravity pull -no turning fitting)) for all solids to be extracted within one-third of the
total volume of water discharged. ASME A112.19.2 plastic ball media (shown for no
apparent utility for wash down WCs tests by Gauley) appeared not to be useful. Possible
breakup type also carried solid materials out the 180 ft drain. An inference - solids that
could not retain the original configuration and broke into smaller pieces may float
and/or carry in minimal trace flows. Uncased materials seem to extend transport out
end of test drain; solids that could not retain original configuration, may not retain
volumetric integrity or stability that appears to result in great carry distances. By
inference, those possibly float and/or carry with minimal trace flows. Issues exist for
test materials that break up; volumetric stability may be required for standards
replications purposes. No discussions on influences of flush profiles (limitations) have
been discussed.
230 ASME A112.19.2 Task Group - Toilet / Drain Line Carry Proposal
Evaluating Different Flow Rates of 0.55gal/sec, 0.4gal/sec and 0.26gal/sec with Volumes of 1.6gal,1.3gal,0.8gal, 0.5gal, 0.3gal using
Australian Test Media, Polypropylene Ball Media, MaP (latex covered), MaP (modified case) & MaP (uncased) with 4" (100mm) Pipe @ 1.67% Grade
Flow
Flow RateRate
g
7% Grade
MaP MEDIA
(Uncase)
1st Flush
1 .6 g a l ( 6 L )
0.55gal/sec (2.1L/sec)
2nd Flush
BALL MEDIA
(Polypropyle
0.4gal/sec (1.5L/sec)
1st Flush
2nd Flush
0.26gal/sec (1L/sec)
MaP MEDIA
(latex cover)
1 .3 g a l ( 4 .8 L )
Note: No further distance achieved with 2nd Flush
1st Flush
0.55gal/sec (2.1L/sec)
2nd Flush
Note: No further distance achieved with 2nd Flush
AUS MEDIA
0.4gal/sec (1.5L/sec)
1st Flush
2nd Flush
0.26gal/sec (1L/sec)
0.55gal/sec (2.1L/sec)
.8 g a l ( 3 L )
u m e ( g a l)
MaP MEDIA
(Mod Case)
1st Flush
2nd Flush
0 4 l/
(1 5L/
)
Drainline distances first and second flushes for differing WC discharge rates are:
Carry Distance (ft)
1.6 gal Flush
0.55 gal/sec
0.4 gal/sec
0.26 gal/sec
48
44
39
78
54
23
37
34
1.3 gal Flush
43
29
36
43
32
23
18
16
12
231 4.2.1Additional Observations
From actual installed rest rooms fixtures (male and female populations) research test
series BOKOR showed observed/instrumented variations of solids. Observed WC actual
usage materials discharges indicated mixes of waste materials for solidity and waste
pieces plus other user disposed materials. Further tests could indicate how discharge
profiles distributions (with and without solids) provide optimal solids extraction with
extended transport results; further investigations (with selected types of test media)
appear necessary.
4.3 Waste transport in cast iron and plastic drains - impacts
Information from related extraction/transport information on extraction and transport
from WSSC data (7) applied for these comparisons. Examination of many 1.6 gal WCs
had test conditions that varied with light and heavy test loadings - of natural and
synthetic sponges and baby wipes. Tests had light (20) and heavy (30) pieces. Drains
were 3 and 4-inch plastic pipes at 1% and 2% slopes with appropriate entry turning
fittings. Final few tests conducted comparisons with cast iron pipe/fitting – note that
most tests reported from almost all facilities use only plastic pipe pipes. Some
additional tests with cast iron drain/fitting for “rougher walls” than plastics
demonstrated significantly different results. Introduction of retrofit WCs for
conservation, rehabilitation/renovation programs and new construction other than
plastic piping materials may apply.
Comparison of a 3.5 gal flush WC
Heavy Loads: WC#17 - 14
indicated total waste solids extraction
Sinker/Natural Sponges, 12 Floater/
with plastic drain and essentially
Sponges, 4 Baby Wipes (30 total)
complete extraction with cast iron drain.
Cast iron results in loss of transport distance
WSSC WC# 3; 3.5 Gal, 4" D. Pitch 1% , Plastic & Cast Iron
Series
and bout total extraction. Note the influence
2
Series
from the pipe/fitting may appear as great and
3
Series
4
perhaps (surprisingly upstream influence).
Series
5
Series
This aspect of feedback upon the WC is
6
Series
contrary to almost all assumptions of
7
downstream effects upon the performance
characteristics of WCs. This needs further
study and tests to become an element of
future standard testing protocols.
80
70
Carry Dist (ft)
60
50
40
30
20
10
0
1
3
5
7
9
11 13 15 17 19 21 23 25 27 29 31
Solid Test Matl Number
Series 2,3, 4 cast iron pipe Series 5,6,7 plastic pipe 232 Comparison of a 1.6 (nominal) gal flush WC that was able to replicate total extractions
with plastic fitting/pipe indicated incomplete extraction capability.
Performance demonstrated inability for fully extracting the same sets of sponges/baby
wipes with cast iron fitting/pipe.
WSSC Heavy Loads Tests #17 (Run #2 0f 3)
Assumption that WC flushing is
1.7 gal, Plastic & Cast Iron Drains, 4" Diam. 1% Slope
unaffected by impacts of installed
elements represents sufficient
concern for detailed further
investigations. If those test
experiences are correct, then,
detailed
considerations
of
potential installation conditions
(existing or new) becomes a
necessarily new aspect for testing
protocols in acceptances. A need also exists for special labeling, or other identifiers.
Lowered flush volumes are a high priority since WCs replacements are essential to all
aspects for conservation programs.
30
NtrlSpngsSnkrs
Cstirn
S olids Transport D istances (ft)
25
SynthSpngsFltrs
CstIrn
20
BbyWips CstIrn
15
NtrlSpngsSnkrs
Plstc
SynthSpngsFltrs
Plstc
10
BbyWips Plstc
5
0
0
5
10
15
20
25
30
35
Individual Solid (Arbitary #)
4.3.1 Additional aspects/factors
The hydraulic solids models is shown with surround water conditions frequently
assumed; modeling for transport generally follows similar construct. Hydraulic model
with surround water conditions frequently assumes following characteristics:
- Wall friction lessened (lubricant water)
- Water dam at trailing part of solid
- Lessened water depths at forward part of the solid
(flow about solid)
- Differences in depths cause driving force in direction
of motion
- Lessened depths occur with transport along the drain
- Declining depths as distance along drain increases; water eventually trickles down
to almost no flow condition and leaves the solid at rest on pipe wall.
233 4.3.1.2 Influences - turning fittings
Turning fitting effects at entry into the
horizontal drain in female pads tests
showed initial deceleration in the
turning section, followed by small
recovery into the sloped drain, then
usual deceleration along the pipe wall
by Waklin & Swaffield. Comparable
simulant human wastes tests are not
available. Effects of diverse solids
mixing (chaotic mixing at entry for flow/solids) unknown manners or interactions have
been observed but no details appear available. There exists a gap and an urgent research
need - those entry conditions into the turning fitting and drain have significance for
subsequent solids order and motion along the pipe.
4.3.1.3 Influences – washdown impacts
Wash down characteristics in new WCs provides forcibly clearing flush from the bowl
surround and (possibly) other trapway priming. Flushing action may serve intended
purposes; however, indications of inadequacies for all purposes, e.g., clear bowl wall
streaks, or solids attached to wall, and thorough spot removals from all exposed bowl
inner surfaces. Current practices with lined marker pen as a marker may not adequately
test such potentials. Different ‘sticky materials’ have found use and also spot methods
applications onto surface locations in testing applications experiences need review and
tests to appraise such needs.
5 Data comparisons for waste transport in drains
WSSC tests were with as received fixtures, water volumes varied widely and data
differed; outliers’ from actual test WC consumption records were dropped Transport
WSSC 3" Diam. , 2% Lites (20 Solids)
Variability of WCs Transport
Solids Trnspt (ft) 1st Flush
WSSC & Cummings (1.67%)
12
13
WC Number
234 %
/2
%
%
/1
%
%
/2
Pipe Daimeters & Slopes
4"
11
.6
7
10
/1
9
4"
8
4"
7
3"
6
3"
5
3"
4
/1
25
3
.6
7
35
%
45
65
55
45
35
25
/1
55
Trnspt (ft) 1st flush RJCT OUTS
T ran sp [o rt
d istan ce (ft)
Transport Dist. (FT)
65
data comparisons for impacts of pipe diameters and slopes so that an assembly sought to
indicate a correlation. The referenced Australian tests were over differing flush rate
ranges (known settings of WCs) hence an averaging of data was required on how to
apply such data. Those results appear in the figure and display trends/levels anticipated.
Smaller diameters and greater slopes known to increase transport distances for carry
along the drain (computer developments, dynamic testing at National Bureau of
Standards (now NIST) forty years ago, Heriot-Watt and other test stations). The figure
shows results for effects of drainpipe diameters and slopes. Anticipated increased
transport indicated is due to greater pipe pitch and smaller diameters. Unanticipated
information from the second flush of 1.6% slope transport data indicates large distances.
The reasons need identification of elements that may contribute to such effects. Those
considerations may include:
a. Solids on inner pipe wall surface (and/or contact between solids) create uniquely
spacing favorable to surround surge water abetting transport condition;
b. Discreet shape variations due to enclosing flexible materials have unusual
benefits for water dams and movements;
c. Tissues selection materials widely alter, when wetted/torn, compressed and/or
sticking to solids, as attached formations, to other materials alters cross-section
areas extensively.
d. Distributions of solids on pipe wall unknown so extent of each solid final location
are unknown governing influences.
Large distances for waste solids from the Australian series needs attention. Since those
solids had different characteristics than sponges (almost cubical) the combination of
shape, dimensions, and especially overall length, may all contribute to those
determinations. The integrity of the
WC Transport - WSSC (1% & 2%; Cummings @ 1.67%)
sample materials, overall solidity (or
All 1.5 gal Flush
breakup), flexibility, and shapes require
further detailed information. From added
150
testing of sets of varied solids, it may be
130
feasible to distinguish necessary
110
90
characteristics for understanding impacts
70
for governing parameters. Furthermore,
50
those efforts would assist in setting
30
alternative test simulants useful for
10
objectivity in standards testing.
3"/1%
3"/1.67%
3"/2%
4"/1%
4"/1.67%
4"/2%
WSSC Lite Slds WC DATA AVG Ft Trnspt (ft) 1st flush
WSSC Lite Slds WC DATA AVG Ft Trnspt (ft) Increemnt 2nd flush
T A N S P O R T (ft)
WSSC Lite Slds WC DATA AVG Ft Trnspt (ft) Total 2nd flush
Pipe Diameters and Slopes
235 5.1 Comments - test materials, different waste transport media in
drains
Efforts for selections of test materials in applications and modeling transport are
required to understand results with clearly differing outcomes. Needs requires attention
to rigidity, shapes, retained volume/mass and flexibility appropriate to solid waste
simulations. Media representations that differ from the circular cylinder model now
applied for computer transport modeling requires expansion options. Contours that
govern contact areas with the pipe wall, irregular shapes, and other shape influences that
impact on bypassing flows around the body, the depth to width ratios and essentially a
mix of differing lengths for mixed media properties.
From WSSC solids’ test data indicated many solids that are almost in contact with each
other may have a major effect for transport for all (second or later) flushes. Those may
present an “artificially large single mass” for water transport in a following flush.
Gormley reported another aspect of nearby solids causing “accordion type” movements
between close solids in drain transport in his dissertation). Those interactions can alter
transport distances overall and may differ for differing size solids. Larger solids with
other characteristics, and/or do not stop in very close proximity to each other in tests
causes other issues of concern for data assembly/averaging for overall determinations.
5.1.1 Other concerns in testing
Understanding initiation of WC flushing action becomes more important as attempts to
reduce consumption in conservation efforts occur. WC modes of operation starts with a
stored tank or flushometer valve supplying inflows; discharge initiates with rim entry
channels from peripheral outlets into the bowl and/or priming jets into the trapway.
Siphon conditions from bowl into trapway and delivery into turning fitting connection
to the pipe drain. Outflow of bowl contents into the connected discharge connection
fitting occurs with considerably chaotic entry conditions into drain.
Typically, WC discharges appear with rapid rise to peak rate followed by a decay
trending from the peak(s). The formation or generation may result from the inherent
condition of decreasing head in the tank (rate varies with square root of head), losses
due to friction/mixing within contoured bowl, and/or siphon entry into trap with
236 elements of siphonic extraction (do solids alter such occurrences?) into drain fitting; that drives (crudely) the “drainage system”. Influences from waste solids, tissues,
tampons,
etc. if any, are unknown to the author (except from
Hoover Codes of 1920’s with limited
information). Do typical discharge curves
with flushometer valves of essentially
constant pressure supply have the same or
similar characteristics?, Recent test data for
flushometer WC pressure supplies indicate
constant pressure conditions maintained
values in the supply lines (shown). How such entry condition is altered (if that does
occur) becomes a first step in understanding the remaining developments in the bowl
and outlet. To what extent does the siphon with waste solids have altered
characteristics? Many other unknown questions require investigations.
Supply Line Study, #7
Wall Mount Wall Outlet WC
90
45
80
40
70
35
#7, 35psi
#7, 80psi
#7, Flow @ 35
#7, Flow @ 80
50
30
25
40
20
30
15
20
10
10
5
0
Flow (gpm)
Pressure (psi)
60
0
0
0.9
1.9
2.9
3.9
4.9
5.9
6.9
7.9
8.9
9.9
Time (s)
6. Conclusions and recommendations
Achieving complete wastes extraction with minimal water consumption remains a target
for research and development. In-depth testing requirements were noted in several parts
(testing methods, solids simulated test materials, standards needs, practicum for
utilizations and methods). The ranges of such determinations remain a necessity for
activities from research, field test periods for practical outcomes of verifications or
needs for revisions.
The quandary of what is required for drainline (lesser or greater) transport distances
remains a topic of great concern. Newly built buildings have many options for installed
systems choices; however, renovation or piecemeal installations require special focus
for performance assurances are neglected. A capability of combining both total solids
extraction assurance with greater carry distances requires an essential basis from greater
field input information as more than simply for further research. Extension into existing
building renewals with installed fixture replacements appears to be only partial, or
neglected, in technical sufficiency thrusts for reduced consumption in water
conservation.
Introduction of initial WC types following EPACT nationally established 1.6 gal (6 l)
flush in the U.S. remains the national requirement. Comparisons of data for differing
drain slopes shown provide some insight to the potentials of drainline transport over the
small range of data available.
237 Effects of the turning fitting at entry into the horizontal drain remains an open item;
conditions occur require investigations for impacts from entry into the drainpipe. Those
provide entry initial conditions for subsequent transport set by the profile. Solids
deliveries (from the WC into the drain) have little depths of understanding of governing
mechanisms (chaos) for establishment of entry conditions into drains.
Additionally, siphon surge studies for wastes extraction from the bowl do not appear.
Test data suggests dependency of solids removals on the entry fitting/pipe wall
roughness. Performance for solids removals capacity with varied wall roughness effects
extends to upstream influences that require additional research from defined testing.
Impacts on requirements applied to standards testing and required pass/fail must have
inputs from realistic simulations for test methods and materials. Tests require realistic
simulations for requirements based upon installation practices and usages. Current
“gravity aided vertical non-connected WCs extraction” requirements for solids result in
dubious findings and little confidence for installations applications.
7. References
1. Cummings Waterless Urinals
2. Austarlin Stds dry urinals
3. ASM E A112.19.2 STD
4. JAS Waklin
5. Gauley
6. Cummings Conf.
7. WSSC Yingling, Galowin
7. Presentation of Author
Dr. Lawrence Galowin - Directed National Plumbing
Research at NBS (NIST), now a Guest Researcher
since 1998. Set total dynamic computer operations
with recording in full-scale laboratory operations of
instrumented plumbing systems. ASME Member
A112 National Plumbing Standards Committee.
238 IV.2
Building rainwater harvesting systems.
Doubts and certainties
A. Silva-Afonso
silva.afonso@ua.pt
Departament of Civil Engineering, University of Aveiro, Campus Universitário
Santiago, 3810-193 Aveiro, Portugal
Abstract
Growing interest is being shown in the utilization of rainwater in buildings in many
countries, not only for reasons of rational use of water but as a way of reducing flood
peaks when it rains, too.
Germany and Brazil, for instance, have already established standards in this area. A
technical specification has also been published in Portugal recently. It is voluntary and
is run by a Portuguese NGO (ANQIP) that promotes quality and efficiency in water
supply and drainage for buildings.
This paper presents a detailed analysis of this new specification; it looks at certain
technical aspects of the conception and design of installation components and the
demands of water quality in light of its various uses.
Maintenance requirements are also examined, and a certification procedure is created
for systems to guarantee the overall technical quality of the installation and the
protection of public health.
Finally, the standards used in Germany and Brazil are compared and some aspects of
those where there is a broad agreement are analyzed. Aspects of those systems where
there is currently a certain amount of doubt are also examined, and the differences
between the standards used in several countries are indicated.
239 Keyword
ds
Water-efficciency; rainnwater harveesting; dom
mestic water.
1. Introd
duction
Rainwater harvesting systems inn buildingss have seen
n considerabble developpment in a
n
in Brazil
B
and Germany,
G
both
b
to encoourage the rational use
number off countries, notably
of water annd to help reeduce floodd peaks wheen it rains.
Portugal has
h also shoown increassing interesst in rainwaater harvestting throughh an NGO
that promootes qualityy and efficiiency in the water sup
pply and drainage
d
forr buildings
(ANQIP – National Association
A
f Quality in Building
for
g Installationns).
It should be
b noted thaat in terms of
o the rationnal use of water
w
the so-called Medditerranean
climate dooes not seem to be faavourable too making the
t most of rainwaterr since the
summers are
a typicallyy hot and dry
d and the winters aree cold and wet. Typical summer
climate usuually lasts tw
wo or three months.
As the nam
me suggestss this climatte is only found
fo
in thee Mediterrannean basin (Figure 1),
though sim
milar conditiions may occcur from tiime to timee in southernn Australia and on the
east coast of
o north andd South Am
merica. Mosst of Portugal, Spain, Ittaly and Grreece enjoy
this type off climate.
F
Figure
2 – Climate
C
chaaracteristiccs in variou
us regions of
o the world
d
Spain and Portugal, however,
h
arre at high risk
r
of hydrric stress inn the short-- /mediumterm (Figuure 2) and soo the harvessting of rainnwater in thee context off promotingg the global
24
40 water efficiency in buildings may play an important part in reducing this stress as well
as helping to reduce flood peaks in the winter.
2. Water efficiency in buildings: The 5R principle
The overall global waste in water use in Portugal is presently estimated at over 3 x 109
m3/year, which is around 39% of the country’s total water requirement.
With specific reference to the urban supply sector (public and building systems), total
waste is reckoned to be 250 x 106 m3/year, costing about 600 x 106 €/year.
In terms of figures per person, this amounts to waste of more than 25 m3/year, i.e., near
70 m3/year per family (average family consist of 2,7 persons in Portugal).
Bearing in mind the short- /medium-term water stress forecasts this situation is
unsupportable and needs urgent intervention through the application of measures to
rationalize water use.
Figure 2 – Hydric stress. Scenario in 2025 according to the World Water council
Rational water use in the urban cycle can be summarized as a principle analogous to but
more comprehensive than the 3R principle (used for waste) which is known as the 5R
principle (Figure 3).
241 - REDUCE CONSUMPTION - REDUCE LOSS - RE‐USE WATER - RECYCLE WATER ‐ RESORT TO ALTERNATIVE SOURCES }
‐ WATER EFFICIENCY IN BUILDINGS
Figure 3 – The 5R principle for water efficiency in apartment blocks
The use of rainwater is included in the fifth R (resort to alternative sources) and, as
mentioned above, ANQIP developed a specific technique for this (ETA specification
0701). Note that ANQIP has already devised a certification and labelling model for
water efficiency for products and it is currently developing specifications for the
recycling of greywater.
3. The Portuguese specification for rainwater harvesting in buildings.
Description and comparative analysis
3.1. Introduction
Obviously, as specification ETA 0701 has been formulated by a non-governmental
body, compliance with it is voluntary.
The specification has 6 chapters (Introduction, Definitions, Legal and regulatory
references, General aspects and certification, Technical Provisions and Maintenance),
and the certification of these installations by ANQIP is recommended.
This recommendation is justified so that the technical quality and public health can be
ensured. It implies the prior assessment of the design by ANQIP and the inspection of
works and certification of the installers by ANQIP.
242 3.2. Technical provisions
ANQIP compiled average rainfall maps of Portugal for the rainfall studies (Figure 4).
Figure 4 – Average rainfall map of Portugal (ANQIP)
One aspect to which special attention was given was the need to divert the first flush as
prolonged dry periods can aggravate the pollution of this water and automatic diverting
systems should be installed.
ETA 0701 allows criteria for the time or season of the rainfall to establish the amount to
divert. In the first case it is held that an amount corresponding to the first 10 minutes of
rainfall, though shorter times (but no less than 2 minutes) may be adopted if the period
between rainfalls is no more than 4 days.
The season criterion takes a reference figure of 2 mm of rainfall, though this may vary
between 0.5 mm and 8.5 mm, depending on local conditions and the period between
rainfall events.
The Brazilian standard also takes 2 mm of rainfall, but the German standard DIN 1989
does not have specific requirements.
243 ETA 0701 also requires the use of appropriate filters in the connection to the storage
tank (to trap leaves, etc.), as to the German and Brazilian standards.
The specification further includes technical provisions to prevent contamination in the
spillage of water from the overflow of the system from the first flush and from the
rainwater filter mesh, whether infiltrated or in a natural water course.
The installation of a device to reduce turbulence and reduce the speed at which the
water enters the storage tank is also required.
Pump suction should be at a slow speed and between 10 and 15 cm below the level of
the water in the storage tank, if possible (or through an equivalent system that prevents
the suction of floating material or sediment in it).
The ETA also contains various constructive provisions; there is a recommendation that
rainwater should be stored in place away from light and heat and that openings should
be fitted with anti-rodent and anti-mosquito devices.
A shut-off should also be installed at the start of the system so that if products that may
harm human health in the catchment area are used or spilt (on purpose or by accident)
the system can be disconnected and the products will not enter the storage tank.
The experts find it hard to agree on the best design for the storage tank. A great many
methods can be used: simple ones (German abridged procedure, German simplified
procedure, Spanish method, English practical method, Azevedo Netto method, and so
forth), and theoretical and probabilistic methods (Rippl method, simulation method,
Monte Carlo method, etc.).
The Portuguese specification proposes that the abridged German procedure be used for
current situations. It is described in the German standard and yields about 1 m3 per
person. Storage period should be no more than 30 days in any case, which is slightly
higher than the period established in the German standard (3 weeks).
Like the German standard, the specification contains a table of consumption per
installed appliance to help calculate the building’s water needs. The Portuguese table is
based on the use of appliances labelled 'A' for water efficiency under the ANQIP
certification system since the use of a rainwater harvesting system by non-efficient
appliances is not regarded as consistent.
These two tables differ essentially in terms of the amounts for watering outside areas,
due to climate differences.
It also stipulates (like the foreign standards mentioned earlier) that the drinking water
and non-drinking water systems should be clearly distinguished. Watering and washing
appliances, both indoor and outdoor, must be identified and marked with symbols (yet
to be defined).
244 It is also recommended that washing or watering taps should have removable
handles/levers (safety key) to prevent improper use.
The installation of a totalizing meter in the section connecting the storage tank to the
block’s system is considered, so that the water that does not enter the drainage system
(i.e. that used for watering gardens, etc.) is not measured.
Questions of quality also arouse significant differences of opinion among the experts.
The use of rainwater for washing clothes, for instance, is not allowed in Brazil but it is
in Germany. This difference of criteria may be due to the various washing temperatures
considered and their effect on microorganisms.
The Portuguese specification is nearer the German standard and considers the following
possible uses:
- Toilet storage tank flushing
- Washing clothes
- Washing floors, cars and so on.
- Watering gardens, lawns, parks etc.
- Industrial uses (cooling towers, firefighting systems, HVAC, etc.)
It is felt, moreover, that the use of untreated rainwater for toilet flushing should only be
acceptable if the water quality is at least up to that of bathing water pursuant to the
applicable European directives (Directive 76/160/EEC, of the Council, dated 8/12). It
may be disinfected with chlorine or a similar process, if necessary.
Clothes should only be washed with rainwater that has had no specific treatment if the
washing water temperature reaches at least 55ºC. A microfilter with a minimum mesh of
100 µm should be fitted if the water is to be used for this purpose.
If the pH of the rainwater is lower than 6.5 then pH correction may be necessary or
appropriate, depending on the materials used in the installation.
Discharges into the storage tank from the drinking water system must be undertaken in
such a manner that this system is not contaminated.
ETA 0701 also contains various notes and recommendations related to the
characteristics of the pumping equipment and its installation.
245 3.3. Maintenance of the systems
Technical Specification ANQIP ETA 0701 contains a maintenance schedule table.
It is analogous to that in the Brazilian standard and less comprehensive than the one in
the German standard.
4. Conclusions
The efficient use of water is an environmental must for every country in the world.
Some countries, like Mediterranean countries, must develop measures to ensure this as a
matter of urgency, since water availability could be significantly affected in the shortmedium-term.
Even though the Mediterranean climate is not really suitable for proper rainwater
harvesting this should still be considered in the context of the 5R of water efficiency in
buildings.
This is why ANQIP, a non-profit Portuguese NGO composed of companies and
universities decided to draw up a technical specification for the harvesting of rainwater
in buildings, similar to those developed in other countries.
Some aspects still have to be clarified in these systems, especially in relation to the
design of storage tanks and, more importantly, in relation to the issue of quality
associated with the possible uses of this water.
The matter of tariffs related to rainwater drainage could also be relevant to the
implementation of these standards. Note that in Germany, for instance,
impermeabilization of the ground is subject to a tax (not yet the case in Portugal) and
this can encourage the development of harvesting systems by weighting the recovery of
rainwater against such a tax.
It is felt that, despite the different climates occurring in Europe, it should be quite easy,
and even desirable, to draw up a European standard for this domain.
246 5. References
1. Oliveirra, A., Silva-Afonso, A.,
A Costa, V., Figueirredo, J., Cooelho, C.,Figueira, E.,,
Pereiraa, S. (2006). Optimizattion of The Water Cyccle in the H
House of the Future off
the Unniversity of Aveiro. Proceeding
gs - Internnational C
Conference RSC 06 Rethinkking Sustaiinable Construction 2006.
2
Next Generationn of Green
n Building.
Sarasotta, EUA: Unniversity off Florida.
2. Castro,, R.; Silvaa-Afonso, A.
A (2007). Integrationn of Sustaainability in
i Sanitaryy
Installaations: The Example off the Aveiro
oDOMUS House
H
of thhe Future. Proceedings
P
s
- SB07 Sustainablee Constructtion – Materrials and Prractices, Vool. 2 (pp.1083 – 1087)..
Lisbon, Portugal:.IIOS.
3. Rodriguues, C.; Sillva-Afonso,, A. (2007)). A Qualiddade na Connstrução ao
o Nível dass
Instalaçções Prediaais de Águuas e Esgo
otos. Situaçção e Persppectivas em
m Portugal,,
Proceeedings - Conngresso Connstrução 20
007. Coimbrra, Portugal: FCTUC.
4. Lança, I., Silva-A
Afonso, A.. (2008). As
A Alterações Climátticas, as Medidas
M
dee
A
ao Nível das Instalaçõess
Eficiênncia Energéética e a Saúde Pública. Uma Análise
Prediaiis. Proceediings - XIII SILUBESA – Simpósio
o Luso-Braasileiro de Engenharia
E
a
Sanitárria e Ambienntal. Belém
m do Pará, Brasil:
B
ABES
S.
5. Silva-A
Afonso, A., Pimentel-R
Rodrigues, C.
C (2008). Water
W
efficciency of prroducts andd
buildinngs: the impplementatioon of certiffication andd labelling measures in
n Portugal..
th
Proceeedings – CIB
B W062 20009 – 34 International Symposium
m on Water Supply andd
Drainaage for Builddings. Hongg-Kong, Ch
hina: HKPU
U.
6. Presen
ntation off Author
Armando Silva-Afonnso is Proofessor off Hydrauliccs at the
Universityy of Aveeiro (Porttugal), Deepartment of Civil
Engineerinng, and Chaairman of thhe Board off Directors of
o ANQIP,
an NGO thhat promotees quality annd efficienccy in water supply
s
and
drainage for
fo buildingss. His speciialisation is urban hydrraulics and
piping systtems. In thiis latter fieldd he is work
king on maathematical
models, suuch as stochastic moddels, for dem
mand forecasting and
the econom
mic design of interior networks. He has recently been
concentratting on imprroving wateer-use efficiency in buildings.
247 2
IV.3
Water conservation at International Airport of
São Paulo in Brazil: the Hidroaer Project
L. H. Oliveira (1), W. C. Sousa Júnior (2), M. S. O. Ilha (3), O. M. Gonçalves (4), M.
A. S. Campos (5), L. G. Pereira (6)
(1) Department of Construction Engineering of Escola Politécnica, Brazil, P.O. Box
61548, University of São Paulo, São Paulo, Brazil, e-mail: lucia.oliveira@poli.usp.br
(2) Department of Hydraulics, Instituto Tecnológico de Aeronáutica, São José dos
Campos, SP, Brazil, e-mail: wilson@ita.br
(3) Department of Architecture and Construction, School of Civil Engineering,
Architecture and Urban Design, University of Campinas, Campinas, SP, Brazil, e-mail:
milha@fec.unicamp.br
(4) Department of Construction Engineering of Escola Politécnica, Brazil, P.O. Box
61548, University of São Paulo, São Paulo, Brazil, e-mail:
orestes.goncalves@poli.usp.br
(5) Department of Architecture and Construction, School of Civil Engineering,
Architecture and Urban Design, University of |Campinas, SP, Brazil, e-mail:
marcussiqueira@yahoo.com.br
(6) Department of Architecture and Construction, School of Civil Engineering,
Architecture and Urban Design, University of Campinas, Campinas, SP, Brazil, e-mail:
galleo@uol.com.br
Abstract
The International Airport of São Paulo - AISP, in Guarulhos, presents the water and the
energy consumption as the similar way to the one of a medium city. It annually takes
care of 17 million passengers and with forecast of amplification for 29 million. Brazil is
one of the countries with largest water resources with 33,000 m3/hab.ano, but spite of
that it suffers water scarcity in some regions, especially in the places of bigger urban
concentration, as the state of São Paulo, whose water resources is of 6,607 m3/hab.ano.
This scene demands greater investment in the efficient management of the water. The
AISP has about 200 sanitary, beyond the frozen water central and of the sector of food.
To make possible a management adjusted with reduction of water consumption is being
developed the Project called HIDROAER that aims to promote the efficient water use
with water saving technologies and, also, the amplification of water offers to an airport
248 plant. To reduce the water demand it is necessary to know the standard of water
consumption of the various sectors of the airport, what it makes possible to specify
efficient systems and components more adjusted. Therefore, the aim of this work is to
present an overview of the water supply system and the water consumption at the AISP
and the plans
for a survey of the water consumption pattern of one sanitary composed by six
restrooms of the international terminal of passengers. The results, beyond directly
subsidizing the water demand management, from the control of water losses, can
influence the development of water supply systems designs in similar airports and other
public plants.
Keywords
Water conservation; airport; HIDROAER; water consumption pattern.
1. Introduction
Brazil presents a favorable average value of water availability, about 33000
m3/person/year. However, this value varies from 1145 to 533096 m3/person/year and in
the state of São Paulo, where the International Airport of São Paulo - AISP is located,
the water availability presents about 2209 m3/person/year [1], which is a critical
situation. This situation implies the necessity of reduction of the water demand in the
AISP to make it possible the future construction of the third air terminal.
There is a world-wide interest in reducing the environmental impact from the part of the
airport administration, in special those related to the energy and water uses. One of the
first initiatives was verified, according to Chouthai et al. [2], in the International Airport
of Stapleton, Denver, Colorado, U.S.A. where a laboratory and field research was
conducted with the aim to verifying the performance of WC cistern of 6 liters and,
mainly, and its ability to transport solids in the sewage system in an efficient way
without high costs of maintenance and/or blockages in horizontal branches.
The San Francisco Airport shall minimize potable water use by deploying water saving
equipment and facilities and shall contribute to improving water quality in the lower
San Francisco Bay through state-of-the-art wastewater treatment and enhanced storm
water management. For this, one of the goals is to maximize water conservation and
minimize water use and waste [3].
The INFRAERO – Brazilian Airports is responsible for the administration of 67
Airports and 81 Air Navigation Support Units throughout Brazil. Aware of its
environmental responsibilities, INFRAERO maintains an Environmental Policy
compatible with the planning, construction and operations of its activities in accordance
to national and international regulations and laws. INFRAERO’s environmental
management system is supported by the Environmental Programs developed by the
Superintendence of Environment and Energy, the environmental units of the Regional
Superintendence and airports, one of them being the Water Resources Program [4].
249 The objective of the Water Resources Program is to systematise procedures in order to
reduce and optimize water consumption at airports, protect watersheds, and preserve
springs and bodies of water in ways compatible with airport activities. Another
objective is to stimulate the adoption of water saving technologies that reduce water use
in new constructions and make current buildings more efficient.
Thus, the International Airport of São Paulo (AISP) is the object of study of the
HIDROAER Project - Efficient water use at airports, which has been developed by the
Instituto Tecnológico de Aeronáutica (ITA) in partnership with the University of São
Paulo and the University of Campinas and financed by the FINEP - Financiadora de
Estudos e Projetos e Conselho Nacional de Desenvolvimento Científico e Tecnológico
(CNPq). The objectives of the HIDROAER Project are to specify saving water
technologies as well as the implement of non-potable water systems for the reduction of
water consumption, from a diagnosis of the water use at the AISP, carried out in 2006
by AISP.
The aim of this work is to present an overview of the water supply system and the water
consumption at the AISP and the plans for a survey of the water consumption pattern of
the restroom 93 of the international terminal of passengers, one of them for men and
another for women. There are also small restrooms for children and for people with
special needs in this restroom 93.
2. Characteristics of water supply and drainage system of AISP
The International Airport of São Paulo (AISP) is located in the city of Guarulhos, about
20 kilometers from the city of São Paulo, in an area of 14000 km2. It has been managed
by the Brazilian Airports (INFRAERO), since 1985, when it started operating.
The AISP can attend about 17 million passengers per year with two air terminals.
However, it takes care of about 12 million of users annually and the construction of the
third terminal of passengers is been prepared. After this, the AISP will increase its
capacity for 29 million passengers per year. It is important to mention that it is one of
the main modes of logistics of air cargo, with the largest cargo terminal in South
America [5].
2.1 Water supply and drainage system
In accordance to the Report of the Action Plan [5], the water supply in the AISP is
carried out by underground water through wells provided with water meters, which
allow the management of the volume of water consumption. The collected water is
conveyed to a treatment plant and later distributed to three reservoirs with 199 m3 of
capacity each. From these reservoirs the water is conveyed by gravity to the airport.
The average static head of the water supply system, in the six areas of the AISP,
presents a range of 30 kPa to 640 kPa, being that in the four of them the average value
of hydraulic head is about 600 kPa, which is considered a high value and, in this way, it
contributes to higher values of flow rates in the water taps.
250 The drainage system
m conveys thhe effluentss from the AISP
A
to tw
wo biologicaal treatmentt
fter the treattment the water is conv
veyed to laggoons and too a stream.
plants. Aft
2.2 Waterr consumpttion distrib
bution in th
he AISP
mative of monthly wateer consump
ption is of 49907
4
m3 aand the maain uses aree
The estim
distributedd in the folllowing poinnts of use [5],
[ as preseented in Figgure 1. Notte that onlyy
the water consumptioon estimatedd for the WC,
W the faucets and the urinals corrresponds too
i
thatt the restroooms are criitical pointss
73% of thhe total water consumpption. This implies
of water consumptionn in the AIS
SP.
Others
Air 7%
Faucet
15%
conditio
ned
20%
WC
51%
Urinal
7%
Figuree 1: Water consumptio
c
on distribu
ution in the Internatioonal Airporrt of São
Paulo [5]
3. Meth
hodology
For the sttudy and specification
s
n of the water
w
savingg technologgies to be replaced
r
inn
restroom 93
9 of the AISP,
A
whichh was choseen as a piloot, located inn the superrior level off
the Terminal of Passsengers and the internaational sectiion. Restrooom 93 is co
omposed off
two restroooms for aduults, one forr women an
nd the otherr for men, tw
wo small restrooms forr
children, one
o for boyys and the other
o
for girrls. In addittion, two sppecial needss restrooms,,
one male and
a anotherr female. Figgure 2 show
ws the plant of the studiied environ
nments.
Drin
nk Fontains
Lavatories
MENS'
RESTROOM
Urinals
BOYS'
RESTROOM
HANDICAPPED
MENS' RESTROOM
HANDICAPPED
LADIES' RESTROOM
GIRLS'
RESTROOM
Water Closets
LADIES'
RESTROOM
Figuree 2: Plant of Restroom
m 93 monito
ored at the International Airporrt of São
Pa
aulo
251 2
The water consumption of the appliances is being measured by 12 water meters
connected to their respective branches, as shown in Table 1.
Table 1 – Water meters installed in restroom 93 of the AISP
Water
meter
H2
Type and
Appliances monitored
Metrological class
Volumetric, class D 1 wash basin with metering faucet + 1 faucet for cleaning
- girls’ restroom
H3
Volumetric, class D 5 wash basins with electronic faucets - men’s restroom
H4
Volumetric, class D 1 wash basin with electronic faucet - men’s restroom
H5
Volumetric, class D 1 water faucet for cleaning - men’s restroom
H6
Volumetric, class C 1 wash basin with metering faucet + 1 faucet for cleaning
(boys’restroom)
H7
Multi jet, class B
6 valve operated water closets(1) - 4 in the men’s
restroom, 1 in the boys’ restroom and 1 in the special
needs restroom (male)
H8
Volumetric, class D 2 drinking fountains - outside hall
H9
Volumetric, class D 1 faucet for cleaning - ladies’ restroom
H10 Volumetric, class D 1 wash basin with electronic faucet - ladies’ restroom
H11 Volumetric, class D 5 wash basins with electronic faucet - ladies’ restroom
H12 Multi jet, class B
8 valve operated water closets(1) - 6 in the ladies’
restroom, 1 in the girls’ restroom and 1 in the special
needs restroom (female)
H13 Volumetric, class C 5 urinals (electronic) - men’s restroom
(1) The duration of flush depends on the user, i.e. the volume is variable.
The water meters are instrumented for remote metering and the data are presented in
electronic sheets. The local of the water meters was chosen considering the branches
that feed just one type of sanitary appliance and also considering the specific difficulties
to install them.
Isis
3.1 Steps of the research
Five steps of monitoring, in function of the water saving technologies, have been
defined to be installed in the water taps, as presented in Table 2. Each step of
measurement is about a period of 15 days. At this moment, step 1 is being implemented,
with the objective of evaluating the water consumption with the existing sanitary
appliances. This step has a greater number of days than the other steps of the study.
252 Table 2 - Steps and respective water save technologies to be implemented in
restroom 93 of the AISP
Step
1
2
Activity
Installation of water meters and instrumentation
Measurement 1: data collect and data analysis
Adjustments in the water taps: flow rates, duration of discharges, volumes
Measurement 2: data collect and data analysis
Installation of:
3
4
5
6
a) electronic valves in the water closets - ladies and men’s restrooms;
b) electronic faucets in the wash basins – special needs ladies´ and men’s
restrooms; and in the boys and girls’ restrooms
c) foot faucet - boys and girls’ restrooms
Measurement 3: data collect and data analysis
Installation of: metering faucets in the wash basins - all restrooms
Measurement 4: data collect and data analysis
Installation of: double flush toilet valves (3/6 L) – all restrooms
Measurement 5: data collect and data analysis
Installation of conventional faucets with aerators and flow rate regulator valves
in all water taps (all restrooms) to compare the results
Measurement 6: data collect and data analysis
The population is being controlled by the number of passengers of each international
flight which has been provided by AISP. The average number of passengers per day in
the influence area of the Restroom 93 was estimated in 3227 in May and 3796 in June
(100024 and 113889 passengers per month, respectively).
3.2 Development of the data acquisition system
The development of a system of data acquisition is also objective of this research which
is being used to collect the data in restroom 93. This system operates from electronic
modules of communication in net, connected to the water meters that generate pulses in
an output pre-established. The system operates in high frequency and data
communication short bands.
The net is dimensioned to overcome prompt failures of transmission, providing highly
reliable data security. Moreover, the system counts on coordinator modules, generating
253 a positive redundancy for the operation. Each device has a storage capacity to
accumulate registers in case of temporary failure of the net, with the guarantee of
recovery of the accumulated information when the operational conditions return.
The use of this technology brings some comparative advantages for this type of
application, such as:



cost of installation and implementation of the reduced net. Therefore the data are
transmitted without cables, not being necessary the installation of cables and
connectors between the modules and the remote central office;
low cost of the modules, once it deals simple technology, that uses low transference
tax of (up to 250 kb/s), more than enough for the application in question;
low consumption of power, which allows the modules to be fed by common
batteries, functioning per years without the necessity of exchange of batteries.
4. Results and discussion
The total daily water consumption (all restrooms) in the first step was 42.4 m3, on
average, with a standard deviation of 19.9 m3, as shown in Figure 2.
80
70
60
Average: 42,4 m3/day
40
3
m /day
50
30
20
10
0
6/10 6/11 6/12 6/13 6/14 6/15 6/16 6/17 6/18 6/19 6/20 6/21 6/22 6/23 6/24 6/25 6/26 6/27 6/28 6/29
Day
Figure 2: Daily water consumption in the restroom 93 of the AISP
254 In this period, 86% of the water consumption, on average, has been occurred in the WC,
as it illustrates in the Figure 3. The distribution of the water consumption of the other
sanitary appliances is presented in Figure 4.
Considering only the water consumption at the ladies and men’s restrooms, the WC
consumption represents between 80 and 90% of the total consumption of each restroom.
100%
90%
Average: 86%
80%
70%
60%
50%
40%
30%
20%
10%
0%
6/10 6/11 6/12 6/13 6/14 6/15 6/16 6/17 6/18 6/19 6/20 6/21 6/22 6/23 6/24 6/25 6/26 6/27 6/28 6/29
H7
H12
H7 – 6 water closets - 4 in the men’s restroom, 1 in the boys’ restroom
and 1 in the handicapped men’s restroom.
H12 – 8 water closets 6 in the ladies’ restroom, 1 in the girls’ restroom
and 1 in the handicapped ladies’ restroom.
Figure 3: Distribution of the water consumption in the WC of the restroom 93 of
the AISP
255 24%
22%
20%
18%
16%
14%
12%
10%
8%
6%
4%
2%
0%
6/10
6/11
6/12
6/13
6/14
6/15
H2
6/16
H3
6/17
H4
6/18
H5
6/19
6/20
Day
H6
H8
H2 -1 wash basin with metering faucet +
1 faucet for
cleaning - girls’ restroom
H3 - 5 wash basins with electronic faucet
- men’s restroom
H4 - 1 wash basin with electronic faucet
- men’s restroom
H5 - 1 faucet for cleaning - men’s
restroom
H9
6/21
H10
6/22
H11
6/23
6/24
6/25
6/26
6/27
6/28
6/29
H13
H6 -1 faucet for cleaning - men’s
restroom
H8 - 2 drinking fountains - outside hall
H9 - 1 faucet for cleaning - ladies’
restroom
H10 - 1 wash basin with electronic faucet
- ladies’ restroom
H11 - 5 wash basins with electronic
faucet - ladies’ restroom
H13 - 5 electronics urinals - men’s
restroom
Figure 4: Water distribution consumption in the other sanitary appliances of the
restroom 93 of the AISP
The water consumption at the urinals is less than 1% of the daily consumption. This
may be due to the lack of wall division between these appliances, with no privacy for
the use. Also, the valve operated WC and the urinals are operated by electronic sensor
with about 1 liter per flush. To illustrate this fact, Figure 5 shows the water consumption
pattern for the WCs and the electronic urinals for a same day and Figure 6 shows the
water consumption in the other appliances on the same day.
Another important piece of information was the peak period (Tp), defined in terms of
the consumed maximum volume in each hour of the day. Considering all the sanitary
appliances of the men´s restroom, it was verified that the peak of water consumption on
this day was between 5 and 6 h (2674 liters in 1 hour). In the ladies´ restroom, 5 periods
of one hour with similar consumption was verified: between 5 and 6 h (2600 L),
between 11 and 12h (2659), between 13 and 14h (2637 L), between 17 and 18h (2346
L) and between 22 and 23h (2166 L).
256 40
water consumption (liters in 1 hour)
35
30
25
20
15
10
5
15:01-16:00
16:01-17:00
17:01-18:00
18:01-19:00
19:01-20:00
20:01-21:00
21:01-22:00
22:01-23:00
23:01-00:00
16:01-17:00
17:01-18:00
18:01-19:00
19:01-20:00
20:01-21:00
21:01-22:00
22:01-23:00
23:01-00:00
14:01-15:00
15:01-16:00
13:01-14:00
12:01-13:00
11:01-12:00
10:01-11:00
09:01-10:00
08:01-09:00
07:01-08:00
06:01-07:00
05:01-06:00
04:01-05:00
03:01-04:00
02:01-03:00
01:01-02:00
00:01-01:00
0
Hour
(a)
2750
2500
water consumption (liters in 1 hour)
2250
2000
1750
1500
1250
1000
750
500
250
14:01-15:00
13:01-14:00
12:01-13:00
11:01-12:00
10:01-11:00
09:01-10:00
08:01-09:00
07:01-08:00
06:01-07:00
05:01-06:00
04:01-05:00
03:01-04:00
02:01-03:00
01:01-02:00
00:01-01:00
0
Hour
(b)
Figure 5: Water consumption (a) in the five electronic urinals appliances and
(b) in the six WC of men’s restroom during June, 12th, 2009 day
The water taps are also being monitored with the aim to determine the number of uses;
however these results are not available yet.
5. Final considerations
This paper presented a survey of the water consumption pattern of one sanitary of the
international terminal of passengers in the International Airport of São Paulo, Brazil.
This restroom is composed of six restrooms, including installation for children and for
people with special needs. The study is right now at first step 1.
The main objective of this study is to evaluate water saving components that represent
major water consumption decrease for other restrooms of the airport. Also, it will make
a proposal for other similar airports in the country.
The results obtained so far (step 1) indicate that the WCs represent 86% of the water
consumption, on average. It is due to the fact of the discharge volume of all WCs
depends on the users. It is also true for the men’s restroom, which has urinals. Urinals
257 are rarely used, maybe because there is no wall division between them (no privacy for
the use).
The water consumption peak period was more concentrated on the men’s restroom. On
the other hand, there are three or more periods of peak with similar values to this
parameter on the ladies’ restroom.
Next five steps of the study will contemplate the evaluation of the water consumption
with different types of water saving components and conventional ones for comparison.
Acknowledgements
The authors thank the Financiadora de Estudos e Projetos - FINEP (Research and
Projects Financing – Brazilian Innovation Agency, Ministry of Science and
Technology) and Conselho Nacional de Desenvolvimento Científico e Tecnológico –
CNPq (The National Council for Scientific and Technological Development, Ministry
of Science and Techonology) for the financial support.
6. References
[8]
AGÊNCIA NACIONAL DE ÁGUAS, Ministério do Meio Ambiente.
Disponibilidade e demandas de recursos hídricos no Brasil. Cadernos de recursos
hídricos 2. Maio, 2007.
[9] CHOUTHAI, A.; VIJAPUR, B.; KONEN, T. A laboratory and field evaluation
of 1.6 gpf water closets in a commercial setting. Hoboken, New Jersey, August,
1992. (Report n. 246, prepared for Denver Water Department, Denver, Colorado).
[10] SAN FRANCISCO INTERNATIONAL AIRPORT. SFO - Environmental
Sustainability Report, 2007.
[11] INFRAERO – Empresa Brasileira de Infra-Estrutura Aeroportuária. Relatório
Ambiental 2005/2006. Brasília 2006. Disponível em: http://www.infraero.gov.br.
Acesso em 21 jun. 2009.
[12] AEROPORTO INTERNACIONAL DE SÃO PAULO. Relatório do Plano de
Gestão de Recursos Hídricos. Vitalux Efifiência Energética Ltda, 2006. Contrato
0053-St/2005/0057.
258 7. Presentation of Authors
Lúcia Helena de Oliveira is a professor at Department of Construction Engineering of
Escola Politécnica of University of São Paulo, Brazil, where she teaches and conducts
researches on building services.
Wilson Sousa Júnior is a professor at Department of Hydraulics, Instituto Tecnológico
de Aeronáutica, where he teaches and conducts researches on Environmental
Engineering, with emphasis in integrated water resources planning and economics.
Marina Ilha is a Head of Department of Architecture and Construction, School of Civil
Engineering, Architecture and Urban Design, Universidade Estadual de Campinas,
where she is the Head of the Building Services Research Group.
Orestes Gonçalves is a professor at Department of Construction Engineering of Escola
Politécnica of University of São Paulo, Brazil, where he is the Head of the Building
Services Research Group.
Marcus André is a PhD student at School of Civil Engineering, Architecture and Urban
Design, Universidade Estadual de Campinas, Brazil.
Leonel is a researcher at School of Civil Engineering, Architecture and Urban Design,
Universidade Estadual de Campinas, Brazil.
259 IV.4
WATER EFFICIENCY OF PRODUCTS.
OUTCOME OF APPLYING A
CERTIFICATION AND LABELLING
SYSTEM IN PORTUGAL
PIMENTEL-RODRIGUES, Carla
Civil Engineer, MSc
anqip@civil.ua.pt
Associação Nacional para a Qualidade nas Instalações Prediais (ANQIP) – Av. Fernão
de Magalhães, 151 – 4ºB, 3000-176 Coimbra, Portugal
SILVA-AFONSO, Armando
Professor, PhD
silva.afonso@ua.pt
University of Aveiro – Campus Universitário Santiago, 3810-193 Aveiro, Portugal
Abstract
Mankind is already using approximately 50% of the fresh water resources available. In
only 15 to 20 years this percentage could rise to 75%. As a consequence the risk of
hydric stress will increase significantly across the entire planet and some countries, such
as Portugal, might experience very serious problems in a large part of their territory by
2025.
In addition, there is a high level of inefficiency in the water supply system in Portugal
(in public and building systems), which amounts to approximately 250 million m³ a
year, representing 60% of total inefficiencies in economic terms. The Government aims
to rectify this situation and has drawn up and published a National Program for Efficient
260 Water Use, anticipating the support of various non-governmental organisations and nonofficial bodies to achieve its goals.
The truth is that the risks of hydric stress and the high level of inefficiency in the
networks require the immediate implementation of various measures, including the
promotion of the use of efficient products in buildings.
With this objective in view, ANQIP (National Association for Quality in Building
Installations), a Portuguese non-governmental organisation dedicated to the promotion
of quality and efficiency in buildings, decided to launch a voluntary certification and
labelling system for product water efficiency in October 2008.
This paper describes the system that is being implemented and the categories attributed
to each product. The results of the first semester of the implementation of the system are
also presented.
Keywords
Certification; water efficiency; labelling.
1. Introduction
Water has become a resource of the utmost importance. Demographic growth and, most
especially, economic development and today’s lifestyles have rendered drinking water
scarce, and its status has changed over the past decades from that of a community and
national asset to that of an economic one.
Climate change has worsened the situation and it is predicted that in certain countries,
such as Portugal, the forecast reduction in rainfall or the alteration of its regime could
have a negative effect on situations of crisis in the short to medium term.
Because water is a finite resource which is essential to life, its rational usage at all levels
is now a priority.
It is reckoned that total inefficiency in water use in Portugal, in all sectors, amounts to
3100 x 106 m3/year, which represents nearly 0.64% of national GDP. About half this
figure can be ascribed to inefficiencies in urban supply (public and building systems).
In Portugal the need for efficient water usage has already been recognised as a national
priority in the publication of a National Program for Efficient Water Use (PNUEA).
Amongst the actions suggested in this Plan are proposals for the labelling of devices in
buildings (flushing systems, showers, etc.) in order to provide consumers with
information as to their water efficiency. The Plan suggests that this measure be made
compulsory after a transitional period.
The National Program for Efficient Water Use (PNUEA) [1] also predicts the
involvement of companies, management organisations and non-governmental
organisations for the implementation of the said measures. ANQIP (National
Association for Quality in Building Installations) is the only large Portuguese
association which focuses on building installations and it covers the sector of
261 businesses, universities, management organisations and technical companies. Its
responsibility is clear in terms of launching the process and its leadership role.
2. Water efficiency certification and labelling in Portugal
2.1 Short description of ANQIP (National Association for Quality in Building
Installations)
ANQIP (www.anqip.pt) is a Portuguese non-profit association and was established in
2007. Its members include several universities, firms from the sector, management
organisations and self-employed technicians, whose basic aims are to promote and
ensure water quality and efficiency in the water supply and drainage fittings and fixtures
of buildings [2][3][4][5].
Under its powers and in accordance with the proposals of the National Plan for Efficient
Water Use ANQIP decided to introduce a product certification system and a water
efficiency labelling scheme in Portugal, starting in 2008.
The model used (described below) was implemented in stages and started with cistern
toilets, since these account for most consumption in building systems in Portugal.
2.2 The water efficiency labelling model proposed for Portugal
The water efficiency labelling of products has generally been implemented voluntarily
in various countries.
In some countries efficiency is not graded, but an efficiency label is awarded when
consumption is less than a specific amount. This is the labelling system in use in the US
and Scandinavia, for example.
In Australia and Ireland (Dublin), however, the label indicates a classification that
varies with the product’s efficiency [2][4].
ANQIP has opted for a voluntary model of the latter kind for Portugal. Figure 1 shows
the labels used. The base colours, which cannot be seen in the Figure, are green and
blue.
"A" signifies the greatest efficiency and is considered ideal. It also takes into account
the user-friendliness and performance of the devices in question. There is a graphic
indication by means of drops, for a better understanding of the symbol, and a small
informative bar at the side.
262 The A+ and A++ ratings are meant for special or regulated applications, as explained
below.
Figure 1: Portuguese water efficiency labels
ANQIP has drawn up Technical Specifications (ETA) for different products so as to
create and establish the necessary benchmark values to be assigned to each letter. These
Technical Specifications also establish the certification testing conditions.
Firms signing up to the system will sign a protocol with ANQIP which will define the
conditions under which they can issue and use the labels.
ANQIP controls the process by randomly testing labelled products on the market, from
time to time. These tests are performed by accredited laboratories or by laboratories
which are recognised by the Association.
2.3 Certification and labelling of flushing cisterns
As mentioned earlier, cisterns were regarded as a priority since toilet flushing cisterns
are one of the biggest consumers of water in buildings in Portugal.
As there is a project for a European Standard for WC and urinal flushing cisterns (prEN
14055:2007), it was decided that the labels of water efficiency to be used in Portugal
should comply with this Standard, where applicable.
The following mechanisms are also regarded as water-saving devices, under this
Standard:
263 a) Double-action mechanisms (interruptible): one action initiates flushing and a second
action stops the flush;
b) Dual-control mechanisms: one control releases the full flush volume and another
control releases a reduced flush volume.
The reduced volume cannot be greater than two-thirds of the maximum flush.
Table 1 presents the categories defined in the Technical Specification ANQIP 0804 for
flushing cisterns.
The minimum permitted volume or discharge amounts in current facilities are limited
for reasons linked to performance, user-friendliness and public health. The use of 4-litre
flushing cisterns, for example, has led to problems in the flushing of solids in building
and public networks. Therefore, their usage requires an alteration of the usual criteria of
the design of the drainage system (which is incompatible with many existing drains).
In addition, the European Norm EN 12056-2 does not allow the use of 4-litre flushing
systems in building systems whose design comply with System I of the said Norm, and
this is precisely the most common system in Portugal, allowed by the General
Regulation.
Furthermore, it must be ascertained if the discharge volume is compatible with the other
characteristics of the cistern toilet. Product performance is usually ensured by
compliance with European norms, meaning that any water efficiency certification must
require prior compliance with the existing norms in terms of the product’s respective
performance (in the case of flushing cisterns, as mentioned above, it is the prEN 14055).
Based on these facts, ANQIP established low volume flushing cisterns belonging to
water efficiency categories A+ or A++, but with the obligation that the label should
warn users of the need to guarantee the performance of the system and compatibility of
the drainage conditions in the building system (Figure 2).
264 Table 1: Water efficiency categories for the labelling of flushing cisterns
Tolerance
Tolerance
Nominal
Water
(Minimum
volume
Type of flush
efficiency (Maximum volume
volume for
– complete
rating
(litres)
water-saving
flushing)
flushing)
4.0
Dual control
A++
4.0 – 4.5
2.0 – 3.0
5.0
Dual control
A+
4.5 – 5.5
3.0 – 4.0
6.0
Dual control
A
6.0 – 6.5
3.0 – 4.0
7.0
Dual control
B
7.0 – 7.5
3.0 – 4.0
9.0
Dual control
C
8.5 – 9.0
3.0 – 4.5
4.0
Interruptible
A+
4.0 – 4.5
-
5.0
Interruptible
A
4.5 – 5.5
-
6.0
Interruptible
B
6.0 – 6.5
-
7.0
Interruptible
C
7.0 – 7.5
-
9.0
Interruptible
D
8.5 – 9.0
-
4.0
Complete
A
4.0 – 4.5
-
5.0
Complete
B
4.5 – 5.5
-
6.0
Complete
C
6.0 – 6.5
-
7.0
Complete
D
7.0 – 7.5
-
9.0
Complete
E
8.5 – 9.0
-
265 Figure 2: Examples of water efficiency labels for low volume flushing cisterns
2.4 Certification and labelling of showerheads and shower systems
Shower systems and showers represent over 30% of the daily average domestic water
consumption volume in Portugal [2] [5].
At this level, efficiency reduces both water consumption and the consumption of energy
required for the production of hot water.
The classification of these devices considers the following:
- Shower heads (showers), individually;
- Shower taps equipped with a hose and a shower head or with a fixed shower
head (shower systems).
For shower systems and showers, the model implemented considers the ideal usage
(letter A) to represent a water usage of between 5.0 litres/minute and 7 litres/minute.
The A and A+ labels applied to shower heads with a discharge which is 5 l/min or less
must bear the indication “Recommended for usage with thermostatic taps”, due to the
increased risk of scalding.
In products which can be regulated by the consumer, certification may be awarded on
the basis of the most efficient position, as long as the criterion is clear to the consumer,
without any risk of confusion, and it must be marked next to the label.
Due to the fact that discharge is dependent on residual pressure, the established
reference residual pressure for all ratings and for the tests was 300 kPa, which
represents the average pressure in Portugal and is the pressure selected by several
recognised laboratories for various tests.
The taps for bathtubs were not rated, because hot water consumption depends on the
volume of the tub to fill, and not on the discharge of the device.
Table 2 presents the various efficiency categories for showers and shower systems.
266 Table 2 – Water efficiency ratings for the labelling of showers and shower systems
Shower
system with a
DISCHARGE
thermostatic
(Q)
Showers Shower systems
tap and an
(l/min)
eco-stop
function
(1)
Q≤5
A+
A+
A++
A++ (1)
5.0 < Q ≤ 7.0
A
A
A+
A++
7.0 < Q ≤ 9.0
B
B
A
A+
9.0 < Q ≤ 15.0
C
C
B
A
15.0 < Q ≤ 30.0
D
D
C
B
30.0 < Q
E
E
D
C
Note (1): Eco-stop functions are not considered of interest in these cases
Shower
system with a
thermostatic
tap or an ecostop function
2.5 Certification and labelling of taps
The certification and labelling of taps are now being study by the technical commissions
of ANQIP.
Taps are the most common device, both in homes and in collective facilities. In an
average home, there are at least 3 to 5 taps installed in the kitchen and bathrooms [2]
[5]. They are used frequently, and their usage is difficult to quantify and varies greatly
in time and space. This variation also applies to the length of time of use, which can
stretch from a few seconds to several minutes.
No specific reference is made to self-closing taps, because according to recent studies
carried out in the United States of America, they do not lead to significant water savings
since, although they might run for less time, the discharge is always at maximum level.
The advantage of such taps lies in their safety aspect and not in water saving.
The case of self-closing taps with a sensor is similar. The advantage of these taps over
traditional ones is that they are more hygienic, but they are no more efficient, usually.
On average, it is estimated that water from taps represents approximately 16% of
consumption in homes in Portugal.
In the case of bathroom taps (in homes), the model which is currently being studied
considers ideal usage (letter A) to be a level of water consumption of 2.0 l/minute,
taking into account the studies performed and proposals made in countries where water
efficiency labelling has already been implemented.
For kitchen taps, the model considers ideal usage (letter A) to be a level of water
consumption of 4.0 l/minute.
Taps with an aerator are recommended for categories A++ and A+.
Kitchen taps with a discharge of under 4 litres per minute and bathroom taps with a
discharge of under 2 litres per minute (in homes) must bear a label with an advisory
note recommending that they be utilised only with an aerator.
267 In public areas, however, the usage of taps discharging a volume higher than or equal to
2 litres per minute might be advisable (usually letter B or above for basic models).
Tables 3 and 4 presents the various efficiency categories for taps, in study.
Table 3 – Water efficiency ratings for the labelling of bathroom taps (in homes)
Discharge
(l/min)
Bathroom
taps
Bathroom taps with
an aerator or an ecostop function
Bathroom taps with
an aerator and an
eco-stop function
Q≤ 2.0
A
A+
A++
2.0 < Q ≤ 4.0
B
A
A+
4.0 < Q ≤ 6.0
C
B
A
6.0 < Q ≤ 8.0
D
C
B
8.0 < Q
E
D
C
Table 4 - Water efficiency ratings for the labelling of kitchen taps
Discharge
(l/min)
Kitchen
taps
Kitchen taps with an
aerator or an ecostop function
Kitchen taps
with an
aerator and
an eco-stop
function
Q ≤ 4.0
A
A+
A++
4.0 < Q ≤ 6.0
B
A
A+
6.0 < Q ≤ 8.0
C
B
A
8.0 < Q ≤ 10.0
D
C
B
10.0 < Q
E
D
C
The certification and labelling systems for urinals and other devices will only be
developed by the end of 2010.
268 3. Results of the implementation of the system in Portugal. The case of
the flushing cisterns
The water efficiency certification and labelling system for flushing cisterns was
implemented in the last quarter of 2008. Approximately 40% of the companies on the
market adhered to the new system from the outset. Initially, 29 flushing models were
certified. Many companies and consumers have complied with the system, and it now
covers about 70% of the national market. 44 flushing models have been certified,
corresponding to 93 commercial references.
Table 6 summarises the certifications awarded per category.
Table 6 – List of certified flushing cisterns according to category
NO. OF CERTIFICATIONS
CATEGORY
AWARDED
A++
0
A+
2
A
86
B
5
C
0
D
0
E
0
The situation presented in Table 6 was expected (i.e. no certifications awarded to the
less efficient categories). In fact, because compliance with the system is voluntary,
manufacturers/importers do not usually request labelling for less efficient categories.
This is not negative for the system; quite the contrary. Since so many companies and
consumers complied with the system, the lack of certification of the said flushing
cisterns will gradually lead to their removal from the market, thus contributing towards
ANQIP’s goals.
269 4. Conclusions
Efficient water use is an environmental priority in all countries of the world. However,
in some countries, such as Portugal, the development of measures in this field has
become urgent because the availability of water could be significantly reduced in the
short or medium term.
Special attention must therefore be given to the use of efficient products, and consumers
must be able to identify these efficient products, leading to the need for a labelling
system which is easy to understand.
In Portugal, ANQIP, a non-profit NGO, has decided to launch a voluntary water
efficiency labelling system for products, similar to those developed in other countries.
Over 70% of the companies operating on the Portuguese market have adopted the first
certification and labelling system, for flushing systems, and 93 certified flushing
systems are available.
This initiative will most certainly provide an answer to the crucial and urgent need for
intervention in the field of rational water use in Portugal, aiming to guarantee in the near
future the essential sustainability conditions desired.
5. References
1. Laboratório Nacional de Engenharia Civil (2001). Programa Nacional para o uso
eficiente da água. Lisboa: Portugal: LNEC.
2. Silva-Afonso, A.; Abrantes, V. (2008). Water-efficiency in the housing sector. The
implementation of certification and labelling measures in Portugal. XXXVI IAHS
World Congress on Housing. Kolkata, Índia: IAHS.
3. Rodrigues, C.; Silva-Afonso, A. (2007). A Qualidade na Construção ao Nível das
Instalações Prediais de Águas e Esgotos. Situação e Perspectivas em Portugal,
Proceedings - Congresso Construção 2007. Coimbra, Portugal: FCTUC.
4. Rodrigues, C.; Silva-Afonso, A. (2008). A implementação da certificação de
eficiência hídrica de produtos em Portugal. Uma iniciativa para a sustentabilidade,
Proceedings - Congresso de Inovação na Construção Sustentável CINCOS 08.
Curia, Portugal: CINCOS.
5. Silva-Afonso, A., Rodrigues, C. (2008). Water efficiency of products and buildings:
The implementation of certification and labelling measures in Portugal. Proceedings
– CIB W062 2008 – 34th International Symposium of Water Supply and Drainage
for Buildings. Hong-Kong, China: HKPU.
270 6. Presen
ntation of the
t Authors
Armando Silva-Afonnso is Prrofessor of Hydraulics at thee
Universityy of Aveiro (Portugal), Departmen
nt of Civil Engineering
E
g.
His speciaalisation is Urban
U
Hydrraulics and Piping
P
Systtems. In thiss
latter fieldd he is working
w
on mathemattical modells, such ass
stochastic models for demand forrecasting an
nd the econoomic designn
b
conceentrating onn
of interiorr networkss. He has recently been
improvingg water-use efficiency
e
inn buildings.
Carla A. Pimentel
P
R
Rodrigues
g
graduated
from
fr
the University of
Aveiro (Poortugal) in Civil Enginneering. Sh
he is currenttly studyingg
for her PhD
D in the areea of water efficiency, again in thee Universityy
of Aveiro. She is a member
m
of thhe technicaal secretariaat of ANQIP
P
– Nationall Associatioon for Qualitty in Building Installattions.
271 2
IV.5
A review of the development and release of
climate change related precipitation data of
relevance to water supply and drainage
systems for buildings
Dr L.B. Jack
l.b.jack@hw.ac.uk
Drainage Research Group
School of the Built Environment
Heriot-Watt University
Edinburgh, Scotland.
Abstract
There is a growing awareness that changes in climate, and in particular the anticipated
variability in temperature and rainfall, predicted across future decades, will have a
significant impact upon the availability of water and the frequency of flood events. It is
also increasingly clear that appropriate responses to these problems demand suitablycast climate change data. In response, the UK government, along with UKCIP (the
Climate Impacts Programme), has recently released probabilistic-based climate
projections based on IPCC (Intergovernmental Panel on Climate Change) greenhouse
gas emissions ‘storylines’.
272 This paper identifies the potential impacts of climate change upon property-based water
supply and drainage networks and reviews the development of, in particular, the release
of precipitation data under UKCIP02 and UKCP09. The paper charts the enhanced
suitability of this data for use as a source to drive simulation models capable of
assessing adaptation strategies for water supply and drainage systems for buildings.
Particular attention is given to rainwater systems and the need for suitably temporally
disaggregated data that corresponds to runoff concentration times.
Keywords
Building, property, drainage, design, climate change
1. Introduction
It is now widely accepted that climate change is likely to have a direct impact upon
society and that this presents fundamental challenges across a broad spectrum of
sectors. In many countries, including the UK, government departments have been
specifically charged with ensuring appropriate responsive action. This action normally
encompasses both mitigation (where reductions in greenhouse gas emissions, or the
activities that result in these, are targeted) and adaptation (where, for example,
engineering or management responses are specifically designed to respond to cope with
predicted changes in climate).
In the UK, the Climate Change Act (2008)[1] outlines legally-binding targets for
reductions in greenhouse gas emissions, as well as detailing specific actions or the
implementation of particular schemes aimed at both mitigation and adaptation.
Interestingly, the Act requires quinquennial reporting of climate change risk in relation
to adaptation. Similarly, in Scotland, the Climate Change (Scotland) Bill (2009)[2],
targets emissions reductions and adaptation strategies whilst also aiming to ensure
sustainable economic growth. Recently, Defra (the UK’s Department for Environment,
Food and Rural Affairs) and UKCIP (the UK Climate Impacts Programme) jointly
recently released new ‘probabilistic-based’ climate change projections, referred to as
UKCP09. Defra, UKCIP and DECC (the recently-established Department for Energy
and Climate Change) envisage this data being used to assist in decision-making
processes across a number of sectors, including that for water.
This paper aims to identify the range of water supply and drainage systems for which
performance is likely to be affected by climate change impacts, and importantly, how
both new and existing methods or data may be utilised in order to develop appropriate
responses. Particular attention will be given to the performance of rainwater systems
and the application of UKCIP09 data.
273 2. Climate change and property-based drainage systems
In determining the potential impacts of climate change upon water supply and drainage
systems for buildings, it is first of all necessary to define the scope of provision referred
to. Within this paper, performance issues at the property-scale will be addressed, thus
encompassing all aspects that normally lie within the curtilage. The potential impacts of
climate change on a number of components parts of this overall provision are outlined
below.
2.1 Water scarcity and the need to reduce the consumption of potable water
Water has always been, and continues to be, a scarce and valuable commodity.
Although there is no evidence that suggests that any one particular drought to date has
occurred as a direct consequence of climate change, the perception is that such events
are likely to become more frequent[3]. This view has refocused attention on the issue of
water conservation – an aspect of water management that has been of concern to those
in the water industry and the affiliated international research community for a number
of decades. Focussing predominantly on the in-building use of water, ie encompassing,
in the main, domestic and commercial usage, it can be seen that driving down
consumption through practice, behaviour and design within both existing and new
properties must be addressed. There are two main components that are required to
implement change. The first is that sanitary appliances within the building must use less
water, whilst retaining high levels of efficacy, and secondly, the overall strategy for the
management of water (of all types, ie rainwater, potable water, groundwater, runoff and
wastewater) must target a reduction in both the demand for potable water, whilst
ensuring that the risk of potential detrimental conditions arising from, for example,
blockage or over-capacity, are avoided.
Much effort has been expended in the drive to reduce water consumption within the
UK, as evidenced, in part, by the introduction of the Water Supply (Water Fittings)
Regulations 1999[4] that target the 30-35% of potable water supplied to households or
for domestic purposes that is used to flush WCs[5]. In addition to a number of
documents with a similar aim, worth specific mention is the CIBSE Guide L[6] and the
Market Transformation Programme (MTP) – a government initiative that provides a
number of briefing notes on water efficiency and performance for a range of
appliances[7]. In addition, the Code for Sustainable Homes[8], introduced in England in
2008, provides a new framework for a mandatory assessment of performance standards
for the overall sustainability of new homes and explicitly encompasses targets for
‘water’ and ‘surface water run-off’.
This progress is encouraging and indicates that potable water consumption levels in the
UK are headed in the correct direction. However, despite the fact that several countries,
worldwide, use far less water than in the UK, there is growing concern, internationally,
that a reduction in water consumption will result in the potential for increased
274 deposition and blockage in associated horizontal, mainly underground, drainage. There
is little, if any, evidence to indicate that such drainline carry problems will occur
however, and it is important to remember that, in curtailing (or attempting to curtail)
water consumption, this merely returns discharge levels to those seen within recent
years and for which many drainage systems were designed. In addition, there is a
substantial body of evidence, much of it based on the work of the affiliated research
community, that shows how flow conditions, solid deposition and any potential for
blockage can all be predicted with accuracy. It should therefore be recognised that any
implications arising from a reduction in in-building water consumption are wellunderstood and that tools exist to predict, with a high degree of accuracy, any risk of
deposition. In addition, these tools allow an exploration of pipework connectivity, with
the aim of identifying the impact of co-incidental or simultaneous flows – an exercise
that can be undertaken with the aim of either re-assessing drainline carry or for the
determination of appropriate network configuration.
With regard to water management within the curtilage, it is also encouraging that
documents such as the Code for Sustainable Homes[8], propose a framework that
addresses different sources and pathways for water. The use of such a framework means
that it is entirely possible, if not probable, that the introduction of greywater and
rainwater harvesting systems will increase as building owners or occupiers are required
to reduce consumption and run-off. For greywater systems, this reduction in water
consumption may be significant, however, it should be noted that their use will clearly
impact upon the overall discharge of wastewater flows to underground drainage.
Rainwater harvesting systems, as a function of their operation, link together the
rainwater drainage provision with that for in-building drainage, and for these systems it
should be noted that previous research publications have shown the importance of
recognising both the time-dependency and the interdependency of property-based water
supply and drainage when determining appropriate adaptation solutions to address
climate change impacts[9].
2.2 The impact of changing precipitation patterns on the performance of rainwater
drainage
The impact of a changing climate on the performance of rainwater systems is likely to
be more direct. Although many conventional, ie gravity, roof drainage systems may
allow for some residual flow that will accommodate more intense storms, siphonic
rainwater systems are designed to suit one particular rainfall event and thus exhibit
little, if any, excess capacity. In reviewing weather patterns, it can be seen that,
conversely, particularly long spells of dry weather may result in an accumulation of
loose debris around the siphonic gutter outlet that, in turn, could be washed into the
small space provided between the baffle plate and the downpipe and causing a sudden
increase in negative pressure. Figure 1 shows some examples of debris accumulation.
275 Figure 1 Debris accumulation around siphonic gutter outlet: dry and wet
Changing rainfall patterns may also have a significant impact upon the performance of
green roofs, where the absorption and evaporation characteristics may be influenced not
only by changes in precipitation but also by varying temperature, wind speed or
direction. Similarly, local pervious areas and SUDs (Sustainable Urban Drainage
systems) may experience changes in saturation levels that mean that runoff
characteristics will differ. These flows can have a significant impact upon conditions in
downstream pipework and can also directly influence environmental factors such as
pollution levels.
3. Climate change impacts and how scenario predictions
Section 2, above, has outlined the range of systems for which climate change may
impact upon performance. Integrating adaptation within the design procedure demands
a two-pronged approach where, firstly, the climate parameters that may alter
performance must be identified and secondly, the change in this variable requires
quantification. The former task is relatively simple, whereby it is clear that
precipitation, temperature, solar radiation, wind and relative humidity will all influence
the performance of one or more of the water supply and drainage systems noted. The
latter, however, requires the application of climatology on such a huge scale that it has
thus far taken large teams of researchers many years to establish complex models that,
only relatively recently, are able to yield predictions based on different greenhouse gas
emissions scenarios.
In 2002, the UK government released, via UKCIP, a number of reports and datasets that
together presented a range of climate change scenarios referred to as UKCIP02[10]. The
UKCIP02 data were based upon four emissions ‘storylines’, where these were labelled
‘high emissions’, ‘medium-high emissions’, ‘medium-low emissions’ and ‘low
emissions’, as mapped onto corresponding IPCC (Intergovernmental Panel on Climate
Change) projections B1, B2, A2 and A1F1 respectively. The climate predictions
embedded within UKCIP02 were based upon analyses by the Hadley Centre and the
276 Tyndall Centre. The Hadley centre is a leading centre for climate change research that is
funded mainly by Defra and DECC and operates as part of the UK Meteorological
Office. The Tyndall Centre is also a leading centre for climate change research within
the UK, and brings together a range of disciplines from the academic sector whose aim
is to develop outputs that will positively influence both mitigation adaptation strategies
to address climate change.
The anticipated increase in temperature – particularly in summer months and in the
south-east of the UK, is well-documented and is confirmed by UKCIP02 data.
Similarly, changes in precipitation patterns are likely to mean a notable decrease in
summer rainfall and an increase in winter precipitation (again, especially in the southeast of the UK). It is anticipated that average wind speeds are likely to become higher in
winter in central and southerly areas of the UK, however, it is recognised that changes
in wind speed with climate are notoriously difficult to predict and thus, any projections
carry with them a high degree of uncertainty. Changes in solar radiation and relative
humidity are similarly difficult to predict, although it is envisaged that cloud cover is
likely to decrease (and solar radiation increase) in summer, and that relative humidity
may decrease throughout the year.
Using a high-resolution model driven by data derived by the Hadley simulation and
based on the IPCC A2 (medium-high) storyline, the UKCIP02 projections were able to
provide information on climate using a grid size of 50km. Pattern-scaling methods were
then used to determine climate data for the three remaining IPCC storylines.
Figure 2 shows two examples of typical UKCIP02 illustrations. The first shows the
predicted summer temperature change for the 2020s, 2050s and 2080s timescales, and
the second the predicted summer and winter precipitation change for the 2080s (both for
the medium-high emissions storyline).
Figure 2a UKCIP02 predicted summer temperature change, ‘medium-high
emissions’. Source: UKCIP02 [10]
277 Figure 2b UKCIP02 predicted summer and winter precipitation change, ‘mediumhigh emissions’, 2080s. Source: UKCIP02 [10]
It is worth noting that changes are always presented relative to baseline data observed
between 1961 and 1990. This ‘change’ or ‘shift’ is often referred to as the ‘anomaly
value’.
It will be appreciated that this information provided a step change in the quantification
of climate variables, particularly rainfall and temperature, that the design engineer for
water supply and drainage might wish to adopt. Less clear however, was an indication
of which particular ‘storyline’ global emissions are likely to follow. Thus, embedded
within any design-making process that uses quantified climate change predictions, is an
assessment of the most applicable scenario – an impossible dilemma given that this will
be determined by choices governed by societal changes such as population growth,
economic growth and technological advancement! Sensibly, however, UKCIP advice
was that more than one scenario is used in order to yield a comparison of possible
change or a scope across which adaptation strategies might differ. In addition, the data
was envisaged as being of use in long-term strategies, and caution when using data was
advised, as the basis for projections, and their limitations and associated uncertainty,
must be fully understood.
Following the release of the IPCC’s fourth report[11], UKCIP and Defra published, in
June 2009, what is referred to as UKCP09 – UK Climate Projections, 2009. These data
are set against three, rather than four, emissions scenarios – termed ‘high’, ‘medium’
and ‘low’, and comprise ensemble data, based on information from the Hadley Centre
model simulations combined with data from other IPCC single-model simulations. This
progression to the release of data in ensemble form, rather than a single result for each
emissions scenario, yields information on a range of possible outcomes, each with an
estimated probability of occurrence.
278 Essentially, the climate change predictions have changed little from those published in
UKCIP02, however, the presentation of information in probabilistic terms provides an
important step forward in the potential applicability by designers to the assessment of
adaptation strategies. The information is also now presented using a grid of resolution
25km, rather than 50km.
Figure 3 Illustration of UKCP09 map of climate variable: % change in winter
mean precipitation (high emissions, 50% probability)
UKCIP provide two routes through which this data may be accessed. The first presents
key findings and mapped information, whereas the second offers the user access to more
technical information and a ‘weather generator’. For the former, a wide range of maps
are provided that represent climate variables of temperature, precipitation, relative
humidity and cloud cover. Wind data is not presented. Figure 3 illustrates the form of
these maps, showing the UK-wide map for the percentage change in winter mean
precipitation for the high emissions scenario.
There is a key difference between the presentation of climate change information in the
UKCIP02 and UKCP09 reports. Whereas in UKCIP02, specific data was presented,
using maps, for variables set against different emissions scenarios, temporal average
and across a 30 year timescale, the UKCP09 maps show changes in climate variable,
also linked to emissions scenarios and timescale, but that relate directly to the
probability associated with the variable under consideration. It is important to note that
the probability values included are cumulative. This means that, for the example shown,
279 where the probability is noted as 50% (referred to as the central probability), the model
output suggests that the change in climate variable is just as likely to be greater than that
shown, as it is likely to be less than. However, with an allocated probability factor of
50%, these values do not necessarily represent the most likely outcome.
This point is important, as proper interpretation of the datasets rests on this
understanding. In addition, this point emphases how the maps provided as part of
UKCP09 are not weather maps, and therefore do not yield, for example, precipitation
data that can be directly applied in the design of water supply and drainage systems for
properties. This confirms the need for an important component of UKCP09, referred to
as a ‘weather generator’. This tool uses stochastic downscaling methods to generate
single-site synthetic daily and hourly time series data with a spatial resolution of 5km.
As part of UKCP09, the variables downscaled by this weather generator include
temperature, rainfall, humidity and sunshine. The key determinant is the rainfall model,
embedded within which is the Neyman-Scott Rectangular Pulse approach – a
mechanism that is widely accepted as providing the foundation for replicating both
clusters and extremes of rainfall[12]. This provides exceptionally useful information,
with the proviso that, as with UKCP09, the application of data is thoroughly understood
and that the inherent limitations of this use are recognised in full[13]. Inherent in this
should be recognition of the fact that the degree of uncertainty associated with these
datasets at a temporal resolution of less than daily, becomes significant.
4. The application of climate change projections for the assessment of
adaptation strategies
Despite these step-changes in the way in which projected rainfall data may be used to
assess appropriate adaptation solutions for water supply and drainage systems for
buildings, it will be appreciated that temporal resolution remains a key consideration.
For example, the typical concentration time for run-off from a roof surface is two
minutes. In addition, the response of rainwater systems – conventional or siphonic –
will inevitably be determined by changes in rainfall patterns of duration shorter than one
hour. Thus, there is a need to disaggregate data yet further, and to develop awareness in
the associated uncertainties of using such information.
However, such a move would be exceptionally beneficial to designers, particularly since
current practice is based upon the application of steady-state principles linked to rainfall
intensity defined by return period. Ongoing work at Heriot-Watt, funded by the
Engineering and Physical Sciences Research Council, endeavours to develop a
framework for use in assessing the efficacy of climate-change initiated adaptation
strategies. A key component in this is the provision of rainfall information temporally
disaggregated to a resolution of 15 and 5 minute intervals and presented as synthetic
times series data. This component of the work will be undertaken by researchers based
at Newcastle University (UK) and similarly represents a step change in the way in
280 which climate change data is, and will continue to be, used in the design of systems for
buildings. As noted above however, a key factor inherent in the use of this data will be
an appreciation of the limitations of this degree of resolution and of the associated
uncertainty. Nonetheless, using climate projections cast in probabilistic terms and
disaggregated to suit system demands will undoubtedly facilitate limited but improved
and realistic predictions of performance that will avoid over-design to accommodate
climate change.
5. Conclusion
In concluding, it is perhaps worth noting that the UKCIP02 publication outlines how
our future climate will, inevitably, vary annually, and how much of this variability is,
and will continue to be – for some time at least – entirely natural. There is a clear
prediction however, that, by the 2050s, ‘most changes in average climate as described as
due to human activities are likely to greatly exceed the natural variability of the UK
climate’. This should leave no doubt of the necessity to incorporate predicted climate,
and in particular precipitation, changes within the design process for water supply and
drainage systems.
This paper has provided a brief overview of the components of these systems that may
be directly affected by climate change impacts, and has outlined how the challenges
presented by climate change impacts increasingly demand appropriate responses from
those charged with design and specification.
It has highlighted how, currently, reference sources for designers rely heavily upon the
use of return periods and corresponding rainfall intensity figures. These intensity figures
are, in turn, applied within steady state calculations assumed to be representative of
system behaviour. It is suggested that the newly-released probabilistic climate change
data allows designers better scope to assess the risk associated with a particular design,
for any given building in a pre-defined geographical location. It has reviewed the
development of climate change predictions and outlined the step changes observed in
the availability and applicability of data (and in particular, precipitation data) that will
be of direct use to designers in assessing appropriate adaptation strategies for water
supply and drainage systems for buildings.
281 6. References
1. UK Climate Change Act, Department for Environment, Food and Rural Affairs,
Published by: The Stationary Office Ltd, 2008.
2. Climate Change (Scotland) Bill, Scottish Parliament, 2009. Published document
available from: Office of the Queen’s Printer for Scotland.
3. Environment Agency, Drought explained. Available at:
http://www.environment-agency.gov.uk/homeandleisure/drought/default.aspx
4. The Water Supply (Water Fittings) Regulations 1999, Statutory Instrument No. 1148
5. Wise A.F.E., and Swaffield J.A., Water, sanitary and waste services for buildings, 5th
Edition, Butterworth Heinemann, 2002.
6. CIBSE (Chartered Institution of Building Service Engineers), Guide L:
Sustainability. 2007, ISBN 978-1-903287-82-8
7. Market Transformation Programme – Supporting UK Government policy on
sustainable products. http://www.mtprog.com/
8. DCLG (Department for Communities and Local Government) Code for sustainable
homes, Technical Guide, 2007. Available at:
http://www.planningportal.gov.uk/england/professionals/en/1115314116927.html
9. Jack L.B., ‘Assessing the impact upon property-based water supply and drainage
systems of rainfall events predicted by climate change scenarios’, CIBW62 Water
Supply and Drainage for Buildings Symposium, Hong Kong, 2008.
10. Hulme M., Jenkins G.J., Lu X., Turnpenny J.R., Mitchell T.D., Jones R.G., Lowe J.,
Murphy J.M. Hassell D., Boorman P., McDonald R., Hill S., Climate change scenarios
for the United Kingdom: The UKCIP02 Scientific Report, Tyndall Centre for Climate
Change Research, School of Environmental Sciences, University of East Anglia,
Norwich, UK, 2002.
11. IPCC (Intergovernmental Panel on Climate Change) Fourth Assessment Report,
AR4, 2007
12. Cowpertwait P.S.P., O’Connell P.E, Metcalfe A.V. and Mawdsley J.A., Stochastic
point process modelling of rainfall. 2. Regionalisation and disaggregation. Journal of
Hydrology, 175, 47-65, 1996.
282 13. Jones P.D., Kilsby C.G., Harpham C., Glenis V., Burton A., UK Climate Projections
Science report: Projections of Future Daily Climate for the UK from Weather
Generator, University of Newcastle, UK, 2009.
7. Presentation of Author
Dr Lynne Jack is Director of Research in the School of the
Built Environment and has been a member of the Drainage
Research Group at Heriot-Watt University since 1993. Her
research interests include the simulation of air pressure
transient propagation in building drainage ventilation
systems and the assessment of property drainage system
performance when subject to climate change impacts.
283 Session V: Drainage I
V.1
Empirical study on terminal water
Velocity of drainage stack, Part 2
(1) C.L. Cheng, Dr. (2) W.J.Liao, Ms. (3) K.C. He, Dr. (4) J.L.Lin, Ms.
(1) CCL@mail.ntust.edu.tw
(2) D9613011@mail.ntust.edu.tw
(3) D9513001@mail.ntust.edu.tw
(4) M9613011@mail.ntust.edu.tw
(1) (2) (3) (4) National Taiwan University of Science and Technology, Department of
Architecture, 43 Keelung Road Sec.4, Taipei, Taiwan, R.O.C.
Abstract
Due to the importance of permit flow rate regulation which is adopted in many
countries, the terminal velocity in drainage stack was seen as one of the crucial issues in
building drainage studies. Several theories and predictions were reported in previous
researches from 1960s. An empirical study on terminal water velocity of drainage stack
was explored from 1996 and reported in 2008 CIBW62 conference HK. However, the
validation issues were remained and need to be further clarified. This paper would
continue the methodology with empirical approach by theoretical study from air
pressure distribution mechanism. Meanwhile, a technology with digital high-speed
video camera and statistic tool will be used to validate the calculation result of terminal
velocity of drainage stack in this research. The results confirm that the theoretical study
can fit the practical sense and the validation also can response the calculation results.
Accordingly, the terminal velocity of drainage stack should be reconfirmed and
redefined its accurate value under the new validation evidences. Furthermore, the long
term regulation and application which is according to the lower terminal velocity and
permit flow rate also need to be reconsidered.
Keywords
terminal velocity, gravity acceleration, high-speed video camera, statistic methodology
284 List of symbols
symbols
content
unit
Qw
Water flow rate
l/s
Qa
Air flow rate in stack vent
m3/s
R
Diameter of stack
m
Vw
Velocity of water flowing
Va
Velocity of airflow
Vt
Terminal velocity of water
w
The coefficient of water resistance in stack

The coefficient of air resistance in stack
a
m/s
m/s
m/s
m/s2
g
Gravity acceleration
t
time interval

Distance
m
SD
The accumulation distance of falling water
m
sec
1. Introduction
An innovation study which explored the terminal velocity in drainage stack was
reported in 2008 CIBW62 annual conference HK. This research continued the remained
issues regarding the experiment and validation process. A high-speed digital video
camera and statistic sampling methodology were adopted to validate the previous
calculation results. This report includes a summary of proposed calculation
methodology and its results. Accordingly, we improved the verification by statistic
sampling process to offer more accurate evidence for the validation.
As the technical reviews, Hunter1) explored the flow phenomenon of drainage stack
in1940s. Afterward, Wyly2)3) & Dawson first issued the theory of the terminal velocity
at 1960s. According to these pioneer research results, the guideline of National
Plumbing Code (NPC) of US was used to set the permit flow rate as the regulation of
drainage system. Following initial work of the HASS 203 of Japan in 1970s, the method
285 of steady flow condition was merged as the provision reference and evaluation
technique, hence, they conducted series researches of steady flow methodology and
greatly contributed to the application of building drainage network. Due to the
importance of permit flow rate regulation, the terminal velocity in drainage stack was
seen as one of the crucial issues in these series researches. Several theories4)5) and
predictions were reported in previous researches. However, the validation and accuracy
were still criticized so far and did not reach the persuasive results. The new evidence
and validation are expected and need to be conducted nowadays.
2. Theoretical Reviews
The theory of the annulated flowing in drainage stack was first reported by Wyly2) in
1960s. Afterward, some researches tried to figure out the velocity of flowing water in
stack by the experimental method and theory, however, no firm results were reported in
that period. In 1980s, Tukagoshi6) conducted electricity to the salt solution in Japan, and
put the sensor of the electricity into the pipe which perpendicular to the pipe’s section
and divided into 1-25 points as observational points, when salt solution flowing into the
vertical stack and pass through the sensor would evaluated the velocity and quantity of
the water flowing. In 1994, Sakaue7) in Japan continuously infused water into vertical
stack for testing the velocity of the water flowing, and to return to original equation for
evaluated the water flowing rate in the vertical stack. However, all these researches
have not reached a clarified and validated conclusion on the terminal velocity of
drainage stack.
According to the previous researches8)9)10) on air pressure distribution, the airflow rate
(Qa) was identified as a critical parameter for a prediction model which can
approximately figure out the falling water phenomenon in vertical drainage stack. The
mechanism of flowing phenomenon within vertical drainage stack is now schematically
understood. Air pressure in vertical drainage stack is caused by series interactions
between downstream water and through-flow air in vertical pipe. Fig.1 illustrates the
image of flow state and the modified interaction, thus it conducts the main parameters
with air pressure, airflow rate, and resistance coefficients, and they are the essential
factors for prediction model of air pressure distribution in vertical drainage stack. These
researches also provided the possible viewpoint to explore the flowing velocity of air
and water in drainage stack. Consequently, the innovative study on terminal velocity
was restarted from 1996. Afterward, an initial study results with empirical approach was
reported in 2008 CIBW62 conference HK. However, the validation issues were
remained and need to be further clarified.
286 Fig.1 Interaction mechanism of water and air in drainage stack
The essential phenomenon of falling water and interaction mechanism in drainage stack
was described in our last report. Fig.1 shows the image of interaction mechanism of
water and air in drainage stack. The velocity of the falling water in the stack is mainly
dominated by the three interaction balance including gravity force (g) and friction drag
of the pipe inside and the air interaction toward the falling water which we mentioned in
our last report. When the falling water in the stack reaches the terminal velocity
situation, which means the interaction inside the stack reaches the condition of balance
and the forces actions are totally equal and neutralized. Herein, we omitted the
conduction details of theoretical process and summarized the result of terminal velocity
function as equation (1). The details of the theoretical study could be referred from our
last report in 2008 CIBW62 proceedings11).
Vt 
g
w
………..(1)
287 3. Experiment and observation
Following the development of observation technology, this research used a digital highspeed video camera to validate the calculation model for terminal velocity. According to
previous report in 2008 CIBW62 conference HK11), an initial experiment was executed
to validate the theoretical terminal velocity. The further precise verification process is
necessary to be conducted in this research. Due to the phenomenon of falling water in
stack is complex and random by time sequence, the improvement of accuracy is a
crucial part for the reliability of validation.
In order to improve the reliability of validation results, a statistic methodology of
sampling conception for random observation is adopted in this report. The same as our
previous report last year, Table 1 is the specification of this digital high-speed video
camera which is used to observe the falling water velocity in stack. Fig.2 is the picture
of experiment in observation place which shows the circumstance and condition of the
operation.
Table 1 Specification of the digital high speed camera
MEMORY
ELECTRICAL CORE COMPUTER LINKS DIAPHRAGM Video camera
VIDEO CAMERA Auto Exposure Control,Color or monochrome
2,100 pictures per second full resolution
Software:
Acquisition, Analytical playback, Measurements,
Image processing and File management
256 mg RAM,for files memory
Diaphragm
Adjustable diaphragm.
Resolution of the screen
288 Fig.2 The experimental device and operation picture
The pictures of the transparent pipe at each floor of the experimental tower were taken
by high-speed video camera. Water discharges are all from 12th floor with the water
flow rate of 1.0 l/s, 2.0 l/s, 3.0 l/s, 4.0 l/s, meanwhile each floor divides into 3 layers so
that each floor can be taken video with three times. This research totally got 128 video
data from the observation of 32 layers. The experimental device includes digital high
speed camera, two lamps, notebook for recording the data and high resolution image
screen. Fig. 3 shows the devices of this observation.
Fig.3 The experimental tower and testing device profile in NTUST
289 According to the visual observation by transparent pipe, the phenomenon of falling
water is complex and no possible to identify its character from the picture. As we slow
down the picture by the high-speed video camera, the behavior of falling water reveals
its feature visually in the picture. An image of falling water picture in stack is as shown
as Fig. 4. Firstly, we can see the annual water along with the surface of stack with a
comparable slow velocity. The other obvious part can be identified is the chunk body of
falling water with comparable high speed velocity mostly in the middle of the stack.
The last obvious feature part can be seen is the drop water with various falling velocity
in the picture.
Fig.4 An image of falling water picture in stack
According to the statistic conception for the sampling process, at least 385 observation
samples are necessary for the accuracy of calculation under 95% reliable level and 5%
permit inaccuracy. Therefore, 129 observation samples are picked from individual part
of these three water features and totally 387 samples are taken into the calculation for
one observation point.
In order to catch the individual sample of water, clear black lines are set by 10 mm
interval on the screen picture to trace the moving of falling water. Accordingly, the
velocity of each sample can be calculated by the moving distance and time interval. The
calculations include average velocity and deviation for three water features. Fig.5 shows
the process flow and details of sampling method. We used the grid meshes to specify
the water sample and trace its moving velocity. Fig.6 shows a case of the different time
location which we pick 25 and 26 recognizable samples for tracing their moving
behaviors. All the velocity of the recognizable water drops can be exactly calculated by
its moving distance and time intervals for each observation point in experiment tower.
As the above mentioned about experiment device, the discharge water is from 12th floor
height, and three observation points are set in each floor area. Accordingly, there are
290 totally 32 observation points and 12 384 selected samples for calculation in this
empirical validation study.
Fig.5 The details of sampling method
291 time 16:42:53.036
Original
image and
pick
25
samples
time 16:42:53.040
Original
image and
pick
26
samples
samples
pick
51
samples
Fig. 6 The sampling process of water drop
The experimental condition for observation is a continuing discharge and repeatable
phenomena in stack. Thus, the profile of water velocity can be expressed as the
distribution of air pressure in stack. As the results, the velocity of each observation
point can be produced from the sampling process and Fig. 7~10 shows the holistic water
velocity profile of water discharge from 12th floor height.
292 Fig.7 W
Water veloccity in stack
k for three Fig.8 Waater velocitty in stack for three
water featu
ures (1.0 l//s)
w
water
features (2.0 l/ss)
Fig.9 W
Water veloccity in stack
k for three
water feattures (3.0 l//s)
293 2
Fig.10 Water vellocity in sta
ack for
threee water features (4.0 l/s)
Fig.11 Water velocity and air
pressure in stack (1.0 l/s)
Fig.12 Water velocity and air
pressure in stack (2.0 l/s)
Fig.13 Water velocity and air
pressure in stack (3.0 l/s)
Fig.14 Water velocity and air
pressure in stack (4.0 l/s)
294 The velocity profiles in figures are including average of individual part of these three
water features which are annual water along with pipe and the chunk body of falling
water and the drop water with various falling velocity in the picture. Meanwhile, the
discharge flow rate are including 1.0 (l/s), 2.0 (l/s), 3.0 (l/s) and 4.0 (l/s). The results
reveal that the terminal velocity increases following the water flow rate and obviously it
shows 4.0 (l/s)>3.0 (l/s)>2.0 (l/s)>1.0 (l/s). Fig.11~14 shows the average velocity for
each observation point and the simultaneous air pressure distribution in stack. It is seen
that the terminal velocity approximately happen in C zone of the air pressure
distribution. That also matches the theoretical conception and the initial assumption of
this research. Namely, the results can validate the accurate terminal velocity of drainage
stack by this empirical approach.
4. Validation and discussion
The previous report in 2008 CIBW62 conference introduced an empirical methodology
to calculate the terminal velocity in drainage stack. The calculation process and results
are summarized in Chapter 2. On the other hand, the terminal velocity is concluded
from the observation by high-speed video camera and statistic sampling process in
chapter 3. The comparison of the results from the different approach is show in Table 2
and Fig. 15. It reveals that the results of individual water flow rate (1.0, 2.0, 3.0 and 4.0
l/s) are approximately closed and validation is reliable. We also compare the positions
of maxima air pressure by calculation and observation. The position of maxima air
pressure is expressed by the distance from discharge floor. The results are shown in
Table 2. It also reveals that the range is approximately matched for the two approaches.
Table 2.the result of the terminal velocity
Water flow The
rate(QW) theoretical
value (m/s)
The
experimental
value (m/s)
The
theoretical The
experimental
value
of
max value
of
max
negative point (m)
negative point (m)
1.0
4.60
4.39
16
13~15
2.0
5.07
4.80
15
12~13
3.0
5.57
5.18
12
10~12
4.0
6.00
5.50
10
8~9
295 Velocity(m/s)
7
6
5
4
3
2
The theoretical value
1
The experimental value(standard deviation)
0
0
1
2
3
4
Qw(l/s)
Fig15. The comparison of calculation and experiment of the terminal velocity
The theory of the annulated flowing in drainage stack was first issued by Wyly2) in
1960s as mentioned in Chapter 2. Regarding the researches of terminal velocity,
Dawson (US) used the Manning equation and conducted calculation model to determine
the terminal velocity by the following equation (2).
0 .4
 Q 
Vt  5.18   w  ………..(2)
2 R
Meanwhile, Wyly (US) continued to explore the falling water in stack and assumed the
flowing mechanism to be annual membrane flow. Consequently, he amended the
coefficient and proposed the calculation equation for the terminal velocity as the
following equation (3).
0.4
 Q 
Vt  4.02   w  ………..(3)
2 R
This paper compares the theoretical calculation results as shown as Table 3 and Fig.16.
The results reveal that the calculated terminal velocity in this research is much greater
than the results which were done by Wyly and Dowson.
296 Table 3 the terminal velocity of water flowing in stack
Previous researches
This study
Wyly type Dowson type
(the experimental value)
1.0 l/s
1.60
2.06
4.39
2.0 l/s
2.11
2.06
4.80
3.0 l/s
2.48
3.20
5.18
4.0 l/s
2.79
3.60
5.50
Velocity(m/s)
Water flow rate Qw
7
6
5
4
3
2
1
Wyly
Dawson
The experimental value(standard deviation)
0
0
1
2
3
4
Qw(l/s)
Fig.16 Comparison of terminal velocity values
It is obviously that the major reason for the different terminal velocity between present
research and previous documents is the calculation assumption. Wyly and Dowson’s
researches assumed that water flow in stack is similar to the open channel with annual
membrane flow in stack and adopted the Manning equation to estimate the terminal
velocity. However, the visual observation significantly shows that the real flow in
drainage stack is more complex and dynamic than annual flow assumption.
Meanwhile, according to the visual observation and the calculated velocity in this
research, it is noticed that even the annual flow along the stack has greater value than
previous researches. Fig.17 shows the comparison of the annual flow velocity. Namely,
the terminal velocity in drainage stack should be reconfirmed nowadays and need to
297 Velocity(m/s)
redefine its accurate value under the new validation results. Furthermore, the long term
regulation and application which is according to the lower terminal velocity and permit
flow rate are also necessary to be reconsidered.
7
6
5
4
3
2
Wyly
1
Dawson
annular water(standard deviation)
0
0
1
2
3
4
Qw(l/s)
Fig.17 Comparison of the annual flow velocity
5. Conclusion
According to the importance of permit flow rate regulation, the terminal velocity in
drainage stack was seen as one of the crucial issues in building drainage researches.
This paper introduces a calculation method with empirical approach for terminal
velocity by theoretical study from air pressure distribution mechanism. An observation
technology with digital high-speed video camera was used to validate the prediction of
terminal velocity of drainage stack in this report. The theoretical study reveals that the
practical sense and the validation approximately responses to the calculation results.
The guideline of National Plumbing Code (NPC) of US was used to set the permit flow
rate as the regulation of drainage system. Following initial work of the HASS 203 of
Japan in 1970s, the method of steady flow condition was merged as the provision
reference and evaluation technique. As the results of this research, the terminal velocity
in drainage stack should be reconfirmed nowadays and need to redefine its accurate
value under the new validation results. Furthermore, the long term regulation and
298 application which is according to the lower terminal velocity and permit flow rate are
also necessary to be reconsidered.
Acknowledgements
The authors would like to thank the Architecture & Building Research Institute of the
Ministry of the Interior of Taiwan (ABRI) and the National Science Council of the
Republic of China (NSC97- 2221-E-011-112)for financially supporting this research.
7. Reference
1) Roy.B.Hunter;BMS 79 Water Distributing System for Building, (1941)
2)R.S.Wyly and H.N.Eaton;Capacities of Plumbing Stack in Building,BMS
Repoet,132(1952)
3) R.S. Wyly and H.N. Eaton : Capacities of Stacks in Sanitary Drainage System for
Building, N.B.S. Monograph 31, (1961)
4)B.J.Pink;A Study of Water Flow in Vertical Drainage Stacks by Means of a
Probe,CIB-W62 Semminar,(1973)
5) Yoshiharu Asano;The basic research of the specific of the velocity in vertical
stack---terminal velocity, the report of the architectural institute of Japan,
278(1979)
6) Tukagoshi;The experimental research method of the specific of the vertical
stack,Transactions of the Society of Heating, Air-Conditioning and Sanitary
Engineers of Japan (1977)
7) Sakaue: The analysis of the variation of the pressure in vertical stack,
Transactions of the Society of Heating, Air-Conditioning and Sanitary Engineers
of Japan (1979)
8) Cheng, C. L., Kamata, M., Kurabuchi, T., Sakaue, K., Tanaka, T., “A Prediction
Method of Air Pressure Distribution of Drainage Stack System in Case of SinglePoint Steady Discharge”, Journal of Archit. Plann. Environ. Eng., No.481, pp8391. (1996).
9) Cheng, C. L., Kamata, M., Kurabuchi, T., Sakaue, K., “Study on Pressure
Distribution of Drainage Stack System in High-Rise Apartment Houses”, Journal
of Graduate School and Faculty of Engineering the University of Tokyo (B), Vol.
XLIII, No.4, pp467-489. (1996, EI)
299 10) C.L. Cheng, Lu, W. H., Shen, M.D., An Empirical Approach: Prediction
Method of Air Pressure Distribution on Building Vertical Drainage Stack,
Journal of the Chinese Institute of Engineers, Vol 28.(2004)
11) C.L. Cheng, K.C. He, C. J. Yen, W.J. Liao, Empirical study on terminal water
velocity of drainage stack, CIB-W62 International Symposium, Hong Kong
(2008.09).
Presentation of Author
Cheng-Li Cheng is the Professor at National Taiwan
University of Science and Technology, and ex-chairman of
Department of Architecture. He is a researcher and
published widely on a range of water supply and drainage
in building. He has published extensively on a range of
sustainable issues, including the water and energy
conservation for green building. Currently he also acts as
referee and member of Taiwan Green Building Evaluation
Committee and National Building Code Review
Committee.
300 V.2
Dry drains: myth, reality or impediment to
water conservation.
Professor J.A. Swaffield FRSE
j.a.swaffield@hw.ac.uk
Emeritus Professor, School of the Built Environment, Heriot Watt University,
Edinburgh, Scotland
Abstract
Waste solid transport from the appliance to the sewer connection is a fundamental
requirement of a successful building drainage system. Reductions in appliance
discharge volumes have raised concerns that this drain line carry is being severely
threatened, leading to the emotive use of the term ‘dry drains’ to encapsulate a
perceived problem that may or may not be real.
This paper will review the demographic and climatic change drivers that have led to the
current international emphasis on water conservation. The central importance of wc
flush volume in the water conservation debate will be reviewed, as will the efforts to
reduce the water usage attributed to wc operation from internationally. The fundamental
prerequisites of low flush wc design will be revisited, including the importance of the
percentage of the flush discharged behind the solid. Similarly the strategies available to
enhance solid transport will be reviewed, including both the simplistic ‘reduce drain
cross sectional area and increase drain slope’ options and the more sophisticated
solutions based on the introduction of flow boost appliances. In addition the importance
of wc flush only transport to the first flow confluence will be emphasised.
An argument will be presented to support a concerted research review to identify the
reality of the concerns being expressed over solid transport that would identify for the
first time the necessary transport distances and strategies to allow successful transport
without reneging on the internationally agreed commitment to reduce water
consumption. This proposed study will require historic comparative water usage data as
301 well as a summary of the architectural details that determine minimum distances to
reach junctions carrying additional flow. Strategies to ensure solid transport will be
supported by simulations utilising the Heriot Watt University DRAINET model, as well
as arguments based on the substantial body of research reported through CIBW62 over
the past 30 years.
Keywords:
Waste solid transport, climate and population drivers, water conservation.
1.
Introduction.
The concept that ‘dry drains’ could be a probable consequence of water conservation
arose in the US where ‘concerned groups’ argue that further reductions in w.c. flush
volume should be resisted as current water conservation measures reduce the
throughflow in the drains to the extent that ‘drainline carry’ becomes impossible.
In March 2009 a Dry Drains Forum was hosted at the ISH exhibition by Jeff Patchell
and speakers presented arguments for and against the concept. The US contribution
harked back to the 1992 Energy Bill that introduced 6 litre w.c. flush volume without
consultation and was concerned that any move towards 4 litres would be achieved in a
similar fashion. The UK and Australian contributions emphasised the role of climate
change and our ability to design for reduced flush volume – quite different approaches.
It is necessary to rationally examine the concept of the ‘dry drain’, -
1.
2.
3.
4.
What effect will climate change have on drain throughflows?
What effect will demographic and social changes have on throughflow?
What can we learn from previous research?
What design or installation strategies might be used to offset any perceived
problem?
While it may be attractive to begin again by defining a series of research objectives –
for which funding would have to be sought in a climate where arguably other priorities
may dominate – including the correlation of laboratory results to field observation and
an assessment of the feasibility of CFD, it is more prudent to start by reviewing relevant
past research – i.e. ‘because you don’t know the answer does not mean that nobody else
does’. Proceeding to ‘develop Request for Proposals to be sent to universities for CFD
work, qualified test laboratories for laboratory work, and plumbing contractors for field
studies to determine the costs so that funding can be sought from the US government
and other entities’, (Private communication, Pete DeMarco, IAPMO April 24th 2009),
while delaying ‘an international literature search pending funding’, appears misguided.
302 This paper will address the four issues above and will suggest a way forward drawing
on previous research and providing a route for collaboration with and through CIBW62
as the organization with the required background experience and publication record.
2.
What effect will climate change have on drain throughflows?
UK government predictions for climate change, June 2009, aimed at industries and
organisations that need to make long term investment decisions that could be influenced
by climate change, suggest that summer rainfall in SE England could fall by 19% by
2050 and possibly (1 in 10 chance) by 41%. Winter rainfall in the West of Scotland
will rise by 15% by 2050 and possibly (1 in 10 chance) by 29%. In launching these new
scenarios Hilary Benn, the UK Environment Secretary said that ‘… climate change is
the biggest challenge facing the world today .. this landmark scientific evidence shows
that we need to tackle the causes of climate change and deal with its consequences…’,
(Guardian report of Benn’s presentation, June 19th 2009).
Figure 1:
Water stress areas in England, Environment Agency in 2007,
showing a North Westerly expansion compared to 2001.
303 Figure 2: Met. Office data as to the rainfall disparity across the UK.
These predictions extend the concern at the water stress assessments in 2001 and 2007
undertaken by the Environment Agency. Figure 1 illustrates the more recent mapping.
It should also be noted that the centres of current population, particularly in the South
East of the UK, lie well within the water stress areas. The new scenarios strengthen the
already known disparity in rainfall across the UK, Figure 2 illustrates recent findings
that show a fourfold difference from North West to South East.
Climate change will introduce wetter winters and higher summer temperatures for the
UK. Water availability will become more problematic and water conservation will
continue to be a major issue for government, industry and the domestic and commercial
building owner. Thus water conservation will continue as a major priority for the UK
that has to be reflected in appliance design and user perceptions.
3.
What effect will demographic and social changes have on
throughflow?
Demographic and user preferences have changed the water use pattern in the UK over
past decades and in the context of the ‘dry drain’ it is useful to determine whether less
water is used now than previously. Publicly available data is instructive, Figure 3,
•
The number of households in England is projected to grow to 27.8 million in 2031,
an increase of 6.3 million (29 per cent) over the 2006 estimate, or 252,000
households per year.
304 •
One person
p
housseholds are projected to
t increase by 163,0000 per year, equating too
two-thhirds of the overall increase in hou
useholds.
•
By 20031, 32 perr cent of hoouseholds will
w be headded by thosse aged 65 or over, upp
from 26 per cent in 2006.
•
By 20031, 18 per cent of thee total popu
ulation of England
E
is pprojected to live alone,,
compared with 13 per cent in 2006.
d
a populattion growthh and the risse in singlee
The combbination of demograph
ic change and
occupancyy suggests that water use will rise rather thhan fall in the immed
diate future,,
Figure 4. It is thereffore useful to attempt to determinne likely w
water usage figures andd
des.
how thesee have changged over thee past decad
In 1976 thhe per capita per day water use in
i the UK was estimaated as 115 litres pcpdd
rising by 2001
2
to 1433 litres pcpd while thee populationn grew by aat least 10%
%, Figure 5..
Current prroposals byy the Departtment for Communitiess and Locall Governmeent (DCLG))
within thee Part G Buiilding Reguulations prop
poses a figuure of 125 liitres pcpd – effectivelyy
a rise cf. 1976.
1
Research identifies thhe appliancces contribu
uting to these figures - the w.c. iss the majorr
i
ally yieldinng a figure close to 33%
3
of thee
contributoor with histtoric data internationa
domestic water use and
a somew
what greater in the com
mmercial buuilding areaa, eg. - UK
K
domestic, Griggs annd Shouler (1994) – 32%, UK commerciaal building `domestic',,
4
Austrralian domeestic, Cox (1997), - 32
3 %, US
S
Griggs annd Shouler (1994) - 40%,
domestic water,
w
AWW
WA Researcch Foundatiion (1999), - 23%.
Figu
ure 3:
Demograph
D
hic predictiions drawn
n from the ccurrent Ho
ousing
Staatistical Reelease 11 March
M
2009 DCLG.
http:///www.statistics.gov.uk
k/statbase/P
Product.assp?vlnk=99
97
305 Figure 4:
Water use per capita per day dependent upon number of coresidents.
Thus demographic changes will increase water usage and the highest % of water use
will remain through the w.c. These results give a clear guide as to the area requiring
action – it continues to be the w.c. in terms of usage and design.
Figure 5:
UK Water use per capita per day 1976 – 2006
306 4.
What can we learn from previous research?
Knowledge of previous research identifies the following priority areas and activities –
•
•
•
•
W.c. design is central,
Identify and understand the parameters defining solid transport and develop
predictive models,
Introduce careful drainage system design to minimise probability of deposition,
Understand and apply 30+ years of research freely available through
international published papers.
Water Conservation will remain a major issue for the foreseeable future. Savings
cannot be achieved by merely relying on a change in public attitudes to water usage –
hoping to change the habits of a population is questionable at best. The solution must
lie in a combination of user attitudes and the provision of improved appliance design so
conservation ‘just happens’ - water conservation with no life style change. Figure 6
illustrates the application of this principle to w.c. design over the past 100 years.
In addition to an understanding of the design parameters governing advanced w.c.
design for efficient low flush volume operation, the mechanisms of solid transport
within the drainage network must also be fully understood. CIBW62 since its first
meeting in 1973 has considered this theme with a full range of international
contributors. Figure 7 illustrates the current understanding of the interrelationship of
waste solid, drain and water flow parameters aided by both laboratory and site testing as
well as computational modeling through simulations capable of representing the
Figure 6:
Reduction in w.c. flush volume over the past century through design
innovation.
307 Figure 7:
Solid transport dependence on w.c. design parameters as well as
both drain and waste solid dimensions.
unsteady partially filled pipe flow conditions and the shear interactions between the
solids in transport and the surrounding water flows. Figure 8 reinforces the importance
of w.c. design and in particular the % of flush water available behind the discharged
solid – this is a major determinant of successful drain line carry.
Figure 8: Solid transport depends primarily on the volume of flush water
discharged behind the solid. Maximising this volume is a major
design objective.
308 Figure 9: Summary of the effect of drain diameter, cross section and slope on solid
transport.
Research over 30 years has identified the relationships that dominate and provides a
basis for computational simulation, McDougall and Swaffield (2007). Drain cross
section is highly significant; as flush volumes decrease, and if throughflow decreases,
the response should be to reduce drain diameter or introduce hydraulic solutions found
in low flow sewers - non-circular cross sections, Cummings, McDougall and Swaffield
(2007). Reduced cross section coupled to steeper slopes will be a potent design option
under low flow conditions, Figure 9.
These mechanisms and dependencies are confirmed by the HWU solid transport
simulation, DRAINET, Figure 10, that predicts the effect of w.c. discharge profile,
drain slope, cross sectional shape and dimension, appliance discharge frequency,
interaction of multiple appliance discharges, including junction design and defective
drain slopes and obstructions, Swaffield and Galowin (1989), Swaffield and McDougall
(1996), Swaffield, McDougall and Campbell (1999), McDougall and Swaffield (2000
and 2003). The simulation draws on solid transport studies, HWU laboratory tests and
site evaluations at Nottingham Teaching Hospital and London Underground in the UK
as well as extended periods of collaboration with NBS Washington in the 1980s. Model
development continued at HWU to include multi solid interactions and surfactant dosed
appliance discharges, Gormley and Campbell (2006a, 2006b and 2007)
309 Figure 10: DRAINET simulation allows solid transport to be
predicted in single and multiple pipe situations, with and without
joining flows or secondary flushes.
Thus previous research provides the necessary tools to predict solid transport as water
conservation is achieved internationally and identifies design priorities to assist solid
transport. Accessing the available research is a major priority. Relevant CIBW62
papers prior to 1992 were summarized in Swaffield and Galowin (1992) while Kiya
(2000) provided a cd compilation of all CIBW62 papers from 1973 to 2000. Since 1992
the topic has continued to be considered by CIBW62 and by established international
research journals - Building and Environment, Building Research and Information and
Building Services Engineering Research and Technology among others
Historically major contributions are due to Galowin in the US through NBS (now NIST)
and the Washington Suburban Sanitary Commission and by the late Tom Konen’s
group at Stevens Institute, Hoboken. WRc in the UK has undertaken laboratory testing
on low flush w.c.s. with Butler at Exeter University, while Caroma have investigated
non-circular drains in domestic sizes and published, individually and in collaboration
with HWU. Any literature search would have to include all these avenues.
5.
What design or installation strategies might be used to offset any
perceived
problem?
Careful drainage system and appliance design can minimise the probability of
deposition while flow booster solutions should also be recognised. Minimum transport
distance to the first joining flow junction is an essential parameter and has design
implications. While accurate data is not available, experience suggests that in the
majority of cases the distance to a flow confluence will be less than 5 metres, the
frequency of w.c. to junction distances following the pattern suggested in Figure 11.
Developing the actual form of Figure 11 would be a priority research objective as the
mean distance to a flow confluence is a major determinant of drainline carry.
310 Figure 11: Distribution of w.c. to junction distances likely in
current practice.
Similarly junction design becomes a major issue as the hydraulic jumps upstream of a
junction of two or more flows presents an impediment to solid transport leading to
deposition. Swept entry junctions should be used and top entry 90o entries banned.
Flow boosting is also of interest, 1980s research concentrated on two alternate designs;
the traditional tipping tank used in the UK from at least the 1860s where an
eccentrically pivoted tank tips a large water volume into the drain at periodic intervals,
and the siphon tank, accepted by Stockholm in 1989 as a design solution to the
installation of 3 litre w.c.’s in city apartment blocks – a 21 litre siphon tank in the
basement intercepted w.c. flushes and delivered its contents to the drain in one surge.
6. Conclusions – myth or reality.
Research over the past 35 years reported primarily through CIBW62 but also in
recognised research journals has fully investigated the mechanisms of solid transport or
drainline carry. This work has included site and laboratory testing and observation and
has led to at least one computational model based on the numerical solution of the free
surface unsteady flow equations defining drainage system flow and wave attenuation.
The challenge is therefore to access that work rather than re-invent the wheel.
However that does not answer the question – Myth or Reality? Climate change and
demographic evidence suggests that, while water conservation will continue to be a
priority, water usage levels are much higher than 20 years ago and will continue to be so
even if current government targets – in the UK Part G Building Regulations – are
achieved. This suggests strongly that the basic perception of dry drains is misguided.
311 In order to deliver a high quality drainage performance it will be necessary to deliver
good design for both w.c.s and the drainage network to mitigate any possibility of solid
deposition and fully support the prime objective which must be to contribute to water
conservation as an essential constituent of our response to climate change – a response
that will need to be truly international to be successful. CIBW62 therefore has the
opportunity to contribute significantly to this debate if it chooses to do so.
Finally therefore this paper submits that Dry Drains are a Myth.
7.
References
AWWA. (1999) ‘Water efficiency making cents in the next century’, AWWA Procs,
Conserv 99, Monterey, Calif., Jan. 31st – Feb. 3rd.
Benn H. (2009) ‘UK rainfall scenarios’, Launch of UK government predictions,
Guardian, June 19th.
Cox D.(1997) ‘Designing with water’, Soc for Responsible Design Newsletter, 48: 4,16.
Cummings S., McDougall J.A. and Swaffield J.A. (2007) ‘Hydraulic assessment of noncircular section drains within a building drainage network, BR and I, 35(3), pp316-328.
DCLG. (2009) Demographic predictions drawn from the current Housing Statistical
Release
11
March
2009
DCLG.
http://www.statistics.gov.uk/statbase/Product.asp?vlnk=997
Edwards K. and Martin L. (1995) ‘A Methodology for Surveying Domestic Water
Consumption’, Jour. CIWEM 9th Oct., pp177-488
Gormley, M. and Campbell, D.P. (2006a) ‘Modelling water reduction effects: method
and implications for horizontal drainage.’ Building Res. & Information 34(2), 131–144
Gormley, M. and Campbell, D.P. (2006b) ‘The transport of discrete solids in above
ground near horizontal drainage pipes: A wave speed dependent model’ Building and
Environment 41, 534–547
Gormley, M. and Campbell, D.P. (2007) ‘The effects of surfactant dosed water on
solid transport in above ground near horizontal drainage systems.’ Building and
Environment 42, 707–716
Griggs J. and Shouler M., (1994) ‘An examination of water conservation measures’,
CIBW62 Symposium, Brighton, September 26-30.
Kiya F. (2000) ‘Cd of CIBW62 conference papers 1973-2000’, available on request.
McDougall J.A. and Swaffield J.A. (2000) `Simulation of building drainage system
operation under water conservation design criteria', BSER&T Volume 21, Number 1.
McDougall J.A. and Swaffield J.A. (2003) The influence of water conservation on drain
sizing for building drainage systems, BSER&T, 24,4 pp229-243.
312 McDougall J.A. and Swaffield J.A. (2007) ‘The transport of deformable solids within
building drainage networks’, BR and I, 35(2), pp220-232.
Parliamentary OST (2000), ‘Water efficiency in the home’, Note 135, 2000
Swaffield J.A. and Galowin L.S. (1989) "Multistorey building drainage network design an application of computer based unsteady partially filled pipe flow analysis'. Building
and Environment, Vol. 24, No.1, January, pp99-110.
Swaffield J.A.and Galowin L.S. (1992) ‘The engineered design of building drainage
systems’, Ashgate, Gower Press, pp405.
Swaffield J.A. and McDougall J.A. (1996) `Modelling solid transport in building drainage
systems' Water Science and Technology, Volume 33, No 9, July
Swaffield J.A., McDougall J.A. and Campbell D.P. (1999) `Drainage flow and solid
transport in defective building drainage networks', BSER&T Vol 20 Number 2
8.
Presentation of the Author
Professor John Swaffield is an Emeritus Professor at Heriot
Watt University. His research interests include water
conservation and the simulation of unsteady flows in
building drainage and vent systems and has been funded by
government and industry for 30 years. He has contributed
regularly to CIBW62 meetings since 1975.
313 V.3
Designing sewers for reduced wastewater
flows
Jeff Broome
jeff.broome@arup.com
Arup
Rose Wharf, 78 East Street
Leeds LS9 8EE
Abstract
There is increasing concern that reduction of wastewater flows arising from intensive
water conservation and demand management may result in an increase in blockage and
maintenance requirements for sewerage systems. A design methodology, developed
primarily to reduce the cost of providing sewerage services to poor communities, has
been demonstrated to function well with very low wastewater flows. It is proposed that
a demonstration project be developed applying these principles to the design and
construction of sewerage for a new housing development that incorporates a high level
of water conservation. The ultimate objective would be to gain acceptance for a simpler
and more rigorous sewer design methodology that can be applied to any flow regime
regardless of the wastewater flow per connection.
Keywords
Sewer design, water conservation, boundary shear stress, simplified sewerage.
1
Introduction
Conservative assumptions and robust construction have until now provided a fairly
reliable and durable solution to the sanitation requirements for a huge number of towns
and cities. Generously sized pipes and robust construction have permitted huge
increases of flow and expansions in the areas and numbers of properties served, but
there is a new consideration that must now be taken into account: reducing wastewater
flow. Scarcity of water resources and the possibility of climate change radically
314 affecting rainfall distribution are motivating demand side measures to curb water
demand so that future we may find ourselves in the position of having to anticipate
reducing flows instead of having to make allowances for future increases. In addition
the move away from combined sewers will remove the additional flow from stormwater
that may provide periodic flushing of solids from the system.
The prospect of reducing sewage flows and changes in sewage composition as a result
of water conservation gives rise to concerns over the effects that this might have on the
sewerage system and sewage treatment.
The British Environmental Agency commissioned a study, Less Water to Waste
(Drinkwater, Chambers and Waylen, 2008) which concludes, among other things, that
design methods will require revision to take account of reduced wastewater flows.
In the UK sewerage design is governed by a number of documents, principally a
standard, a code of practice and building regulations. The stipulations of these
documents are not entirely consistent and there is a fourth report, Design of sewers to
control sediment problems (Ackers, Butler and May, 1996) giving additional
recommendations that are intended to minimise the problems caused by sediment in
sewers. A review and revision of the sewer design practices is therefore justified, if only
to clarify and simplify the process.
This paper proposes a methodology for designing sewers to account for reduced flows
and a strategy for testing this approach.
2
Objectives for a revised sewer design methodology
Any sewerage design methodology has to address the three basic requirements of:



Hydraulic capacity;
Transport of solids; and
Prevention of blockage
In addition it would be worthwhile establishing the principles that any methodology
should be as simple as possible by setting values for a minimum number of parameters
and that, if possible, the performance of sewers under conditions of reduced wastewater
flows shall be no worse than sewers designed by established methods and experiencing
typical existing flow rates.
2.1
Hydraulic capacity
The hydraulic capacity of a pipe conveying liquid by open channel flow are the pipe
diameter, gradient and the maximum permitted depth of flow, (d/D)max. Generally the
depth of flow is limited to 0.75 or 0.8 of the pipe diameter to permit ventilation above
the liquid surface and to provide an additional factor of safety against surcharging.
There does not seem to be any reason to depart from this practice in devising a new
design approach.
315 2.2
Transport of solids
The transport of sediment must be considered separately from the transport of gross
solids as the mechanisms are different. Sediment consists of small particles that are
transported in suspension, as a bedload close to the invert of the sewer, or deposited as a
bed. The requirement for sediment transport has generally been accommodated by using
a concept of a self cleansing velocity.
An alternative criterion, boundary shear stress, has been proposed as a means of
ensuring that sewers are capable of transporting sediment. The average boundary shear
stress is the average shear stress exerted on the sewer wall by the moving liquid and is
given by the expression
τ = ρgRi
(1)
Where τ is the average boundary shear stress, ρ is the density of the wastewater, g is the
acceleration due to gravity, R is the hydraulic radius and i the slope.
Although not the first person to advocate the use of boundary shear stress in sewer
design, Yao (1974) appears to have produced the first design procedure using this
concept. Yao noted that experimental results showed that the velocity required to
transport low concentrations of sand through pipelines varied as the square root of the
pipe diameter and so a single value of self cleansing velocity could not be applied in all
cases, whereas a single value for boundary shear stress should be valid for all pipe
diameters, other factors such as sediment composition and load remaining unchanged.
Boundary shear stress is also the basis of the recommended procedures presented in
Ackers, et al. (1996). This report, published by the Construction Industry Research and
Information Association had the objective ' . . . to develop a standard methodology for
the hydraulic design of sewers to control sediment problems, and to produce from it
appropriate guidance on the subject for design engineers.' In addition to reducing
sediment problems in sewers, the approach was expected to produce more economic
designs for small diameter sewers which tend to be over-designed using a single value
for self-cleansing velocity.
It would therefore seem that any future design methodology for sewer design should
accept the basic proposal put forward by Yao and include a minimum value for
boundary shear stress (τmin) and have no need of any consideration of velocity.
2.3
Prevention of blockage
There has been a significant amount of research into the transport of larger solids,
particularly related to the design of drainage within buildings, see for example Wise and
Swaffield (1995). This has concentrated on examining the distance that test objects are
transported by a standard pulse of water, designed to simulate the discharge from a WC,
in pipes of differing diameter and gradient. The experimental results clearly show that
solids are transported further in smaller pipes than larger ones and in pipes with steeper
gradients than ones with less slope. This is explained by the concept of damming of the
316 water flow behind the object and the resulting forces overcoming friction and moving
the object forward at each successive passing wave.
One way to ensure that the pipe diameter is not too large is to ensure that a minimum
depth of flow (d/D)min is achieved at the daily peak flow and that the smallest available
pipe size is used that will accommodate the anticipated maximum peak flow. In practice
this may well be smaller than the minimum size permitted under sewer design
regulations, which in some places is 200 mm or even larger.
This assumption is also supported by findings of an investigation into the causes of
sewer blockage in the UK (Lillywhite and Webster, 1979), which concluded that the
main factors contributing to blockage were the degree of utilisation, which is equivalent
to the depth of flow, and defects in construction.
2.4
Sewer gradient
The same study found that sewer gradient appeared to have little effect on the rates of
blockage, which may sound surprising, but one length of 100 mm diameter sewer that
was investigated was found to have a gradient of 1:1200 and to be operating without
problems.
A minimum value for gradient can be determined using the set value of boundary shear
stress and minimum depth of flow, however, it is difficult to determine the flow rate to
use for the upper reaches of sewers where there are few connections. The average flow
will be very low and even a high value for peak flow factor will not yield a useful result.
A valid approach would be to use a value for the minimum peak flow that corresponds
to the peak flow from a single connection. This would be equivalent to the discharge
from a single WC. This concept is included in the National Appendix to EN 752 (BSI,
2008) which states a minimum value for flow should be taken as 1.6 l/s.
This is in contrast to current UK practice where minimum gradients are stipulated for
different pipe diameters and flow rates. Also for a 100 mm drain the number of
connections permitted is limited to 10, which is contrary to the findings of Lillywhite
and Webster (1979) that the flow should be maximised.
In conclusion, the parameters that should be considered to minimise sewer blockage
will include a minimum value for the relative depth of flow (d/D)min, a minimum pipe
diameter, Dmin and a value for the minimum flow rate, qmin. It is also is important that
the minimum pipe diameter that will meet the hydraulic requirements is selected to
ensure that the level of utilisation is maximised.
The only other requirement to explicitly determine the required pipe diameter and
gradient for a given flow regime is the peak flow factor Fp. This should be determined
as accurately as possible, from measurement of flow rates in similar sewer lengths if
possible. Because of the need to maximise the level of utilisation, there should be no
hidden partial safety factor incorporated in this value to account of unforeseen
development or infiltration. These must be explicitly estimated and incorporated into the
value for estimated future flow.
317 2.5
Potential design methodology
The proposed methodology will therefore depend on setting values for 6 parameters
discussed above and developing the necessary design equations based on a standard
pipe flow formula. This may taken as the Manning-Strickler equation using the
commonly adopted value for n of 0.013 for slimed sewer pipes, regardless of the
material. The extra complication introduced by using more accurate flow formulae, such
as Colebrook White, is not justified, but this would be an equally valid approach.
As noted above, the use of boundary shear stress for sewer design has been advocated
for over 35 years, but has not gained wide acceptance. Before making such a major
change in design approach, it would be useful to have some demonstration of the
effectiveness of the methodology and experience of its widespread application.
Fortunately there is such evidence available through the development of a design
philosophy in Brazil that substantially reduces the costs of providing sewerage so that
low-income households can afford water borne sanitation.
3
Simplified Sewerage
The simplified sewerage concept was originally developed in the northeastern Brazilian
state of Rio Grande do Norte as a means of providing an affordable water borne
sewerage service to low income housing areas, both planned and un-planned. The key to
this approach is that the hydraulic design is rigorous and the system is hydraulically as
efficient as possible.
Design was based on a self cleaning velocity of 0.5 m/s and a minimum peak flow of
2.2 l/s. This resulted in a minimum gradient of 1:167 for a 100 mm diameter sewer (de
Andrade Neto, 1985; UNCHS, 1986). Because one of the major cost saving strategies
was to locate sewers though private land within the housing block, it was known as
condominial sewerage.
Other state water companies in Brazil quickly adopted this approach and to suit local
conditions constructed sewers beneath sidewalks or front gardens as well as in-block. At
the same time the design basis was changed to one based on boundary shear stress,
known as tensão trativa (tractive tension) (Machado, Neto and Tsutiya, 1985). The
simplified sewerage concept has been incorporated in the Brazilian Sewer Code (ABNT,
1986) and has been applied in other countries, mainly in south Asia and Latin America.
3.1
Design equations
The values for the parameters identified above and as applied in Brazil are:
Minimum boundary shear stress
Minimum proportional depth of flow
Maximum proportional depth of flow
Minimum peak flow rate
Minimum pipe diameter
τmin
(d/D)min
(d/D)max
qmin
Dmin
318 1 N/m2
0.2
0.8
1.5 l/s
100 mm
Peak flow factor
Fp
typically 1.8
As
measured,
The minimum diameter suggested here is for collector sewers, however, 75 mm may be
more appropriate for the property connection, but this has seldom been applied, largely
due to the lack of drainage fittings for 75 mm pipe.
Based on the Manning-Strickler equation, design equations for minimum gradient and
pipe diameter can be derived, as shown in Bakalian, Wright, Otis and Azevedo-Neto
(1994), Mara (1996) and Mara and Broome (2008). The minimum gradient can be
determined from of flow, τmin and (d/D)min:
imin = [(1/n)kakr−2]6/13(τmin/ρg)16/13 q–6/13
(2)
Where
ka =
θ
n
θ – sinθ 
8
and
kr =
1  sin θ 
1 

4
θ 
angle subtended at the centre of the sewer by the wastewater surface (radians)
Manning's roughness coefficient (usually taken as 0.013)
Substituting the values commonly used in Brazil this simplifies to:
imin = 2.33 × 10−4q–6/13
(3)
Similarly the pipe diameter required for a given maximum or minimum flow rate,
gradient and proportional depth of flow can be determined:
D = (nq)3/8ka−3/8kr−1/4imin–3/16
(4)
The minimum diameter is determined from the maximum peak flow and maximum
proportional depth of flow, taking the next larger size of commercially available pipe.
The maximum diameter is determined from the minimum peak flow, which will usually
be the initial condition with only partial development of the catchment, and the
minimum proportional depth of flow.
It can be seen that the design equations lead directly to explicit solutions for minimum
gradient and pipe diameter and so are relatively simple to apply. However, a key
advantage in developing the equations governing hydraulic design from set values for a
minimum number of parameters is that the design is consistent. Once the appropriate
values are selected for those parameters, all aspects of the hydraulic design are defined
and there is no need for any of the empirical or arbitrary limits on gradient or numbers
of connections that are common in self cleansing velocity design procedures. If it is
found, for instance, there is deposition of sediment, it would be a simple matter to adopt
a higher value for boundary shear stress, without any need to adjust any of the other
parameters.
319 For detaileed description of the application
a
of simplifieed seweragee principless see Mara
and Broom
me (2008), which inccludes instrructions forr a simple design tem
mplate for
selecting pipe
p diameteer and gradiient.
3.2
Applications off simplified
d sewerage
Simplified sewerage was develooped to redduce the co
ost of wateer borne seewerage to
ge of half that of convventionally
affordable levels and the cost iss typically in the rang
a constructed seweraage. While most waterr companies in Brazil have only
designed and
applied sim
mplified sew
werage prinnciples to low-income
l
e areas, the state waterr company
serving thee Federal District
D
of Brasilia
B
is using
u
the ap
pproach in high incom
me areas as
well (see Fig
F 2).
In many of the projeccts where thhe simplifieed seweragee approach has
h been addopted, the
l
haave been veery low. Oft
ften the WC
Cs connecteed will be oof the pour
hydraulic loadings
flush type, in that theyy are not prrovided withh a flushing
g cistern andd the flushinng water is
a poured into the paan. In many
y cases the water usedd is sullage
carried to the toilet and
wer than woould be expected from
from otherr uses and so the wasteewater flow is even low
the 1.5 to 2 l used per flush (see Fig
F 2).
Figuree 1: Simpliffied seweraage being in
nstalled at the
t back off the pavem
ment in
Brasilia
Figure 2: Pour flu
ush squat ty
ype WC
32
20 In Orangi, a district of Karachi, Pakistan, simplified sewerage was introduced
successfully in an area where the water consumption was measured at the start of the
project as only 27 lcd obtained from intermittent piped supplies and vendors.
Admittedly the gradients are probably higher than in most simplified sewerage project
areas, but the success shows that sewerage will function with very low wastewater
flows.
4
Design for low flows in industrialised countries
Simplified sewerage is a mature technology that has been applied in various parts of the
world with a great deal of success. Admittedly there have been failures, but under
similar conditions, conventional sewerage systems also often fail.
It would therefore seem reasonable to assume that a design methodology that can
produce successful sewerage systems with very low flows in developing countries could
also be applied in other contexts, such as in industrialised countries where there are
concerns about reducing wastewater flows. However, there are serious obstacles to
overcome. In developing countries there is often strenuous opposition to adopting
alternative design criteria that may be seen as inferior.
The idea that smaller pipes are less likely to block than larger ones seems counterintuitive to many people and so the need to reduce the minimum sewer diameter is
probably the greatest impediment to the widespread adoption of the simplified sewerage
design approach. An example of this occurred in Faisalabad, Pakistan where the
minimum sewer diameter was maintained as 225 mm (9 inches) (Khatib Alam and
Parkinson, 2002), when a key feature of simplified sewerage is that pipe diameters
should be selected to maximise utilisation and so a 100 mm, or at most 150 mm
diameter should have been selected.
Given that in the UK rules of thumb from more than a century ago, Maguire's Rule for
example, are enshrined in standards and codes of practice (BSI, 2008; WRc, 2006) and
the European standard is based on self-cleansing velocity approach, it may prove
difficult to gain acceptance of alternative design methodologies and philosophies.
4.1
Additional supporting evidence
In addition to the success of many simplified sewerage schemes in developing
countries, there is evidence from the USA and the UK that there is some convergence in
approach and some elements of simplified sewerage design are already in use.
In the US state of Nebraska there has been a long and successful history of using ‘flat
grade sewers’. Due to the typically flat topography and high water tables in the state,
150-mm and 200-mm diameter sewers have been laid at gradients as flat as 1 in 900,
without any ‘unusual’ maintenance requirements (Gidley, 1987).
Mara and Broome (2008) show that the simplified design method of Ackers et al. (1996)
gives very similar results for the minimum gradient for small diameter sewers and
321 drains when compared with the simplified sewerage design procedure, as applied in
Brazil.
The UK National Annex to EN752 (BSI, 2008) contains a provision that is equivalent to
the minimum flow criterion of simplified sewerage, that the minimum flow for the
design of drains and sewers serving small numbers of dwellings should be taken as 1.6
l/s (i.e., close to the value of 1.5 l/s used in Brazil). This is not included in the code of
practice (WRc, 2006) which instead stipulates average flow rates and implicitly
prescribes a very high peak flow factor of 6.
Pressure from housing developers to reduce the cost of construction of sewerage has led
to the substitution of sealed access fittings for manholes and also the relatively recent
relaxation of some rules that govern the number of connections permitted on 100 mm
diameter drains (DEFRA, 2002). These developments represent cautious moves towards
some aspects of the simplified sewerage design philosophy.
4.2
Research versus demonstration
The basis of the UK Environment Agency's report on the impact of low flows on
sewerage systems (Drinkwater, 2008) was to seek evidence from existing systems
where flows are reducing and so failed to identify the development of simplified
sewerage as a methodology that has been shown to operate in areas with very low water
consumption rates.
As a result of this, the report recommends a variety of investigations as a necessary
input to revising design methods, but does not give much indication of what the ' . .
practical investigations into the effects of reduced WC flush volumes on the parts of the
drainage system most susceptible to blockages due to low flows . . ' might actually
achieve (Drinkwater et al., 2008). Nor does the suggested rig based testing hold much
hope of simulating actual conditions in the upper reaches of drainage systems where
flows are intermittent and solids are regularly stranded. It therefore follows that the only
way to really test the proposed approach is to construct an actual sewerage system
serving an area where all connections serve properties with very high levels of water
conservation. This is, after all, the way that current design standards were developed.
Since there is likely to be resistance to the idea of a new design methodology, a
convincing demonstration will be required to convince sceptical engineers, other
professionals and members of the public. There will undoubtedly be an unwillingness to
abandon all the accumulated experience of traditional practice and there is also a huge
difference in the cultural and hygiene behaviour between industrialised countries and
those where simplified sewerage has been operating successfully.
If such a drainage system is to be constructed, it will be necessary to identify a type of
development where the water conservation objectives are combined with a sufficiently
extensive drainage system that would form a suitable pilot area for the project. One
important aspect of prestige 'low carbon' and 'zero carbon' housing developments is
water demand management. This would make such developments an ideal testing
ground for alternative approaches to sewer design.
322 4.3
Obstacles to developing a pilot scheme
There are clearly going to be a lot of obstacles to using a new design basis, not the least
of which is that EN 752 (BSI, 2008) explicitly states that self cleansing velocity
methods should be used (see section 9.6.3). However Section 8.7.3 does state that
'Drains and sewers shall be designed to provide sufficient shear stress to limit the build
up of solids . . .' Local planning laws, building regulations and water company policy
are also almost certainly going to be an impediment to a radical departure from
established practice. The successful development of a pilot scheme is therefore going to
require a high level of commitment from the developer and its financiers, political
support and a flexible approach by all the regulatory bodies involved.
The duration required for such a demonstration project to yield results and the possible
longer term consequences, of increased maintenance requirements for example, means
that it will need significant resources in addition to those required for actual
construction.
4.4
Possible experimental design
Any demonstration project would need to demonstrate that:



The system works as well, or better than existing standard practice;
The design methodology is simple to apply and the resulting sewerage system is
simpler, or is at least no more difficult to build than a conventional one; and
The resulting sewerage system will be more economic.
A possible sequence of activities could include a review of the performance of
simplified sewerage systems, development of standards for the UK that have the
potential to meet the expectations of water companies and regulatory bodies, but
without compromising the principles of simplified sewerage, comparison of costs for
conventional and simplified sewerage systems and finally the construction of a
sewerage network based on the proposed methodology.
Assessment of the success of the design would be based on monitoring of water
consumption and sewage flows, requirements for maintenance, the CCTV inspection of
sewers. It is extremely improbable that suitable data will be available to compare the
experimental systems with conventionally designed sewers. Not only is comprehensive
data on sewer blockage, routine cleaning and other maintenance unlikely to be available
for comparable sewerage networks, any available data would be difficult to correlate
with water consumption and wastewater flows. It would therefore be necessary to
include within the demonstration scheme a control area designed strictly in accordance
with established practice, codes and standards.
The outputs from the demonstration project should include a number of individual
pieces of academic research and a summary of the development of the project and
research findings. In addition there should be a draft standard for sewer design that
would hopefully be suitable for designing sewerage both for areas with conventional
plumbing fixtures and average wastewater generation rates and for areas where high
levels of water conservation and greywater recycling are implemented.
323 4.5
Risks
There are a number of potential risks that will need to be addressed before a developer
and other participants in a demonstration project are likely to be willing to proceed.
These include:





Failure to obtain the necessary waivers or derogations of regulations and
standards;
Opposition from water companies (adoption is not likely to be possible in any
case);
Possible additional maintenance requirements;
Difficulty of selling an experimental system to potential residents; and
Resourcing a rather long term project.
However, with sufficient support from agencies responsible for water resources
management, housing and infrastructure provision, it should be possible to reduce these
risks to acceptable levels and implement a successful project.
5
Conclusions
A methodology developed in Brazil to provide sewerage services to low income
communities has been proved to work well with low wastewater flows and which has
been successfully implemented in a number of countries. There seems little reason to
doubt that this approach, known as simplified sewerage, could also be applied in
industrialised countries to accommodate reduced wastewater flows arising from water
demand management and greywater recycling.
Rather than continue to promote research through laboratory testing and small scale
studies of existing conventional sewerage systems, a demonstration project would be far
more effective in influencing sewer design methods. Although promoting a
demonstration project will be far more difficult than continuing individual small scale
research projects, largely because of the scale necessary, it is probably the only way to
effectively validate the design principles proposed.
Adoption of simplified sewerage design methods would greatly simplify sewer design
by reducing the number of individual regulations and standards to a minimal set of
parameters that define all aspects of hydraulic design.
324 6
References
ABNT (1986). Projeto de Redes Coletoras de Esgoto Sanitário. Rio de Janeiro:
Associação Brasileira de Normas Técnicas.
Ackers J. C., Butler D., and May R.W.P. (1996). Design of Sewers to Control Sediment
Problems. London: Construction Industry Research and Information Association,
Report No. 141.
De Andrade Neto C.O. (1985). Uma solução eficaz e de baixo custo para o esgotamento
sanitário urbano. Engenharia Sanitária, 24(2), 239−241.
Bakalian A., Wright A., Otis R. and Azevedo-Netto J. (1994). Simplified Sewerage:
Design Guidelines. Washington DC: World Bank
BSI (2008). BS EN 752: 2008, Drains and Sewer Systems Outside Buildings. London:
British Standards Institution.
DEFRA (2002). Protocol on design, construction and adoption of sewers in England
and Wales. London: Department for Environment, Food and Rural Affairs.
Drinkwater, A., Chambers, B., Waylen, C. (2008). Less water to waste: Impact of
reductions in water demand on wastewater collection and treatment systems, Bristol:
Environment Agency.
Khatib Alam, S.M., Parkinson, J. (2002). Appropriate design standards and
construction specifications for tertiary systems. Faisalabad, Pakistan: Faisalabad Area
Upgrading Project.
Gidley J.S. (1987). Ericson, Nebraska Flat Grade Sewers, Case Study No. 11.
Morgantown: National Small Flows Clearinghouse, West Virginia University.
Lillywhite M.S.T. and Webster C.J.D. (1979). Investigations of drain blockages and
their implications on design. Journal of the Institution of Public Health Engineers, 7(4),
170−175.
Machado Neto J.C.O. and Tsutiya M.T. (1985). Tensão trativa: um critério econômico
de esgoto. Revista DAE, 45(140), 73−87.
Mara, D. (1996). Simplified sewerage: simplified design. In Mara , D. (ed.) Low-cost
Sewerage, (pp169−174): Chichester,Wiley,.
Mara, D., Broome, J. (2008). Sewerage: a return to basics to benefit the poor. Municipal
Engineer. 161(ME4), 231 – 237.
UNCHS (1986). The Design of Shallow Sewer Systems. Nairobi: United Nations Centre
for Human Settlements,.
Wise, A F E and Swaffield, J E, (1995). Water and sanitary waste services for buildings,
fourth edition. Harlow,.Longman Scientific and Technical.
325 WRc (2006). Sewers for
f Adoptioon: A Desiggn and Con
nstruction Guide
G
for D
Developers,
6th editionn, Swindon: Water Reseearch Centree.
Yao, K.M. (1974). Sew
wer line dessign based on
o critical shear
s
stress,, Journal off the
Environmeental Engineeering Divission, Proceedings of th
he Americann Institutionn of Civil
Engineers, 100(EE2), April 19744.
7
Preseentation of
o Author
Jeff Broom
me is a senioor engineer with Arup in
i their Leeeds office.
While he is now emplloyed on suupervising construction
c
n of water
p
he has
h extensivve experiencce of both urban
u
and
treatment plants,
rural wateer supply and sanitaation in deeveloping countries,
c
mainly in Africa.
A
32
26 Session VI – Drainage I
VI.1
Study of Performance Evaluation of Drainage
Systems with Air Admittance Valves for
Single-family Housings
Effects of Swivel Fittings with Air Admittance Valve Function to Improve
Drainage Capacity
(1) Kazutoshi Suzuki (2) Masayuki Otsuka (3) Norihiro Hongo
(4) Koichi Kawasaki
(1) ka-suzuki@kitz.co.jp
KITZ Corporation, Japan
(2) dmotsuka@kanto-gakuin.ac.jp
Department of Architecture College of Engineering, Kanto Gakuin University,
Dr. Eng, Japan
(3) hongou@astro.yamagata-cit.ac.jp
Lect. Dep. of Architectural Environment System Engineering,
Yamagata College of Industry and Technology, Japan
(4) k-kawasaki@kitz.co.jp
KITZ Corporation, Japan
Abstract
In Japan, standardization of methods to design and construct drainage systems for
single-family housings is not sufficient, and design and construction of such drainage
systems are often left to workmanship. The current situation also indicates that design
and construction manuals are insufficiently provided to housing constructors and users’
complaints are many about the trouble of broken traps within their drainage systems.
327 Based on a drainage system normally used for three-story single-family housings, this
study proposes a drainage vent system developed with air admittance valve function for
improvement of drainage capacity and easier installation of drainage systems. It is
provided with function of moderating airflow resistance when drain water flows down
stack drain pipes.
The study also evaluates drainage capacity of the proposed drainage vent system, under
an actual drain load, in comparison with drainage capacity of conventional vent-type
and non-vent-type drainage systems. Being more specific, two types of loads, which are
constant flow and variable drainage from fixtures, are experimentally applied to the
drainage stack system provided with different stack pipe sizes and different stack vent
opening configurations to compare and discuss resultant differences in the drainage
capacity.
Keywords
Drainage systems, Drainage performance, Air systems, Swivel fittings with air
admittance valve function, Single-family housings
1. Background
In Japan, engineering standards to design and construct drainage systems for low-rise,
single-family housings have been less defined than those for apartment houses, and it
disturbs establishing detailed methods of drain and ventilation works for their stack
drain systems. The problem of limited installation space often causes undesirable
plumbing works such as fully blocking vent openings which results in no airflow, or
reducing pipe opening sizes and fitting up vent valves within walls. Where the 2nd and
3rd floors of single-family housings have been thus plumbed, disturbed airflow might
cause extremely negative drain pressure to induce siphon effect and result in seal break
of traps on lower floors.
2. Objective
We have developed a swivel type fitting with an air admittance valve (“a swivel air
admittance fitting” in short below) for improvement of capacity of a house drainage
system. Its design is overviewed in Fig.1. This fitting is designed with a provision of
air-intake function of a vent valve and swirling flow of drainage, while a care to solve
the problem of limited installation space for single-family housings is taken by means of
a built-in air admittance valve. With a horizontal drain branch eccentrically connected,
328 this air admittance fitting swivels water flow to secure an airflow path along a stack
drain. Our objective was to evaluate the performance of this fitting installed on the top
floor to improve drain capacity of a three-story single-family housing. Our evaluation
was made by comparing the performance of this system with those of conventional vent
type and non-vent type plumbing methods.
151
110
Air admittance valve
97
Airflow guide
83
275
180
83
40
40
97
(2) Side
(1) Front
Fig.1. Overview of a swivel air admittance fitting
3. Summary of Experiments
3.1 Test arrangement
Two series of experiments were performed. One was the constant flow drainage load
test (“the constant flow test” in short below) conducted with the standardized load
(Fig.2 Part I). Another was sanitary fixture drainage load test (“the sanitary fixture
test” in short below) carried out with variable loads drained from fixtures
(Fig.2 Part II).
329 (I) Constant flow test piping
(II) Sanitary fixture test piping
[Unit:mm]
Fig.2 Test drainage stack piping
Fig.3 indicates locations of horizontal drain branches and sanitary fixtures used for the
sanitary fixture test. The stack vent system for a three-story building was deployed with
JIS-DT fittings. Drain pressure was loaded from the top floor, which was the toughest
test condition. Sizes of test drain stack pipes were 50A, 65A and 75A. Sizes of
horizontal drain main pipes were 65A, 75A and 100A, one size larger respectively than
sizes of stack piping. Gradients of horizontal drain branches at the 3rd floor were 1/50
for 50A and 65A, and 1/100 for 75A, which are the minimum requirements specified in
SHASE-S206 (*2). Horizontal drain main pipes were also installed with the same
gradients.
330 1600
300
950
250
65A
480
65A 50A
65A
LWashbasin
(L)
B
50A
50A
350
400
50A
Bathtub (B)
Trap
Gradient:1/50
Unit:mm
175
(With 50A and 65A piping)
1600
300
400
300
Toilet
75A
75A
1000
Washbasin
(L)
50A
350
400
50A
65A
Bathtub (B)
Trap
Gradient:1/100
Unit:mm
175
(With 75A piping)
Fig.3 Layouts of horizontal drain branches and sanitary fixtures on 3rd floor
3.2 Test vent opening configurations and items of measurement
Table 1 shows 4 test models (a), (b), (c) and (d) to cover the stack vent opening on the
top of the stack piping, with arrows indicating direction of airflow. The vent opening
was either blocked with a cleaning cap (a), or covered with a bell mouth (b) or a vent
cap (c). Or, a swivel air admittance fitting (d) was installed to the fitting section of the
stack on the top floor.
Table 1. Test vent opening covers
Name
Cleaning Cap
Bell Mouth
Vent Cap
Swivel Air
Admittance Fitting
Symbol
a
b
c
d
Elevation
view
* The arrows indicate directions of airflow.
331 For both the constant flow test and the sanitary fixture test, pressure in drain on each
floor was measured with a small pressure converter, while piping air velocity was
measured with a hot-wire anemometer. For the sanitary fixture test, test traps were
installed on horizontal drain branch ends of the 2nd floor. 2 types of traps were used as
shown in Table 2. One was P trap used for a washbasin with 60mm deep seal water and
1.00 cross-section area ratio (“ P trap” in short below). Another was a trap for the
overflow tray of a washing machine with 50mm deep seal water and 0.88 cross-section
area ratio (“Floor trap” in short below).
Table 2. Tested traps
(I) P trap used for
a washbasin
Type
(II) the overflow tray of a
washing machine
62.0
50
32.0
60
Cross-section view
145.0
32.0
64.0
86.5
80.0
Seal water depth
60[mm]
50[mm]
Cross-section area ratio
1.00
0.88
Fluctuation of the level of trap seal water was measured with a water level sensor, while
seal loss ( h) was measured with a visual ruler. It was checked and confirmed that seal
loss of each trap is within 50[%] of its original seal water depth and instantaneous or
complete seal break has not occurred After each measurement, test traps were refilled
with seal water to the original level.
332 3.3 Test conditions
(1) The constant flow test
According to SHASE-S218 (*3), pressure in drain for the constant flow test was limited
±400[Pa]. Constant drainage load was increased from 0.5[L/s] to 2.5[L/s] with 0.5[L/s]
increments. Drainage load flow rates [L/s] are given by vent opening configurations and
stack drain sizes in Table 3.
Vent
opening
Drain stack
piping size [A]
a
b,c
d
Drain load [L/s]
50
0.5, 1.0, 1.5
65
0.5, 1.0, 1.5, 2.0, 2.5
75
0.5, 1.0, 1.5, 2.0, 2.5
50
0.5, 1.0
65
0.5, 1.0
75
0.5, 1.0, 1.5, 2.0
50
0.5, 1.0, 1.5, 2.0, 2.5
65
0.5, 1.0, 1.5, 2.0, 2.5
75
0.5, 1.0, 1.5, 2.0, 2.5
* Drainage load was limited so that negative pressure would be
maintained -400[Pa] or higher.
(2) The sanitary fixture test
A washbasin (L), a bathtub (B) and a toilet bowl of ultra water-saving type (WC)
installed on the 3rd floor were subjected to fixture drain load. Measurement of drainage
characteristics and control of drainage time lag were discussed beforehand as introduced
below:
1) It was decided to measure drainage characteristics of each sanitary fixture, according
to the records of our previous study (*4).
2) The timing for drain loading was decided as follows: After drain load has been
released from the washbasin (L) installed on the upstream on the 3rd floor to record
the maximum Psmin, bathtubs (B) and toilet bowls (WC) installed on the downstream
are subjected to the load with a time lag of 1 second each so that fluctuating pressure
in the 2nd floor piping may record the maximum Psmin.
3) It was also decided to subject the sanitary fixtures installed as illustrated in Fig.3 to
single drainage load from one fixture and to combined drainage load from multiple
fixtures with an appropriately set time lag. Patterns of sanitary fixture drainage loads
are given in Table 4.
Table 4. Patterns of test fixture drainage loads
Drain stack
piping size [A]
Single drain
load
Combined drain loads
50
L, B
L+B
65
L, B
L+B
75
L, B, WC
L+B, L+WC, B+WC, L+B+WC
333 4. Reviewed Results of Experiments
4.1 The constant flow test
Fig.4 compares distribution patterns of pressure in drain, which depends on stack vent
opening configurations. Our finding on 75A stack drain is reported here as an example
to compare the minimum pressure recorded on the 2nd floor. Blocked opening (a)
resulted in the lowest drain pressure of -873[Pa], while use of a bell mouth (b) and a
vent cap (c) resulted in the lowest drain pressure of around -455[Pa]. Both results
exceeded the evaluation criterion of -400[Pa]. On the other hand, use of a swivel air
admittance fitting (d) resulted in the lowest drain pressure of -143[Pa], the highest
negative pressure among all easily meeting the criteria of ±400[Pa] specified in
SHASE-S218 (*3).
a
4F
b
c
d
Pmin
Pmax
Drainage load 2.0[L/s]
Floor
3F
2F
Floor beneath drainage
1F
Criterion value±400[Pa]
0F
-1200
-1000
-800
-600
-400
-200
0
200
400
Pipe pressure [Pa]
* with 2.0 [L/s] drain flow rate through 75A stack piping
Fig.4. Drain pressure distributions by stack vent opening configurations
(Constant flow test)
Fig.5 compares drainage performances of each vent opening cover combined with
different sizes of stack drain piping. Considering the criteria of the maximum allowable
flow rates specified for stack vent system and stack drain piping sizes in SHASE-S206
(*2), use of a swivel air admittance fitting for 50A piping resulted in about 2.5 times as
high as the criterion. Similarly, experiments with 65A and 75A piping recorded about
2.1 times and 1.4 times as high as SHASE criterion respectively. An extremely good
drainage performance of our swivel air admittance fittings was thus verified.
334 Drainage capacity [L/s]
3.0
a
b
c
d
2.5
2.0
Allowable flow rate 1.80[L/s]
1.5
Allowable flow rate 1.20[L/s]
1.0
Allowable flow rate 0.61[L/s]
0.5
0.0
50A
65A
75A
Fig.5. Compared drainage performances by drain stack piping sizes
and stack vent opening covers (Constant load test)
4.2 The sanitary fixture test
(1) Testing fixture drainage characteristics of sanitary fixtures
Fig.6 shows how drainage characteristics of tested sanitary fixtures fluctuated. Their
drainage characteristic values are given in Table 5. Calculation of the averaged
drainage volume of tested fixtures (qd), the piping design index, was made by the
formula given below.
qd 
0 . 6  W …Formula (1)
Td
W: Fixture drainage flow rate [L/s]
Td: Time required for complete drainage [L/s]
1.5
td=7.3[s]
80[%]
1.0
2.5
8
2.0
6
4
0.5
20[%]
0.0
2
1.5
80[%]
1.0
20[%]
10
20
30
Test duration [sec]
With Toilet bowl (6L)
Fig.6-1. Drainage characteristics curves of test sanitary
335 2
0
0
With washbasin (6L)
6
4
td=2.1[s]
0.0
10
20
30
Test duration [sec]
10
8
0.5
0
0
Flow rate
Dainage volume
Dainage volume [L]
Flow rate [L/s]
2.0
10
Flow rate [L/s]
Flow rate
Dainage volume
Dainage volume [L]
2.5
2.5
150
Flow rate
120
F lo w rate [L /s]
Dainage volume
1.5
90
1.0
60
20[%]
0.5
30
td=168.6[s]
0.0
D ain ag e v o lu m e [L
2.0
80[%]
0
0
50
100
150
200
250
Test duration [sec]
300
350
With bathtub (150L)
Fig.6-2. Drainage characteristics curves of test sanitary fixtures
Table 5. Drainage characteristics values of test sanitary fixtures
Sanitary fixture
Flow
W[L]
T[s]
Td[s]
Toilet bowl of ultra
water-saving type (WC)
BC-320SU
6.0
40.0
2.1
1.7
1.9
145.1
374.9
168.6
0.5
0.7
5.9
66.4
7.3
0.5
0.6
Bathtub (B)
Washbasin (L)
Body
JVV1616*W4RMK
Overflow tray
SP1390SA
FRN-603R(Y)/PIFW
qb[L/s] Qmax[L/s]
W: Drainage flow rate T: Time required for draining [s]
[L] [L]
Td: Time required for complete draining [s] (amount of time required until drainage is completed)
qd: Averaged fixture drainage volume [L/s] Qmax: Instantaneously highest drainage flow rate [L/s]
(2) Sanitary fixture drainage loads test
1) Setting time lags for fixture drainage
The time lags for experiments were set to record the highest negative pressure.
Fig.7 shows the correlation between the negative pressure and the time lag set for
combined drainage with 50A, 65A and 75A stack drain piping. With 50A and 65A
piping, the highest pressure of combined drainage (L+B) was recorded at 1 second
after the washbasin (L) was drained off. With 75A piping, the result of the
combined drainage (L+B) was recorded the same, but the highest pressure of other
combined drainage (L+WC) or (B+WC) was recorded at 5 seconds after the
washbasin (L) and the bathtub (B) were drained off. Experiments were conducted
with these time lags.
336 -800
-800
Minimum system value Psmin[Pa]
-700
L+B(65A)
-600
-500
L+WC
L+B 5[s] setup
B+WC
-700
Minimum system value Psmin[Pa]
L+B(50A)
1[s] setup
-400
-300
-200
-100
-600
-500
1[s] setup
-400
-300
-200
-100
0
0
00s
1s
1
2s
2
3s
3
0s
0
(I) For 50A and 65A stack drain piping
1s
1
2s
2
3s
3
4s4
5s
5
6s6
(II) For 75A stack drain piping
Fig.7. Time lags set for combined fixture drainage
2) Testing drainage loads of one fixture vs. combined fixtures
Fig.8 shows fluctuations of negative pressure and floor trap seal loss ( h) caused
by differences in stack drain piping sizes and stack vent opening configurations in
correlation with single or combined fixture loads .
.
a
-700
b
c
d
-600
-500
a
-800
Pmin
Minimum system value Psmin [Pa]
Minimum system value Psmin [Pa]
-800
Criterion value
-400
-300
-200
-100
-700
B
Single drainage
-500
Criterion value
-400
-300
-200
-100
L
L+B
L+B
Combined drainage
(Psmin with 65A piping)
60
60
a
b
c
a
d
Complete breakage
Trap seal loss ⊿hmax[mm]
Trap seal loss ⊿hmax[mm]
B
Single drainage
Combined drainage
(Psmin with 50A piping)
40
Criterion value
20
10
50
b
c
d
Complete breakage
40
30
Criterion value
20
10
0
0
L
B
Single drainage
L
L+B
B
Single drainage
Combined drainage
(Seal loss with 50A piping)
L+B
Combined drainage
(Seal loss with 65A piping)
337 d
0
L
30
c
-600
0
50
b
Pmin
Minimum system value Psmin [Pa]
-800
a
b
c
d
Pmin
-700
-600
-500
Criterion value
-400
-300
-200
-100
0
L
B
WC
L+B
Single drainage
L+WC
B+WC L+B+WC
Combined drainage
(Psmin with 75A piping)
60
a
Trap seal loss ⊿ hmax [mm]
50
b
c
d
Complete breakage
40
30
Criterion value
20
10
(Seal loss with 75A piping)
0
L
B
WC
L+B
Single drainage
L+W C
B +W C
L+B +W C
C om bined drainage
(Seal loss with 75A piping)
Fig.8 Fluctuation of Psmin and trap seal loss by stack drain piping sizes
Fig.8 shows the following facts:
Single fixture drainage from a bathtub (B) with blind covered 50A piping exceeded
the evaluation criterion of -400[Pa], while all combined fixture drain (L+B) from 50A
piping covered with a cleaning cap, a bell mouth or a vent cap also exceeded the
criterion.
In the case of combined fixture drain (L+B) with 65A piping, however, negative
pressure values with all vent opening configurations except a cleaning cap were 200[Pa] or lower and could meet the criterion, thanks to enlarged stack drain pipes.
Where 75A piping was used, provision of a swivel air admittance fitting resulted
much better under any test condition, and, in particular, combined drainage from all 3
sanitary fixtures was almost 6 times as high as the other stack vent opening
configurations. Thus, extremely high drain performance of a swivel air admittance
fitting was verified.
Fig.8 also shows, where negative pressure exceeded the evaluation criterion of 400[Pa], seal loss ( h) was recorded higher than 25[mm] causing complete seal break.
However, provision of a swivel air admittance fitting recorded seal loss of as low as
6.5[mm] with fully combined fixture drains, and its outstanding feature to minimize
concern of seal loss was verified.
338 5.
Conclusion
From the results of our constant flow tests and sanitary fixture tests as introduced here,
we could confirm that use of our new swivel air admittance fittings helped considerably
moderate the pressure generated by the whole drainage system of a single-family
housing and remarkably improve its drainage performance. Comparative survey of stack
vent opening configurations and stack drain piping sizes was a key factor for our
successful project of this time.
6.
Acknowledgements
This study was conducted partially based on “A Study of a Drainage System which
Enables the Free Planning of Water Distribution Space”, (Chief researcher: Masayuki
OTSUKA), Grant-in-Aid for Scientific Research (C) 2009 by the Ministry of Education,
Culture, Sports, Science and Technology.
7. Previous studies
1) HONGO, Norihiro et al.; Study on Drainage Capacity Using a Vent System for
Drainage Systems of Detached Houses (Report 2) The Drainage Capacity of the
Fitting with Air Admittance Valve, Summaries of Technical Papers of Annual
Meeting Architectural Institute of Japan (Sep. 2008)
2) The Society of Heating, Air-Conditioning and Sanitary Engineers of Japan; SHASES206-2000 Plumbing Code
3) The Society of Heating, Air-Conditioning and Sanitary Engineers of Japan; SHASES218-2008 The Current Status of Drainage Systems for Apartment Housing
Complexes and Proposals
4) OTSUKA, Masayuki et al.; Experimental Study on the Evaluation of the Drainage
Capacity of Drainage Systems for Low-rise Housing - Discussion about Vent
Opening Resistance and Influence of Drainage Load on the Drainage Capacity Journal of Architecture, Planning and Environmental Engineering (Transactions of
Architectural Institute of Japan) (Aug. 2001)
8. Presentation of Authors
(1) Kazutoshi Suzuki is a member of KITZ Corporation. He carries out
studies on drainage and ventilation of stack systems for single-family
housing. He is also interested in studies of hot-water supply systems.
(2) Masayuki Otsuka is a professor at Department of Architectural
Environmental Engineering, Kanto Gakuin University. He is a
member of AIJ (Architectural Institute of Japan) and SHASE
339 (Society of Heating, Air-Conditioning and Sanitary Engineers of Japan). His current
research interests are the performances of plumbing systems, drainage systems with
drainage piping systems for SI (support and infill) housing.
(3) Norihiro HONGO, a SHASE member, is a lecturer at Department of
Architectural Environment System Engineering, Yamagata College of
Industry and Technology.
(4) Koichi Kawasaki, a SHASE member, is an engineer of Product
Development Department, KITZ Corporation, Japan. His
assignments at KITZ Corporation include development of plumbing
products for residential buildings.
340 VI.2
A Study of a Prediction Method for Drainage
Performance of Drainage Stack Systems
Using a Horizontal Fixture Drain Branch
System with an Air-Admittance Valve
(1) Masayuki Otsuka
(2) Zhe Zhang
(5) dmotsuka@kanto-gakuin.ac.jp
Department of Architecture College of Engineering, Kanto Gakuin University,
Dr. Eng, Japan
(6) m0943006@kanto-gakuin.ac.jp
Graduate Student, Graduate School of Engineering, Kanto Gakuin University
Keywords
Prediction Method, Drainage Performance, Drainage Stack Systems,
Fixture Drain, Branch System , Air-Admittance Valve
Horizontal
Abstract
This study intends to understand improvement effects on drainage performance when an
air-admittance valve is installed to a horizontal fixture drain branch and to propose a
method to predict such improvement effects. This paper firstly (1) discusses the
characteristics of drainage pipe pressure when an air-admittance valve is installed to
each of multiple horizontal fixture drain branches which are connected to a high-rise
single stack system, and then (2) proposes a piping model for predicting pipe pressure
variation, which is an evaluation index for drainage performance. This paper then
finally (3) actually uses the piping model to predict the drainage performance of a
drainage stack system compliant with SHASE S 218 so as to discuss and report the
effectiveness of the model.
341 1. Background of the study
In Japan, the use of air-admittance valves, as shown in Fig. 1, was approved by the
Construction Ministry under the former Building Standards Act, Article 38, providing
that an air-admittance was installed to the top end section of a vent pipe and/or to a
sanitary fixture pipe or a horizontal fixture drain branch for low-rise 3-storey housing.
However, since then, the approval of air-admittance valves has been focusing much on
the prevention of self-siphonage when they are installed to sanitary fixture pipes and
horizontal fixture drain branches, and not so much about improving the drainage
performance of stack systems. The reason being that at the time of the approval, there
was insufficient data and theoretical basis to discuss drainage performance improvement
in relation to the use of air-admittance valves.
Conventional vent cap
Open into the air
Air-admittance
valve
Treated inside
Inlet opening
Under the roof
Outside wall
Roof
Access opening
Typical roof vent installation
Typical sanitary fixture plumbing
(low-position air-admittance valve)
Fig. 1 Typical air-admittance valve installation in
Hence, at present, if any existing high-rise apartment houses suffer trouble with
drainage vent stack systems after they have been built, causing poor drainage
performance and broken traps, it is difficult to install new vent pipes to improve
drainage performance and such renewal work should therefore be avoided. Moreover,
the adoption of drainage systems with special fittings and with a two-pipe system
creates an issue, as they are costly, although they prevent trouble and ensure good
drainage performance. Therefore, it is necessary to develop a simple drainage vent stack
system with much focus on small size and capability of recovering the drainage
performance from trouble without causing too much interruption, or providing good
drainage performance similar to the capacity of drainage systems with special fittings
and with two-pipe system.
342 2. Overview of the experiment
(1) Experimental system
The environmental construction simulation tower at Kanto Gakuin University is used in
the drainage experiment. The tower comprises test drainage stack systems, as shown in
Fig. 2 and Fig. 3. The test systems are drainage vent stack systems which use JIS-DT
fittings. Fig. 2 shows a system with a straight house drain and Fig. 3 shows a system
comprising a house drain which is bent horizontally and installed 1m away from the
centre of the drainage stack. These two structures are called Fig. 2: standard system
(straight), and Fig. 3: standard system (bent). The systems on the 5th and 7th floors, i.e.
intermediate floors, each comprise an air-admittance valve which is installed to the end
section of the horizontal fixture drain branch, and these systems are called the horizontal
drainage vent system (straight) and the horizontal drainage vent system (bent). Hence,
four test drainage systems are used in the experiment, as described above.
vent system installed
W
P
Stack Dia.100A
P
P
4F
Horizontal fixture drain branch Dia.75A
P
Gradient 1/50
3F
100A
5F
P
4F
P
3F
P
2F
P
1F
GL
100×125A
125LL
P
House drain Dia.125A Gradient1/150
00
115
1F
GL
House drain Dia.125A Gradient1/150
00
105
unit[mm]
Fig. 2 Standard system (straight-pipe
vent stack system)
343 3750
125A
2F
600
P
3000 1200
6F
P
W
P
3000
3000
P
7F
3000
5F
P
3000
Horizontal drainage
8F
3000
6F
W
P
P
10
00
600
P
75LL
Drainage load 0.5~2.5[L/s]
W
9F
3000
3000
7F
P
3000
W P
8F
3000
P
P
Drainage load 0.5~2.5[L/s]
Pv
P
3000
W
P
W
9F
3000
Drainage load 0.5~2.5[L/s]
Pv
P
3750
Drainage load 0.5~2.5[L/s]
3000
P : Pressure sensor
3000 1200
W
: Heat-wire anemometer
0
100
Fig. 3 Standard system (bent-pipe vent
stack system)
(2) Airflow characteristics of the test air-admittance valve
Fig. 4 shows the test air-admittance valve (diameter: 75mm) which is installed to the
horizontal fixture drain branch, and Fig. 5 shows the airflow characteristics of the test
valve. These characteristics clarify the relationship between the airflow rate and the
airflow resistance coefficient of the test valve, ζAV, which can be used for the piping
calculation later in this paper. Incidentally, the pressure which lifts the cap of the test
valve, thus, activating it, is approx. -30 to -40Pa while the airflow rate is approx. 5 to
6L/s.
Fig. 4 A test air-admittance valve installed to the horizontal fixture drain
branch
344 Airflow resistance coefficient ζ 2BV 、ζ 2BV'
1.5[L/s] 2.0[L/s] 2.5[L/s] 3.0[L/s] 3.5[L/s] 4.0[L/s] 4.5[L/s] 5.0[L/s]
Pressure loss [Pa]
1.5[L/s] 2.0[L/s] 2.5[L/s] 3.0[L/s] 3.5[L/s] 4.0[L/s] 4.5[L/s] 5.0[L/s]
- 60
ζ2BV,
100
- 50
- 40
60
- 30
Start-up Pressure range
40
- 20
20
Pressure loss
P[Pa]
Airflow resistance coefficient
ζ
80
- 10
0
0
0
2
4
6
8
10
12
14
16
5
Airflow rate Qa[L/s]
Fig. 5 Airflow characteristics of the test air-admittance valve (dia.
75mm)
(7) Method of applying drainage load
Consistent with SHASE-S218, one floor worth of constant flow load of 0.5 to 2.5L/s is
applied from the top floor at 0.5L/s intervals. A drainage load of 2.5L/s is also applied
from the top floor while further drainage load is added from lower floors in the same
manner.
(8) Items to measure and measuring methods
Pipe velocity variation in the vent pipe section and the horizontal fixture drain branch
section is measured with a heat-wire anemometer, and pressure variation in the
horizontal fixture drain branch is measured with a pressure sensor on each floor. The
reference value for pipe pressure variation is within ±400Pa which is the trap breaking
point. Airflow rate variation is calculated from the pipe velocity variation.
345 3. Experiment results and discussion
(1) Pipe pressure distributions and airflow rates in comparison
Fig. 6 compares the pipe pressure distributions (ave, max. and min. values) of the
standard system (straight) and the horizontal drainage vent system (straight). The graph
presents the results from applying a constant flow load of 2.5[L/s] from the 8th and 9th
floors, i.e. 5.0[L/s] in total. The pipe pressure distributions are distinctive with the
horizontal drainage vent system in the sense that the pipe pressure becomes reduced on
the floors where the air-admittance valves are installed, but immediately after that,
negative pressure develops again on the lower floors. Taking a closer look at the
minimum system pressure, Psmin, which is the drainage performance index, it is 1200Pa with single stack vent system whereas it is reduced to -380Pa with the
horizontal drainage vent system. This suggests that the installation of the air-admittance
valves reduces Psmin by 30% compared to the single stack vent system.
Fig. 7 similarly compares the results between the standard system (bent) and the
horizontal drainage vent system (bent). The results clarify that the standard system
(bent) reduces the pipe pressure by 30% compared to the horizontal vent system (bent).
However, when looking at the maximum system pressure, Psmax, it is +300Pa on the
standard system (bent) whereas it increases to +370Pa on the horizontal drainage vent
system (bent). Hence, it is clear that Psmin is reduced but Psmax increases by 30%
compared to the standard system (bent).
Fig. 8 and Fig. 9 compare the total airflow rates of the experimental systems. In Fig. 8,
the total airflow rate of the standard system is 27.6[L/s] when the house drain is straight
whereas the total airflow rate of the horizontal drainage vent system is 35.7[L/s]. In Fig.
9, when the house drain is bent, the total airflow rate of the standard system is 27.8[L/s]
whereas the total airflow rate of the horizontal drainage vent system is 33.7[L/s]. These
findings confirm that the installation of the air-admittance valves to the horizontal
fixture drain branch increases the flow rate of sucked air. Hence, when the same load
flow rate is applied, the level of hydraulic jump (airflow resistance), which develops in
the house drain section, is more or less the same both on the standard system and on the
horizontal drainage vent system. This suggests that the positive pressure increases by
the same level as the airflow rate does, probably casting an issue to be addressed.
346 Vent system
Min. Pressure
Ave. Pressure
Max. Pressure
Single stack system (standard system)
8F
Air flow
7F
6F
Floor [F]
Floor [F]
Air flow
5F
4F
4F
3F
3F
2F
2F
Psmax
1F
-1600 -1200
-800
-400
0
400
Fig. 6 Total airflow rates of the
standard system and the
horizontal drainage vent
system in comparison (house
drain: straight)
-1600 -1200 -800
-400
0
400
800
Fig. 7 Total airflow rates of the
standard system and the
horizontal drainage vent system
in comparison (house drain:
bent)
40
40
Horizontal drainage vent system
Horizontal drainage vent system
35
35
Single stack system
(standard s stem)
25
20
15
10
T0tal airflow rates Qav
30
Air flow
5F
1F
T0tal airflow rates Qav
Air flow
7F
6F
30
25
Single stack system
(standard s stem)
20
15
10
5
5
0
0
Fig. 8 Total airflow rates of the
standard system and the
horizontal drainage vent
system in comparison (house
drain: straight)
Fig. 9 Total airflow rates of the
standard system and the
horizontal drainage vent
system in comparison (house
drain: bent)
347 Load-applying
Psmin
8F
9F
floor
Load-applying
9F
floor
Horizontal drainage vent system
(2) Proposal and analysis of the pipeline model
Fig. 10 and Fig. 11 show the pipe pressure distributions of the horizontal drainage vent
system and the piping model for the system. The airflow rate of the vent stack is Qa1,
and the airflow rate of the horizontal fixture drain branch is measured in two locations,
Qa2 and Qa3. Three closed loops; (f(Qa1, Qa2), g(Qa1, Qa2, Qa4) and h(Qa1, Qa2,
Qa3), are also shown in the diagram, providing balance equations for pressure
resistance and suction force which are generated inside the pipes. They are analyzed
using successive approximations to acquire airflow rates Qa1, Qa2 and Qa3.
Airflow Classification
Stack venting only
(1)+stack venting(1 location)
(1)+stack venting(2 location)
(1)
(2)
(3)
Qa1
[m]
P2
P1
Balance equations
f (Qa1、Qa2、 )=0
g (Qa1、Qa2、Qa3 )=0
h (Qa1、Qa2、Qa3 )=0
P1V P1S
9F+8 F
1VBV Region
1
P1BV
P4
P3BV P2BV
1SBV Region
1VBV'Region
2
'
P4BV
1SBV'Region
3
(-)
P4BV’
P5BV
(+)
[Pa]
Fig. 10 Horizontal drainage vent system model
by pipe pressure classification
1V Region
1S Region
2 Region
3 Region
4 Region
2 BV Region
3 BV Region
4 BV Region
ζ1V
ζ1S
Drainage
Qa2
PLV1
Qa4 ζ
2 BV
ζ3 BV
ζ1VBV
ζ2 SBV
f(x)
Qa3
( ⊿P
⊿L ) ζ1VBV'
BV
2 BV
' Region
ζ2BV' ζ1SBV'
3 BV'Region
ζ3BV'
4 BV'Region
( ⊿P
⊿L )'
BV
5 BV'Region
ζ5BV'
g(x)
Qa5
h(x)
Fig. 11 Piping model for the horizontal
drainage vent system
Within the structure of the piping model in Fig. 10, the negative pressure, P2BV’, which
is generated immediately after the air flows sucked in from three vent pipe sections join
together, is the most characteristic airflow resistance, and the relationship between the
coefficient of that airflow resistance, ζ2BV’, and the total airflow rate, Qa, is shown in
Fig. 12. Based on this, it is observed that Qa does not create much impact, i.e. being
constant with each total drainage load flow and it is therefore treated as the average
value. The other airflow resistance regions and suction regions are sorted in the same
manner.
In addition, Fig. 13 shows the total airflow rate, Qa, both with the straight house drain
pipe and with the bent house drain pipe in relation to the airflow resistance, ζ5BV’, which
348 Airflow resistance coefficient ζ2BV’
is acquired from the positive pressure, P5BV’, which is generated in the horizontal fixture
drain branch on the 1st floor. It has become apparent that with the straight house drain
pipe, the total drainage flow rate and the airflow rate do not create much impact whereas
with the bent house drain pipe, the airflow resistance increases by two to six times.
20
1.5[L/s] 2.0[L/s] 2.5[L/s] 3.0[L/s]
3.5[L/s] 4.0[L/s] 4.5[L/s] 5.0[L/s]
15
13.61
12.22
12.20
11.67
10.44
10
7.98
5.46
5
0
0
5
10
15
20
25
30
35
Total airflow rate Qa[L/s]
Airflow resistance coefficient ζ5BV’
Fig. 12 Total airflow rate Qa in relation to airflow resistance coefficient ζ2BV’
in 2BV’ region in each of which an air-admittance valve is installed
100
Drainage system
3.0[L/s] 3.5[L/s]
4.0[L/s]
4.5[L/s]
5.0[L/s]
Straight
80
Bent
60
y=-3.48x+175.89
y=-0.84x+78.10
y=-3.12x+144.60
y=-3.3167x+140.31
40
y=-1.56x+77.51
20
9.70
0
26
28
30
32
34
36
38
40
Total airflow rate Qa[L/s]
Fig. 13 Total airflow rate Qa in relation to airflow resistance coefficient ζ5BV’
in 5BV’ region in each of which an air-admittance valve is installed
349 (3) Qa1, Qa2 and Qa3 of each system and measured and predicted Psmin and
Psmax in comparison
Shown in Fig. 14 (1)(2)and Fig. 15(1)(2) are predicted and measured values of airflow
rates, Qa1, Qa2 and Qa3, when the house drain pipe is straight and when it is bent. It is
noticeable that errors occur when values of Qa2 and Qa3 from the horizontal fixture
drain branch are small. This is because unsteady fluctuation is caused by the repetitive
open/close action of the air-admittance valve cap at a pressure of 30 to 40 Pa. These
predicted values are used for acquiring the average pressure on each floor, fluctuation
components are added, and then the minimum system pressure, Psmin, the reference
value for negative pressure, and the maximum system pressure, Psmax, the reference
value for positive pressure, are calculated. Predicted and measured values with each
system are compared in Fig. 16 and Fig. 17. The differences between the predicted and
measured values roughly fall within the error tolerance of ±10%, thus, confirming the
effectiveness of the prediction method.
30
30
Qa1
Qa2[L/s]
1.5[L/s] 2.0[L/s] 2.5[L/s] 3.0[L/s]
25
3.5[L/s] 4.0[L/s] 4.5[L/s] 5.0[L/s]
3.5[L/s] 4.0[L/s] 4.5[L/s] 5.0[L/s]
20
20
Qa3[L/s]
1.5[L/s] 2.0[L/s] 2.5[L/s] 3.0[L/s]
15
Predicted value
Qa2,Qa3[L/s]
Predicted value Qa1[L/s]
1.5[L/s] 2.0[L/s] 2.5[L/s] 3.0[L/s]
25
10
5
0
15
3.5[L/s] 4.0[L/s] 4.5[L/s] 5.0[L/s]
10
5
0
0
5
10
15
20
25
Measured value Qa1[L/s]
30
0
5
10
15
20
25
30
Measured value Qa2,Qa3[L/s]
(1)
(2)
Fig. 14 Predicted and measured values of airflow rate Qa1,Qa2,Qa3
in comparison (house drain: straight)
350 30
30
25
3.0[L/s] 3.5[L/s] 4.0[L/s] 4.5[L/s] 5.0[L/s]
20
Predicted value
Qa1[L/s]
15
10
5
0
0
5
10
15
20
25
Qa2[L/s]
3.0[L/s] 3.5[L/s] 4.0[L/s] 4.5[L/s] 5.0[L/s]
Predicted value Qa2,Qa3[L/s]
Qa1[L/s]
25
Qa3[L/s]
3.0[L/s] 3.5[L/s] 4.0[L/s] 4.5[L/s] 5.0[L/s]
20
15
10
5
0
30
0
5
10
15
20
25
Measured value Qa2,Qa3[L/s]
30
Measured value Qa1[L/s]
(2)
(1)
Fig. 15 Predicted and measured values of airflow rate Qa1,Qa2,Qa3
in comparison (house drain: bent)
500
1.5[L/s] 2.0[L/s]
2.5[L/s] 3.0[L/s]
-300
Predicted value Psmax[Pa]
Predicted value Psmin[Pa]
-400
3.5[L/s] 4.0[L/s]
±10%
4.5[L/s] 5.0[L/s]
-200
-100
0
3.0[L/s] 3.5[L/s] 4.0[L/s]
400
4.5[L/s] 5.0[L/s]
±10%
300
200
100
0
0
-100
-200
-300
-400
0
Measured value Psmin[Pa]
Fig. 16 Predicted and measured values
of the minimum system
pressure Psmin in comparison
(house drain: straight)
100
200
300
400
500
Measured value Psmax[Pa]
Fig. 17 Predicted and measured values
of the maximum system
pressure Psmax in comparison
(house drain: bent)
(4) Prediction of drainage performance
Fig. 18 shows the total drainage load flow rate, Qw, both with the standard system
(straight) and with the horizontal drainage vent system (straight), in relation to Psmin
and Psmax which provide criterion indices. The table lists both predicted and measured
values, and they are more or less identical at all total load flow rates. When the drainage
351 performance of the standard system (straight) is 2.0L/s, the drainage performance of the
horizontal drainage vent system (straight) is 5.5L/s, showing 2.75 times better
performance. The allowable flow rate of the loop vent system (drainage stack diameter:
100A), which is specified by SHASE-S 206, is 6.66L/S, and the installation of the
horizontal drainage vent system in two locations certainly enables better drainage
performance by approx. 80% max. compared to the SHASE-S 206 allowable flow rate.
Fig. 19 shows the results which were acquired on the standard system (bent) and the
horizontal drainage vent system (bent). When comparing to the drainage performance of
the standard system bent, the drainage performance of the horizontal drainage vent
system bent is 5.0L/s, enabling 2.5 times improvement.
In addition, the installation of the air-admittance valves to the horizontal fixture drain
branches, which are installed in two locations, enables approx. 75% max. better
drainage performance than that of the SHASE-S 206 loop vent system (drainage stack
diameter: 100A).
Single stack vent system
(House drain:straight)
Psmin
Psmax
Horizontal drainage vent system 2 locations
(House drain:straight)
Measured value
Predicted value
Measured value
Predicted value
400
200
0
※SHASE-S 206 Two-pipe system
-200
Psmax, Psmin
[Pa]
-400
-600
-800
SHASE-S 218
-1000
-1200
-1400
-1600
-1800
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0
Total load flow rate Qw [L/s]
Fig. 18 Total drainage load flow rate Qw in relation to
Psmin and Psmax (house drain: straight)
352 Single stack vent system
(House drain:bent)
Psmin
Psmax
Horizontal drainage vent system 2 locations
(House drain:bent)
Measured value
Predicted value
Measured value
Predicted value
600
400
200
SHASE-S 218
0
Psmax, Psmin
[Pa]
R f
-200
-400
-600
※SHASE-S 206
-800
T
-1000
i
t
-1200
-1400
-1600
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0
Total load flow rate Qw [L/s]
Fig. 19 Total drainage load flow rate Qw in relation to
Psmin and Psmax (house drain: bent)
4. Conclusion
This report proposes a method to predict the drainage performance of horizontal
drainage vent systems for high-rise buildings (providing that they are installed in two
locations along a drainage system), which each employ air-admittance valves, as well as
examining the effectiveness of the method. Consequently, the following points have
been clarified:
(1) The horizontal drainage vent system can improve the drainage performance of the
single stack vent drainage system by 2.5 to 2.75 times.
(2) The horizontal drainage vent system ensures 75 to 80% better drainage performance
than the drainage performance of the SHASE-S206 loop vent system.
(3) A piping model for predicting the drainage performance of the horizontal drainage
vent system has been proposed and its effectiveness has been confirmed.
(4) On the single stack vent drainage system, even when the drainage performance
declines, the compact size and good constructability of the horizontal drainage vent
system are beneficial and the horizontal drainage vent system can therefore be
positioned as a technology which enables the recovery of drainage performance.
353 Acknowledgement
This study was conducted partially based on “A Study of a Drainage System which
Enables the Free Planning of Water Distribution Space”, (Chief researcher: Masayuki
OTSUKA), Grant-in-Aid for Scientific Research (C) 2009 by the Ministry of Education,
Culture, Sports, Science and Technology.
References
1) Masayuki Otsuka, Jian MA, Yuta TAKAHASHI; Study on the Performance
Evaluation of Air Admittance Valves by means of the Drainage Capacity Prediction
method for Drainage Stack Systems – Influences on vent characteristics of air
admittance valves (AAVs) and drainage capacities (Transactions of Architectural
Institute of Japan Jun. 2008)
2) Masayuki OTSUKA and Jian MA; Studies of a Testing Method for Air Admittance
Valve Characteristics and Design for Vent Pipes, CIB W062 Symposium, Czech
Republic (2007.9)
3) Masayuki OTSUKA, Heizo SAITO, Fudetoshi UEDA; Experimental Study on
ventilating efficiencies of drain pipe system (part 1) about limit loading of variable
vent system (Transactions of the Society of Heating, Air-Conditioning and Sanitary
Engineers of Japan Oct. 1989)
4) Masayuki YAMADA, Masayuki OTSUKA; Study of the Drainage Performance
Improvement Technique of Drainage Stack System by means of Air Admittance
Valves(Part1) Study on Horizontal Fixture Branch Vent System.(Transactions of the
Society of Heating, Air-Conditioning and Sanitary Engineers of Japan Sep. 2008)
5) Zhang Zhe, Masayuki OTSUKA, Masayuki YAMADA; Study of the Drainage
Performance Improvement Technique of Drainage Stack System by means of Air
Admittance Valves(Part2) An application of the plural Air Admittance Valves to the
horizontal fixture branch(Transactions of the Society of Heating, Air-Conditioning
and Sanitary Engineers of Japan Sep. 2009)
6) Koichi KAWASAKI, Masayuki OTSUKA, Kazuhiro SHIGETA and Norihiro
HONGO; Study on How to Improve the Capacity of a Drainage System Air
Admittance Valves, CIB W062 Symposium, Brussels, Belgium (2005.9)
354 Presentation of Authors
Masayuki Otsuka is a professor at the Department of
Architecture, Kanto Gakuin University. He is a member of
AIJ (Architectural Institute of Japan) and SHASE (Society of
Heating, Air-Conditioning and Sanitary Engineers of Japan).
His current research interests are the performances of
plumbing systems, drainage systems with drainage piping
systems for SI(Support and Infill)housing and the
performance evaluation of water saving plumbing systems.
Zhe Zhang is a master of the Otsuka laboratory, Kanto
Gakuin University. His current study interests are the
Drainage Performance Improvement Technique of Drainage
Stack System by means of Air Admittance Valves and
especially in an application of the plural Air Admittance
Valves to the horizontal fixture branch.
355 VI.3
Application
of
Computational
Fluid
Dynamics to Simulate the Behaviour of Fluids
inside
Vertical
Stack
of
Building
Drainage System
Eric Wai Ming Lee
ericlee@cityu.edu.hk
Department of Building and Construction, City University of Hong Kong,
Tat Chee Avenue, Kowloon Tong, Hong Kong (SAR), People’s Republic of China.
Abstract
A numerical method is presented to simulate nonlinear free surface flow in a vertical
stack building drainage system by computational fluid dynamics (CFD). The volume of
fluid (VOF) method is adopted to track the air/water interface, and the effect of the
interfacial surface tension is simulated by the continuum surface force (CSF) model.
The numerical results are validated against full-scale experimental data. The results
demonstrate the behaviour of the air pressure and the annular drain flowing along the
vertical stack. They also show the reduction in air pressure when the air flow passes
through the branch inlet and the existence of positive air pressure at the base of the
vertical drainage stack. The CFD simulation results are similar to the experimental
results.
Keywords
Building Drainage; Computational fluid dynamics; Continuum surface force; Drainage
stack; Multiphase flow; Volume of fluid
356 1. Introduction
Since the outbreak of severe acute respiratory syndrome (SARS) in 2003, the vertical
drainage pipe system that connects with the ventilation system within a building has
been postulated as a possible channel that facilitates the cross contamination of this
deadly disease [1]. To minimise the risk of the disastrous spread of SARS, much
attention has been directed towards better understanding the free surface flow structure
to improve the existing design of building drainage ventilation systems. Conventionally,
design guidelines that are formulated based on statistical theories are used in the design
of practical drainage systems [2, 3]. However, because these guidelines do not address
the interior nonlinear free surface flow structure, they are inadequate for the design of
complex drainage systems in super-high-rise buildings. A numerical model (AIRNET)
was developed by Jack and Swaffield [4] in their pioneering work in building drainage
simulation. However, the model is a one-dimensional unsteady flow model, and the
nonlinear interactions between the air and liquid phases inside the drainage pipes are not
described in detail. This paper presents a detailed methodology to investigate the free
surface flow behaviour in a vertical pipe system using the volume of fluid (VOF) method.
The model predictions are compared against the experimental data of a single vertical
stack system that are obtained from Lu [5]. Particular attention is directed towards the
visualisation of the complex flow structure.
2. Mathematical Modelling
The air flow in a vertical stack is initiated by the shear between the annular water film
and the air core [6]. To resolve the hydrodynamics of air and water, which travel at
different velocities, an inhomogeneous two-fluid model that consists of two sets of
conservation equations is adopted. Appropriate inter-phase constitutive relationships are
applied to close the inhomogeneous two-fluid model for the air and water flow. These
inter-phase constitutive relationships represent the drag and non-drag forces that act on
such a flow. The non-drag forces are modelled according to the virtual mass, lift, wall
lubrication and turbulent dispersion.
The interface between the air and water is tracked using the volume of fluid (VOF)
method. This method uses a volume fraction function to define the volume portion of
one particular fluid inside a cell volume. As the sum of the volume fraction is equal to
one and has no inter-phase mass transfer, a transport volume fraction conservation
equation is given by the following equation where
travelling velocity vector of phase .
1
·
is the density and
is the
0
This equation represents the advection algorithm to advance the interface, which
simultaneously conserves the volume fraction. To account for the effect of the
357 interfacial surface tension, the continuum surface force (CSF) model is adopted. By
combining the surface force with the volume force that is concentrated at the interface,
the surface tension force is expressed as
of which
σ and
, is the surface tension coefficient,
is the normal vector that
goes from phase to phase , is the surface curvature of the interface and
is the
gradient operator at the interface. The turbulence of the air and water flow is simulated
by the inhomogeneous shear stress transport (SST) model, which incorporates the k-
model for the flow within the boundary layer and the k- model for the flow in the free
stream.
3. Numerical Results and Validation
12m The numerical simulations were carried out using the generic CFD code ANSYS CFX
10. The predictions of the numerical models were then validated against the
measurements of the full-scale single-pipe system of Lu [5]. The experiment was
performed on an experimental tower attached to a building 40 m in height. In the tower,
a main vertical stack 100 mm in diameter was installed and connected with branches (75
mm in diameter) to each floor and the building’s drain pipe (125 mm in diameter). The
average air pressure was measured at each floor using small diffuse semiconductor air
pressure transducers. Water flow was discharged at the ninth floor (i.e., 27 m above
ground level) at a flow rate of 2.0 l/s. Figure 1 is a schematic diagram of the
experimental setup.
2m 2.0 lit/s
75mm Horizontal
27m 100mm Vertical Stack 125mm Horizontal
Outflow
2m
Figure 1: Schematic of the Experimental Setup
358 A three-ddimensional computatioonal modell containingg roughly 22.3 million tetrahedrall
elements was
w generaated over thhe entire do
omain. The water discharge at th
he inlet wass
specified using CFX
X commandd language (CCL). Forr the top open end an
nd outlet, a
relative avveraged staatic pressurre of zero was speciffied. Figuree 2 depictss the meshh
distributioon of the coomputationaal model at the junctionn between tthe branch and
a verticall
stack.
Figure 2:
2 mesh disstribution of
o the comp
putational model
m
at th
he junction
n between
the branch and
d vertical stack.
s
The measuured and prredicted air pressure vaalues at eacch floor are presented in
i Figure 3..
In generall, the prediccted pressurre variation along the vertical
v
stacck agrees well with thee
experimenntal data. Thhe negative air pressuree is created by the entrry loss at the top of thee
drainage stack.
s
The wall frictioon further reduces the air pressurre from the top of thee
stack dow
wn to the braanch inlet. Itt is captured
d in the lenggth from 400m down to 26m abovee
the stack base
b
as shown in Figure 3. As sho
own in Figuure 4, the w
water dischaarging from
m
the branchh inlet occuppies a portion of the crross-sectionnal area of thhe pipe thro
ough whichh
the entrainned air is passing.
p
Thiis reduction
n in the effeective air passage areaa acts as ann
orifice andd reduce thhe air pressuure at the im
mmediate downstream
d
of the bran
nch inlet. Itt
can be obbserved at 24m
2
high in
i Figure 3. The existtence of thee positive air
a pressuree
located att the immeediate upstrream of thee stack basse is also ccaptured by
y the CFD
D
simulationn as shown from 0 to 5m
5 high in Figure 3. Detailed
D
flow
w structure is depictedd
in Figure 5.
5 The air enntrained intto the stack is blocked by
b the wateer curtain at the base off
the stack. The air acccumulates there untill its pressurre is sufficiently high enough too
‘blow’ thrrough the water curtaain. At thee downstreaam of the water curtaain, the airr
pressure reesumes to atmospheric
a
c pressure. Referring
R
too the air preessure profille in Figuree
3, at the middle
m
of the vertical sttack (i.e., 10
0 to 20 m high),
h
the air pressure was
w slightlyy
359 under-predicted. This could be attributed to the uncertainty in the drag force calculation,
in which the air entrainment may be underestimated because of the shear of water. It
will be further investigated in future.
40
+
x
+x
HEIGHT ABOVE STACK BASE (M)
35
+x
+x
30
+ x
25
+
x
+x
20
+ x
15 +
x
+
10
x
+x
x +
5
x+
+
x
0
-20
-10
0
10
20
AIR PRESSURE (mm H 2O)
Figure 3: Air pressure profile along the height of the vertical drainage stack.
Figure 4 shows the free surface flow structure at the branch/stack junction. The free
surface is highlighted in blue. As the figure shows, water discharging from the branch
becomes annular flow at the downstream of the junction. The air is drawn by the
annular drain from top of the drainage stack to the core of the drainage stack by the
shear at the interface between water and air of the annular drain. Figure 5 depicts the
flow structure at the base of the vertical drainage stack. It shows that the annular drain
from the vertical stack forms a water curtain at the base of the stack. The water curtain
causes an occlusion of the air. When the air at the upstream of the curtain (i.e. 0 to 5m
high of the stack) gains sufficient positive pressure, it blows through the curtain and
reach the horizontal drain at which the air pressure is close to atmospheric pressure.
360 Air is drawn from atmosphere to the stack by the shear of the annular drain
Vertical drainage stack
Branch inlet
Flow of branch inlet occupies a portion of the cross‐
sectional pipe Annular drain
Figure 4: Detailed flow structure at the branch/stack junction
100 Vertical Drainage Stack
Water curtain covers the drain 100 Horizontal Drain Pipe Z
Y X
Figure 5: Isometric view on detailed flow structure at the base of the vertical stack
361 4. Conclusion
A detailed methodology for solving the complex nonlinear free surface flow in a
vertical stack building drainage system is presented. The numerical results are validated
against the full-scale experimental data of Lu [5]. Comparison of the predicted and the
measured results shows that they are in good agreement. Also, the simulation results
capture the following 5 important observations of vertical drainage stack flow as
follows.
a)
Annular drain is formed at the downstream of a branch inlet.
b) The annular drain draws air from atmosphere to the stack by the shear at the
interface of air and water of the annular drain.
c)
Air pressure is reduced at the immediate downstream of the branch inlet due to the
orifice effect.
d) Water curtain forms at the base of the vertical stack.
e) Air pressure becomes positive at the immediate upstream of the water curtain.
This study demonstrates the advantage of using the volume of fluid (VOF) method to
investigate the free surface flow behaviour in a vertical drainage stack. The visualisation
of three-dimensional flow that is realised by the VOF method offers a great way to
improve the existing design of building ventilation drainage systems, which may
effectively prevent cross contamination of fatal diseases via the building drainage
ventilation systems of high-rise buildings.
Acknowledgements
The work described in this paper was fully supported by a grant from the Research
Grants Council of the Hong Kong Administrative Region, China [Project no. CityU
115507].
362 5. References
1.
WHO (2003). Risk factors involved in the possible environmental transmission of
severe acute respiratory syndrome (SARS) in specified residential buildings in the
Special Administrative Region of Hong Kong, World Health Organisation Report,
World Health Organisation.
2.
Whitehead A. (2002). Plumbing engineering services design guide. The Institute of
Plumbing, Hornchurch, Essex, UK.
3.
BS EN 12056:2 (2000). Gravity drainage systems inside buildings – part 2:
sanitary pipework, layout and calculation, UK, British Standard Institute.
4.
Jack, L. B. and Swaffield J. A. (1999). Developments in the simulation of the air
pressure transient regime within single stack building drainage system. Proceedings
- 25th International Symposium of Water Supply and Drainage for Buildings CIB
W062 (pp. 1-11). Edinburgh, UK: Heriot-Watt University.
5.
Lu, W. H. (2005) Prediction method of air pressure on two-pipe stack building
drainage system, PhD thesis, National Taiwan University of Science and
Technology.
6.
Gormley, M. (2007) Air pressure transient generation as a result of falling solids in
building drainage stacks: definition, mechanisms and modelling, Building Services
Engineering Research and Technology (pp. 55-70).
6. Presentation of Author
Eric Lee is an assistant professor at City University of Hong
Kong in the Department of Building and Construction. He is
specialized in plumbing and drainage system design. He is also
the designer of the plumbing and drainage system of TwoInternational-Financial-Center which is the highest existing
building in Hong Kong.
363 Session VII: Drainage III
VII.1
IKT Comparative Product Tests
as a basis for asset management investment
decisions
B. Bosseler
bosseler@ikt.de
IKT - Institut für Unterirdische Infrastruktur, Gelsenkirchen, Germany
Abstract
To manage investment needs, while meeting technical and regulatory requirements,
reliable information and performance indicators are required for all kinds of techniques
and products used for construction, rehabilitation and operation of sewers. The objective
of IKT Comparative Product Tests (CPT) is to provide network operators with reliable
and independent information on the properties of commercial products. Such
information has been almost completely missing for the sewer construction and
rehabilitation area until now. The clients attain information on product characteristics
almost exclusively from advertisements and the offerers’ brochures, who try to convince
potential customers of the alleged quality of a product.
A central aspect of CPT is the practical product quality evaluation, e.g. under operating
conditions. The focus of the examinations is not only the compliance with individual
standards or bodies of rules and regulations, but the reliable fulfilment of network
operator requirements during construction and operation. The service life under the
expected conditions and loads, such as e.g. groundwater, earth pressures, volume of
traffic or high-pressure cleaning, are the focus of attention. As a result the network
operators are provided with independent, practice-related, and technically well-founded
information concerning the strengths and weaknesses as well as areas of application and
limits of the tested products.
Keywords
Product tests; investment decisions; house connection lines
364 Introduction
Maintenance of drain and sewer networks as an element in Asset Management is of
increasing importance for waste-water management companies all over the world.
Reliable information and performance indicators are needed concerning all types of
methods and products used for the construction, rehabilitation and operation of drains
and sewers, in order to permit appropriate investment management.
To meet this need, IKT - Institute for Underground Infrastructure has developed a tool
which compares the quality of techniques and products available on the market. This
tool integrates the requirements of network operators in order to apply a defined testing
program: it bears the name IKT-Comparative Product Test (IKT CPT).
IKT-CPTs are intended to provide water management companies with reliable and
impartial information on the strengths and weaknesses of products and methods, esp. in
waste-water management technology. They are performed in cooperation with network
operators, who are able to follow the tests in a series of meetings. A central aspect of
IKT-CPTs is practical product-quality evaluation, under installation and operating
conditions. The focus of the investigations is not only on compliance with individual
standards or rules and regulations, but also on reliable fulfillment of network operator
requirements during e.g. construction and maintenance. Service-life under anticipated
conditions, such as load, groundwater, soil pressure, volume of traffic and high-pressure
cleaning, is the prime concern. Network operators are thus provided with impartial,
practice-orientated and technically well-founded information on the strengths and
weaknesses of the products tested, and also on their potential applications and
limitations.
Waste-water management companies are informed quickly and comprehensively on
product quality, complete with an easily comprehensible evaluation procedure and a test
seal. After conclusion of an IKT-CPT, the products and methods tested are all
evaluated, receiving grades ranging from VERY GOOD to POOR.
Test participants and procedure
Customers/Municipalities
Waste-water management utilities and municipalities invest large amounts of money in
their networks. Numerous methods and products are available on the market. The
financial risk is borne by the network operators, however.
For this reason, a group of waste-water management companies and municipalities
commission IKT-CPTs. These network operators need to know what methods are best
for use in their drain and sewer networks. Their representatives closely monitor the
whole test procedure in project meetings, which coordinate all test elements, from
365 selection of products via the test program, through to evaluation. Eighty-one network
operators have up to now participated in IKT-CPTs. The benefits for them:

Selection of methods and products to be tested

Input of their own specific requirements

Interchange of information with other network operators

IKT observation and support on operators' sites

Tailor-made invitation-to-tender documentation

Minimization of investment risks, security of investment
In view of the great importance of this topic for public finances, various ministries of a
number of German states also participate in these tests, including, for instance, the
Ministry of the Environment, Nature Conservation, Agriculture and Consumer Affairs
of the State of North Rhine-Westphalia and the Ministry of the Environment of the State
of Baden-Württemberg, who are, indeed, currently involved in an IKT-CPT [30].
Project Manager
IKT, an neutral, independent, non-profit, research, testing and consulting institute, is the
project manager for the comparative product tests. IKT's work focuses on questions of
the installation and maintenance of underground networks and systems for gas, water
and waste-water. As an independent and reliable partner for utilities, drain and sewer
network operators, water boards and industry, IKT provides highly specialized,
advanced research and testing technologies on a practical and application-orientated
basis. Cooperation with water-management organizations has been IKT's standard
practice since its establishment in 1994. Well-known water-supply utilities and
drain/sewer operators cooperate in knowledge and know-how transfer between science,
research and industry, within the framework of the organizations which support IKT
(see IKT-Fördervereine at www.ikt.de/mitglieder).
The industry
The industry is integrated into the IKT-CPTs via the use of the products and processes
under examination. In every case, payment is made for the utilization of the relevant
products and processes, and the suppliers therefore have no influence on the test
procedure or on the individual tests. They are, however, able to select the contractor for
performance of the work necessary for the tests. This contractor is then able to work
[30] IKT- CPT “Repairing methods for main pipes”; IKT - Institute for Underground
Infrastructure; Gelsenkirchen (current project).
366 under optimum conditions, in, for example, the IKT test facility (no exposure to the
effects of weather, no time pressure, etc.).
Test procedure
The participating network operators firstly nominate the products for the IKT-CPTs.
The products selected are generally tested on three criteria: Quality assurance, system
tests and site investigations.
Quality assurance
IKT also validates the product suppliers' Quality Assurance systems. Quality Assurance
can have a great influence on the results obtained using the various methods and
products. The items investigated include:

The supplier's process description, e.g. procedure manuals

Training courses for qualification of operational personnel

Practical tests, e.g. test use in a drain or sewer section

Certification for the methods and products, e.g. a construction-supervision
procedure approval by government bodies such as the Deutsche Institut für
Bautechnik (German Construction Technology Institute, German abbreviation:
DIBt).
System tests
The products selected for the IKT-CPTs are evaluated on the criterion of their
applications (e.g. in experimental drain or sewer networks). Defined boundary
experimental conditions which reflect practical service conditions are specified jointly
with the participating network operators. A specially developed test program is used for
analysis of potential applications and limitations.
The use of the products in situations as authentic as possible in the system tests make
these the most important module in the IKT product trials.
Site investigations
The use of the various products and methods is in all cases accompanied in the context
of an IKT-CPT by in situ observation (e.g. on construction sites) with the participating
network operators. These in-situ evaluations serve the purpose of verifying whether the
use of the product or method in the context of system tests at IKT is genuinely
comparable with use, for example, in existing drain/sewer networks under in-situ
service conditions (e.g. traffic loads, weather, time pressure). The participating network
operators can themselves nominate specific installation projects, which are then
supported and observed by IKT.
367 Product assessment
Comparative assessment of the products is performed on the basis of the test results.
The central issues of Quality Assurance and system tests form the basis for the grading.
The main focus of the assessment is on the system tests, however. Site investigation
results cannot be included in determination of the test gradings, since conditions vary
from site to site. This data is therefore provided only as additional information.
Publication
All results are presented in a comprehensive Test Report, which provides detailed
information on test apparatus, the tests performed and the evaluation procedure. The
central element of this report is the IKT-CPT Table (see Table 1). The IKT Test Seal
symbol can be awarded, depending on the test results (see Figure 1).
In agreement with the clients, the results are published irrespective of the test verdicts
(VERY GOOD or INADEQUATE), in the IKT eNewsletter, on the IKT Homepage and
in the technical press. Publication involves open discussion of results, with the
participation of all clients, representatives of the industry, and the press.
Table 1 Example of an IKT-CPT test table; in this case, the results of the
IKT-CPT "Tube Liners for Lateral Pipes"; Test project: Repair of a so-called
"standard" situation
368 Figure 1:
Test seal grade: VERY GOOD
Examples of IKT-CPTs
A total of eighty-one network operators from all parts of Germany have up to now
participated in IKT-CPTs. Both small, medium-sized and large municipalities and
network operators have been involved. This is also reflected in the population sizes of
the municipalities/network operators involved, which vary between 20,000 and several
million. The items tested included, for example, products intended for installation of
new drain and sewer networks, repair and renovation procedures, and inspection
systems for drains and sewers. Various tests and their results are discussed below by
way of example.
IKT-CPT "Repair methods for lateral connections"
Twenty-six drain and sewer network operators from the cities of Ahlen, Alsdorf,
Beckum, Bergisch Gladbach, Braunschweig, Dinslaken, Dortmund, Düsseldorf,
Espelkamp, Essen, Gladbeck, Hamburg, Hamm, Hemer, Hilden, Iserlohn, Kamen,
Kempen, Monheim am Rhein, Neuss, Mönchengladbach, Recklinghausen, Rietberg,
Troisdorf, Tönisvorst and Warendorf participated in the IKT-CPT "Repair methods for
lateral connections" [31].
Seven injection and six top-hat liner methods were tested and comparatively evaluated.
The test was closely monitored by the twenty-six drain and sewer network operators in
a series of seven meetings.
The primary emphasis in the tests was on the use of injection methods and top-hat liners
for repair of simulated damage in test pipes. Various damage scenarios, the so-called
[31] Bosseler, B.; Kaltenhäuser, G.: IKT-CPT “Repair methods for lateral
connections”; IKT - Institute for Underground Infrastructure; Gelsenkirchen, June
2004; short report in English on www.ikt.de.
369 “standard” and “extreme” situations, were simulated. Completion of the repairs was
followed by testing and exposure to operational loadings.
The illustration below shows a simulated damage scenario and an example of a
completed repair.
Figure 2:
Simulated damage for evaluation of injection methods, and an
example of a visually satisfactory repair result. A Initial situation. B View from the
main pipe. C View into the connecting pipe.
The IKT-CPT "Repair methods for lateral connections" permits the following basic
conclusions:

The repair results are normally not comparable to those achieved with new
installation. Many of the methods tested fail to fulfill the test requirements. The
lateral connections were in most cases permeable immediately after the repair,
and nearly always after high-pressure flushing. High-pressure flushing stress
with added granulate in some cases resulted in substantial damage to the repair.
In many cases, the repair was unable to restore the disposal reliability of the
lateral connection, particularly in the case of so-called extreme damage; in a few
cases, the repair actually worsened the situation (e.g. imminent danger of
obstruction due to formation of wrinkles and edges).

The tightness of the repaired lateral connections can under no circumstances be
ascertained merely by means of visual inspection. Numerous repairs leaked
during the test, despite their visually satisfactory appearance. It is possible to
perform leak tests on drain and sewer networks only at considerable cost,
however. In addition, the value of such tests is reduced, due to sources of error,
e.g. discharge of the test fluids between the bladder and the inner wall of the
pipe. More development work on equipment and procedures for reliable
tightness tests is therefore necessary.

The result of the repair depends not only on the type of damage, but also on the
adhesive properties of the repair materials used vis-à-vis the particular pipe
materials (concrete, vitrified clay, etc.). Some injection methods, for example,
exhibit better results for repair of extreme damage in concrete pipes than for
repair of standard damage on vitrified clay pipes. This is due to the adhesive
properties of the materials used vis-à-vis the particular pipe material.
370 
The necessary process-engineering conditions to permit repair of defective
lateral connections in closed systems basically exist. The methods applied in
case of so-called extreme damage did, however, demonstrate the limits of the
repair methods during the test. Both the qualifications of the technician
performing the work and the preparatory and adjustment work itself (e.g.
cutting, high-pressure cleaning, etc.) have a significant influence on the result of
the repair.

The test findings from the IKT-CPT "Repair methods for lateral connections"
indicated the need for substantial improvements to the repair method. The test
grade GOOD was awarded only once; the methods predominantly received the
grade SATISFACTORY, ADEQUATE and even POOR.
IKT-CPT “Inspection systems for domestic drain/sewer systems”
Fourteen drain and sewer network operators from the cities of Alsdorf, Bergisch
Gladbach, Cologne, Dinslaken, Düsseldorf, Gladbeck, Göttingen, Hilden, Neuss,
Mönchengladbach, Quickborn, Recklinghausen, Warendorf and Würzburg participated
in the IKT-CPT "Inspection systems for domestic drain and sewer systems" [32].
In North Rhine-Westphalia, Section 61a of the Regional Water Regulations [33]
specifies that the owner of a site must have laterals and base lines inspected for leakage
by not later than December 31, 2015. In recent years, the industry has reacted and has
developed special inspection systems for use in domestic drain and sewer networks.
These remarkably small and manoeuvrable cameras are particularly suitable for
inspection of the narrow and highly ramified systems branching off from the main drain
or sewer, or the demarcation chamber/manhole. But what are these systems capable of?
The IKT-CPT answered this question, with detailled examination of six inspection
systems.
[32] Bosseler, B.; Kaltenhäuser, G.: IKT-CPT “Inspection systems for domestic
sewer networks”; IKT - Institute for Underground Infrastructure; Gelsenkirchen,
September 2005; short report in English on www.ikt.de.
[33] Wassergesetz für das Land Nordrhein-Westfalen - Landeswassergesetz - LWG
vom 25. Juni 1995, , zuletzt geändert am 06.12.2007. GV. NRW. S.926 / SGV.
NRW. 77. GV. NRW. S. 708, hier §61a Private Abwasseranlagen.
371 The
GöttingerThe
GöttingerZK Kanalwurm 70/500
ZK Kanalwurm 70/500
with rotary/pivot head
The Aaligator
The
Göttinger
ZK
The Lindauer Schere (mini)
Kanalwurm 50/300 (mini)
Figure 3:
ORION L
The inspection systems tested
On the basis of the results achieved, all the inspection systems tested were awarded the
“GOOD” grade on the score card ultimately drafted. They nonetheless all have their
advantages and disadvantages for use in domestic drain and sewer networks.
The IKT-CPT "Inspection systems for domestic drain and sewer networks" permits the
following basic conclusions:
Comprehensive Quality Assurance: Supplementary products, services and functions for
the inspection system form the basis for Quality Assurance. These provide support for a
high-quality inspection procedure. The requirements formulated in the test with respect
to Quality Assurance were fulfilled practically completely by all system suppliers, with
the result that the “Very good” and “Good” grades were awarded.

Deployment capability OK: To what extent were the systems capable of entering
and inspecting the various system zones? Three systems achieved “Good” for
this question. A further three systems achieved only a “Satisfactory” result,
however. None of the systems was capable of completely inspecting one of the
three domestic sewer networks simulated in the large-scale test facility.

Practically all incidents of damage detected: All the inspection cameras
deployed were able to detect numerous points of damage in the system zones
inspected in the three networks. This signifies that the points of damage were
372 visible on the film images. The units’ degree of detection ranged between
“Good” and “Very good”. None of the cameras recorded all of the points of
damage present in the system zones inspected in each case, however.

Deficient detection quality: The documentation supplied on the inspections
performed disclosed certain deficiencies. The image quality of the inspection
films was, for example, relatively poor in a number of systems. Despite the fact
that the same networks were inspected, differing system maps were generated in
the majority of cases, i.e., divergent pipe routings were documented for one and
the same piping system. Only one system provided “Good” detection qualities:
detection quality was “Satisfactory” in two systems, and only “Adequate” in the
case of three systems.

Practical deployment successful: All inspection systems are suitable in principle
for detection of the condition of site drainage networks. This was confirmed by
observation of system use in the cities of Gelsenkirchen, Göttingen and
Würzburg. The on-site impressions underline the results of system tests
performed at IKT’s large-scale test facility. The differences in practical handling
and use of the inspection systems, including, for example, the physical burdens
imposed on the technicians, became apparent during the in-situ research.
None of the inspection systems completely fulfills the drain/sewer network operators’
quality requirements. The test results illustrate the differences in the various systems
and equipment. A decision in favour of or against any system must therefore take into
account the boundary conditions of the particular inspection task and the required
quality of the inspection results.
IKT-CPT "Tube liners for lateral pipes"
Lateral pipes provide the link between public and private drain and sewer networks. As
an element in domestic waste-water management, this link is increasingly attracting
attention from the drain and sewer industry. According to a survey by the German
Association for Water, Waste-Water and Waste (DWA) conducted in 2004 [34], private
drain and sewer networks are in considerable need of improvement. Maintenance of
lateral pipes is, however, a matter not only for private connection users [33, 35, 36], but
also for public network operators [37].
[34] Berger, C.; Lohaus, J.: Zustand der Kanalisation in Deutschland, Ergebnisse der
DWA-Umfrage, Hennef, 2004.
[35] Strafgesetzbuch (StGB) vom 15. Mai 1871 in der Fassung der Bekanntmachung
vom 13. November 1998, zuletzt geändert durch Art. 1 G am 22. August 2002, hier
§§324 ff.
[36] Gesetz zur Ordnung des Wasserhaushalts (Wasserhaushaltsgesetz - WHG), vom
27. Juli 1957 in der Neufassung der Bekanntmachung vom 12. November 1996.
[37] Verordnung zur Selbstüberwachung von Kanalisationen und Einleitung von
Abwasser aus Kanalisationen im Mischsystem und im Trennsystem (Selbst373 Tube liners are increasingly coming into use for renovation of private drain and sewer
networks. The extent to which these liners meet the requirements made on them is not
totally clear at present, however. Both drain and sewer network operators and private
land owners require information to assist them in the selection and use of suitable tube
liners.
IKT has performed a comparative test of liners for lateral pipes [38], with participation
by the cities of Alsdorf, Bergisch Gladbach, Cologne, Dinslaken, Düsseldorf, Gladbeck,
Göttingen, Hilden, Neuss, Mönchengladbach, Quickborn, Recklinghausen, Warendorf
and Würzburg.
In this test, too, primary emphasis was on the use of the products in experimental
drain/sewer networks with simulated damage scenarios. The task involved was that of
renovating a so-called “standard” and a so-called “extreme” situation. Completion of the
necessary repairs was again followed by testing and exposure to operational loadings.
überwachungsverordnung Kanal - SüwV Kan); Gesetz- und Verordnungsblatt für das
Land NRW, Nr. 49: S. 64- 67; Düsseldorf 1995.
[38] Kaltenhäuser, G.: IKT-CPT “Tube liners for lateral pipes”; IKT - Institute for
Underground Infrastructure; Gelsenkirchen, November 2005; short report in English
on www.ikt.de.
374 Standard situation
Extreme situation
Figure 4:
Renovating task in IKT’s large-scale test facility (length: 18 m,
width: 6 m, depth: 6 m); Simulation of on-site practice on a 1 : 1 scale (laterals on
the second or third of a total of three levels; showing twelve of a total of thirty-six
lines)
Figure 5:
Inversion in the IKT
large-scale test facility
Figure 6:
Measurement of
wrinkling after removal of the liners
The IKT-CPT "Tube Liners for lateral pipes" assesses eight tube liners for the
refurbishment of lateral pipes. It permits the following basic conclusions:

Foil seals the Tube Liner: A comparison of the results from the so-called pipe
train test (acceptance test as per DIN EN 1610) against those from the APS [39]
test show that in many cases, it is the inner foil which seals the liner tight. If this
foil is removed in places – as is normal with the APS test – then tightness is also
lost. This is demonstrated by numerous leaking liner samples.
[39] APS-Prüfrichtlinie erschienen im IKT-eNewsletter „Schlauchliner: Dicht oder
doch nicht dicht?“; September 2004.
375 
Quality fluctuations: All tube liners exhibited fluctuations in liner properties.
These fluctuations were observed both across the circumference of the liners,
e.g. in measuring wall thickness, and also across liner length, e.g., measuring
density. The results of the tightness test in accordance with the APS Code [39]
underline the liner quality fluctuations. In some cases, the spread in the results
even produces apparent test-result contradictions. These fluctuations, for
example, indicated that the BRAWOLINER - FIX achieved better results for the
"extreme" system test situation (grade "VERY GOOD", (1.2)) than for the
"standard" system test situation (grade "GOOD", (1.6)).

Negligible operating-load influence: The loads resulting from high-pressure and
mechanical cleaning (spiral-type machine, with various fittings) applied during
the test indicated no obvious influence on liner quality. The spread in material
properties appears to predominate in influencing the results of the tightness tests.
The loads applied generally only roughened or in some cases damaged the inner
foil. No modifications to the actual material of the liner were observed.

Conflict of aims between operability and tightness: Nearly all tube liners in the
test produced better results for operability than for tightness. The prerequisite for
good operability of the renovated lateral pipe is that the liner must manifest only
slight wrinkles and edges, if any, after the renovation. The liner material must be
correspondingly flexible at bends to permit this. Such flexibility may, however,
be in contradiction to material tightness. This was particularly apparent in the
test, in cases in which the liner suppliers used different tube liners for
refurbishing of the "standard" and "extreme" situations. The "DrainPlusliner"
and "BendiLiner" products used only for refurbishment in the "extreme"
situation exhibited far fewer wrinkles at bends than the "DrainLiner" and
"SoftLiner" used for refurbishing in the "standard" situation, but their poorer
performance in terms of tightness resulted in an overall less satisfactory test
result.

Quality Assurance of preparation: Only one provider implemented convincing
quality assurance, achieving the grade "VERY GOOD" (1.5), 20% of the total
result. Most suppliers submitted incomplete documentation, if any at all. In
some cases, the documentation related to materials other than those used in the
test. Many suppliers did, however, indicate that they were currently engaged in
improving their Quality Assurance arrangements. Three suppliers, for example,
have applied for certification of their tube liners by the Deutsche Institut für
Bautechnik (DIBt).

Practice-oriented installation: The site tests confirmed the impressions gained of
the tube liners during installation at IKT's large-scale test facility. The
procedures applied are also suitable for installation of the tube liners under
practical conditions (confined space, time pressure) and are therefore also
fundamentally suitable for refurbishing of lateral pipes. Random tightness tests
performed on liner samples did, however, again disclose differences in quality
across the length or circumference of the liners. The question of whether it is
376 possible at all to fulfill on-site the tightness criteria stipulated by the drain and
sewer network operators therefore arises in the case of many of these liners.
The test results indicate that the liner suppliers must still achieve many improvements.
Although the test confirmed that the tube liners can fundamentally be used even for
drains and sewers with numerous bends, with resulting refurbishment restoring
operability of the lateral pipe, the majority of the tube liners only rarely fulfilled the
tightness requirements made by the drain and sewer network operators. In addition, the
tests performed revealed considerable fluctuations in liner quality, both across the
circumference and the length of the liners. There are also deficiencies in Quality
Assurance procedures, the majority of which still remain at the preparatory stage. For a
number of producers the CPT results have been a starting point for product innovation.
The currently offered products are tested in a new CPT with test results being expected
for 2010.
The influence of IKT-CPTs on decision-making and on product innovation
The first direct benefit of these tests is the generation of market clarity, providing
customers with information on the quality of the available products. A longer-term
benefit derives from the potentials for improvement disclosed for the individual
products. Conclusion of the tests is followed by corresponding market pressure as a
result of publication of the results. This pressure also means that the product suppliers
will actually use the potentials for improvement of their products.
A number of tests (see [40], [31], [38]) disclosed that many products do not yet fulfill
the drain/sewer- network operators’ requirements adequately or, in some cases, at all. In
many instances, the results of the tests indicated that the product suppliers still have
much to do. These tests, however, not only provided indications of the utilization limits,
but also of the utilization potentials of these products. The result is a gain in investment
safety for drain/sewer- network operators and the fact that the use of these products can
be systematically increased.
In many cases, the industry itself has also taken the IKT-CPTs performed up to now as
constructive criticism and has accelerated improvement of its products (see [41], [42],
[40] Bosseler, B.; Kaltenhäuser, G., Puhl, R.: IKT-CPT “Lateral connections”; IKT Institute for Underground Infrastructure; Gelsenkirchen, June 2002; download in
German on www.ikt.de.
[41] Homann, D.; Kaltenhäuser, G.: IKT-CPT „Flexoset-Anschlusselement B“;
following the test procedure of the IKT-CPT “Lateral connections”; IKT - Institute
for Underground Infrastructure; Gelsenkirchen, June 2003; download in German on
www.ikt.de.
[42] Kaltenhäuser, G.: IKT-CPT “Strobel-Concrete-Procedure; following the test
procedure of the IKT-CPT “Repair methods for lateral connections”; IKT - Institute
377 [43]) and has in some cases actually developed new products (see [44] and [45]). This,
of course, underlines the importance of the IKT-CPTs.
In addition to this information on individual products, the IKT-CPTs also permits
statements concerning the various groups of products tested. The results of the IKTCPT can, therefore, influence rehabilitation and renovation strategies and the operation
strategies of network operators in cases, for example, in which complete product groups
fail to fulfill certain test criteria.
Ultimately, the results of the IKT-CPTs completed up to now confirm the need for the
assessment of the available waste-water technology products and their installation
methods in the context of comparative quality tests:

This will make it possible to select from the large range of alternatives the most
suitable product for the particular project, thus reducing investment risk.

The network operators' requirements must form the basis for development of
products, as improvement potentials are identified and documented in the tests.
IKT-CPTs can thus achieve in a "closed product-improvement circuit", which will
stimulate innovation and an improved market for these products and services.
Conclusion
The IKT, in cooperation with network operators, will in the future continue to test drain
and sewer technology products to the utmost.
The IKT-CPT “Repair Methods for main pipes" is, for example, currently being
conducted. Twenty-six drain/sewer network operators from all parts of Germany, the
Ministry of the Environment, Nature Conservation, Agriculture and Consumer Affairs
for Underground Infrastructure; Gelsenkirchen, November 2004; download in
German on www.ikt.de.
[43] Kaltenhäuser, G.: IKT-CPT „Janssen-Repairing-System with resin“; following
the test procedure of the IKT-CPT “Repair methods for lateral connections”; IKT Institute for Underground Infrastructure; Gelsenkirchen, February 2006; download in
German on www.ikt.de.
[44] Kaltenhäuser, G.: IKT-CPT „Janssen-Repairing-System with mortar“; following
the test procedure of the IKT-CPT “Repair methods for lateral connections”; IKT Institute for Underground Infrastructure; Gelsenkirchen, February 2006; download in
German on www.ikt.de.
[45] Kaltenhäuser, G.: IKT-CPT „Ceramic lateral connection C100 – 150“; following
the test procedure of the IKT- CPT “Lateral connections”; IKT - Institute for
Underground Infrastructure; Gelsenkirchen, September 2006; download in German
on www.ikt.de.
378 of the State of North Rhine-Westphalia, and the Ministry of the Environment of the
State of Baden-Württemberg, are participating in this test. The test focuses this time on
three groups of procedures: Short liners, injection methods and internal sleeves [30].
The IKT-CPT “Tube liners for laterals” Part 2 , has already been started, the results
being expected for 2010. This project involves testing of products improved or newly
developed as a result of the first IKT-CPT “Tube liners for laterals”.
An IKT-CPT “Tube liners for sewer mains" is also planned for the future. This test will
submit the numerous tube liners for main drains and sewers available on the market to a
comprehensive program of testing. Current IKT test results of in situ samples [46]
illustrate the necessity of also submitting tube liners for main pipes to an IKT-CPT.
Please visit the IKT site (www.ikt.de) for more information and downloads.
[46] Waniek, R. W.; Homann, D.: IKT-Linerreport 2006: Glass clearly ahead?,
download on www.ikt.de.
Presentation of Author
Dr.-Ing. Bert Bosseler studied Civil Engineering at Bochum
University, Germany and at the University of Campinas, Sao Paulo.
He was employed with one of the largest water and waste water
management companies in Germany, the Emschergenossenschaft.
In 2000 he was appointed Research Director of IKT – the Institute
for Underground Infrastructure, Germany (www.ikt.de). Moreover,
since 2006, he lectures at Hannover University on sewer and
pipeline construction, maintenance and rehabilitation.
379 VII.2
From Desktop to Plant Room:
Development of an innovative system
for mapping and assessing trap seal
vulnerabilities in building drainage
systems – lessons from the field.
Dr. M.Gormley
Mr. C. Hartley
m.gormley@sbe.hw.ac.uk
School of the Built Environment, Heriot-Watt University, Scotland, EH14 4AS.
Abstract
The use of the reflective wave technique has been shown to be an effective method of
mapping and detecting water trap seal vulnerabilities in building drainage systems.
Research in sanitation engineering must however be about more than theoretically
proving the efficacy of a particular technique, it is an integral part of the process to
implement the introduction of such methodologies in a practical sense, particularly in
our world of rapidly changing needs, in order to show both the importance of such
research and to provide real solutions to serious problems.
Site based validation of the theoretical methodologies is key to the acceptance of any
innovation in building drainage and has been placed at the centre of the development of
the reflected wave technique right from the very beginning. This has proved immensely
effective since site validation of theoretical results not only provides confidence in the
theoretical approaches, but in turn, informs the theoretical advances in a virtuous cycle
of development, validation and review’. This approach to implementing innovative
ideas in building drainage has been carried out in five major studies leading from the
relative simplicity of a university laboratory to a complex hospital environment in
constant use.
380 The introduction of a conceptual framework for the development of this system is
presented here which highlights the benefits of using an iterative and incremental
approach to problem solving complex testing techniques in building drainage systems.
Keywords
Site testing, results validation, confidence building, conceptual models.
1. Introduction
This research charts the process of moving from the theoretical to the practical and
highlights the difficulties along the way. The acceptance of end-users and building
management personnel is of vital importance in this respect and gaining their
confidence provides a significant milestone for the industry as a whole. The integration
of a new technology into a sometimes rather rigid regulatory framework poses another
challenge. Integrating academic rigour with technological development isn’t new,
however there are many pitfalls and finding a conceptual model to cover this process is
an important element in the development process.
There are many methodologies [1] for charting progress and streamlining the innovative
process. The field of software engineering has considered this to be essential in complex
system analysis and design. A brief look at some of the methodologies used in that area
of engineering shows several potentially useful conceptual models for development of a
technology for application to the detection of defective water trap seals since there is a
considerable software component to the system. The brevity of this paper precludes a
lengthy discussion of these methods here however several have been chosen for a brief
introduction as follows;

The Waterfall method – This method is a top down methodology whereby a
design is produced from a system brief which is implemented, tested and
verified. This is a rigid system and is only suitable for a mature technology
where a reasonably accurate brief can be produced.

The Shashimi method – This again is a top down system based on a rigid brief,
however allowance is made for some refinement of the brief during the
development phase. The methodology is still rigid and again suitable for mature
technologies.

Iterative and incremental development – In this approach a cycle of ‘brief’,
design, implementation, testing and review are all performed in a cyclical
manner to fine-tune a design to the required outcome in an iterative and
incremental way.
381 These methods are all features of complex software system design where many
elements coalesce to form an integrated whole. A critique of these systems in the
context of a system with elements of software, hardware and a complex pressure wave
tracking methodology leads to a conclusion that the iterative and incremental
development model is more appropriate to this field of interest, since there are many
unknowns in the field. The ‘top down’ waterfall method requires prior knowledge of the
eventual operation of the system, which is absent from a development such as the
DETIS at its inception. The Shashimi method includes an element of iteration since
there are overlapping elements leading to a process of feedback to the waterfall method,
however it is still a ‘top down’ system. The iterative and incremental provides the best
chance of producing a robust design in an iterative process best suited to applications in
building drainage systems.
2.
A Conceptual framework for development.
A conceptual framework for the evaluation and development of an innovative
technology provides a useful insight into the way in which systems can be developed,
particularly if the initial concept has a root in theoretical engineering in an academic
setting. Figure 1 below illustrates the ‘development cycle’ and shows the elements of
design and review which need to be included if a technology with a useful, practical
application is to be developed. The conceptual framework is a modification of the
development and iterative method discussed above and provides enough feedback and
checks in order to highlight deficiencies and inefficiencies.
Design/ Refine Evaluate/ Simulate
Peer Review Laboratory trials
Site Figure 1
A conceptual framework for design based on theoretical, laboratory
and site investigation, based on a modification to the iterative and
incremental development method as applied to complex software
system design.
382 The elements of the process are obvious in many cases however it is useful to be
reminded that each of these steps is important to produce a robust design. The elements
are described in more detail below.
Design/ Refine
This is the initial design and subsequent refinement. The motivation for the design may
be a concept based on a known need or a response to a perceived need in the field.
Evaluate/ Simulate
Central to the work of academic engineers is the evaluation of system design and
performance. Linked to this, although often separate, is the modelling of system
operation using mathematical, numerical and computer techniques to allow fast
evaluation of conceptual designs and subsequent changes/refinements in realistic
modelling environments. The power of this element of the process lies in the ability to
evaluate without having to instigate costly and time-consuming laboratory or site
evaluations on embryonic designs, the process is thereby speeded up considerably.
Laboratory Trials
Modelling and simulating operation cannot however highlight some of the nuances of
system performance, particularly if the system under consideration is new with limited
data sets of usage information from which to drive the models. Laboratory
investigations highlight the first practical issues in any development process, as well as
providing useful data on actual performance from which the models can be refined
Site Trials
One of the criticisms often aimed at academics is an inability to operate in real
situations, as professionals in industry have to. The instigation of site trials is therefore
very important in the development cycle since it not only provides more practical
application information and additional performance data to drive the simulation models,
but provides the first real critique of the system from professionals in the field. This
critique is invaluable. Professionals, particularly in the building drainage field, have a
sense of what can work, what building managers and authorities will accept and how the
system will be viewed by the profession at large.
Peer Review
The production of research papers for peer reviewed academic journals and relevant
conferences provide an academic integrity to the development process. It is often
considered inappropriate to publish details on innovative technologies because of
intellectual property (IP) issues, however, if these IP issues can be overcome the benefit
of expert opinion and criticism is of immense value to any development process. If
383 handled properly the ‘peer review’ part of the process should feed into the conceptual
framework for design and enhance the overall outcome.
These elements are not exhaustive and indeed there are ‘loops within loops’ within this
process, however overall these are considered the main elements required to produce a
robust technology with every chance of succeeding since the iterative process has the
effect of ‘weeding out’ inefficiencies and promoting best practice in both academic and
industrial spheres of interest.
3. Case Study
3.1 Introduction
In order to illustrate the effectiveness of this conceptual framework for design and
development, a case study will be considered. The system under consideration is the
defective trap identification system (DETIS). The purpose of the discussion is not to
explain how the system works, that has been dealt with elsewhere both at CIBW62
symposia and in peer review journals,[2],[3],[4] but rather to consider how changes,
refinements and improvements have been made possible by application of the
methodology set out in the conceptual framework above. The iterative process of
‘passes’ around the cycle is further enhanced by an increase in complexity in the site
trial phase. It is the site trial element which forms the focus of the case study.
3.2
First iteration – proof of concept
Prior to going on site for the first time the process began with a concept based on well
known and understood pressure wave propagation techniques. The first iteration
involved a truncated version of the framework above involving only the design,
simulation and laboratory investigation elements. This is usual in the very early stages
of design however it is important to highlight the iterative process even at this early
stage. The proof of concept went through several stages of this process using only very
simple laboratory test rigs, increasing in complexity as the extent of the capability of the
DETIS became apparent.
384 Design/ Refine Simulate Laboratory investigation
Figure 2 Initial Proof of Concept
The initial proof of concept was carried out on the laboratory test rig at Heriot-Watt
University (HWU) as shown in Figure 3. The simple ‘single shot’ piston transient
generator was used to prove that any empty water trap seal could be located using the
reflective wave technique. This system was also used in the initial simulations using
AIRNET. This iteration proved that the methodology worked, that any termination
could be detected and that the system produced a unique ‘signature’ which could be
used to effectively assess if any given system had a defect.
Stack AAV
termination
Trap
Trap
Transient Generator 60m of 75 mm dry stack
pipework
Figure 3 Laboratory test rig used for initial ‘proof of concept’ iteration.
3.3
Second iteration – Site trial housing block, Dundee.
The first site trials were carried out in a 14 storey building in Dundee Scotland. The
results of this trial have been discussed in previous CIBW62 papers [4] and on the
whole the trial was very successful. In terms of application of the iterative and
incremental conceptual framework this trial raised several important issues which did
not emerge in the laboratory or simulation phases, however they were verified
385 afterwards by another pass of the simulation/laboratory investigation phase. The main
issues raised were:
1.
The single pulse methodology used successfully in the laboratory may actually
cause water trap seals to be displaced in certain circumstances. This was an
invaluable discovery and one which instigated some research on non-steady
friction in water trap seals in response to high frequency air pressure waves [5]
[6]. This led to the development of the sinusoidal wave technique, a nondestructive method of testing.
2.
While it was well known that waves divided at junctions [7] the implications for
this application weren’t fully appreciated. The dimensions of the input junction
in relation to the main stack height are critical, as is the location of the device
itself.
These issues were addressed in the laboratory, in modelling and the outcomes reported
in conference papers and peer reviewed journals.
Sink
Bath
WC
Stack
To Sewer
Figure 4 Housing Block in Dundee used in first real site trial.
386 3.4
Third iteration – HWU School of the Built Environment building
The lessons learned from the second iteration were implemented in the third iteration
and the second real site test. The site was chosen because access was available for
continuous observation and because it is in a busy campus; these facilities are in
constant use. At face value the system appears simpler than the Dundee test however
there were unforeseen complications with this system which produced valuable insights.
This, again, highlights the importance of the iterative and incremental approach. The
main findings from this research iteration were:
1.
The sinusoidal wave technique was verified as non-destructive, repeatable and
reliable, even in the busy campus building.
2.
It was difficult to ascertain a defect in water traps at the ends of the branches,
again this is due to the division of waves at junctions and the cumulative effect
of this phenomenon. Further investigation and simulation of the system using
AIRNET produced a solution whereby additional transducers were required on
branches in order to produce the required resolution.
3.
This research also highlighted the unusually high rate of evaporative depletion
of water trap seals used to collect condensate from boilers, a cause for concern.
Boiler Room Male Toilet Shower Room
Male Toilet Figure 5
Shower
Room
Site Trial at Heriot Watt University
387 3.5
Fourth iteration – RBS building, Glasgow
Armed with the valuable insights from the HWU trial another building was procured for
a field trial. The Royal Bank of Scotland kindly allowed the Drainage Research Group
from HWU to trial the system in a building in Glasgow. The building was 7 storeys tall
and was a busy commercial office block. The field trials in this building provided more
invaluable insights into the extent to which the technology could be applied. The main
findings were;
1.
The test methodology was robust, repeatable and non-destructive. Tests were
carried out when the building was occupied without disruption to normal
building operation.
2.
It was found that more than one vertical stack could be tested using a single
input point for the air pressure wave. The installation of addition pressure
transducers on every stack was sufficient to locate defects on any given stack.
This expanded the reach of the device to any stack directly connected to a
collection drain.
3.
A phenomenon of apparent time delay on the reflected wave return time was
observed in this installation. The increased range of the identification system
amplified this phenomenon to the point where the previously accepted method
of calculating defect location broke down. Extensive model simulation and
laboratory investigation of this phenomenon highlighted the need for a means of
calibration to overcome the apparent delay phenomenon. This was included in
the system control programme due to Kelly [4] known as ‘Tracer’.
The iterative and incremental methodology employed in this phase of the research was a
powerful means of overcoming considerable challenges in the field.
AAV
AAV
AAV
AAV
AAV
AAV
Plant Floor
Floor
Floor Floor
Floor Floor Ground
Stac
Figure 6
Stack Stack Stack Stac
The installation at the RBS building in Glasgow
388 Stac
3.6
Fifth iteration : Royal Infirmary Edinburgh (RIE)
Overcoming previous challenges has produced a very robust technology defensible in
any sphere of engineering development. The iterative and incremental approach has
proved a powerful methodology and confidence is high that solutions to future
challenges can be found using this method.
The next phase of this research will install a test system in the New Royal Infirmary in
Edinburgh, a prestigious building owned by Consort Healthcare, operated by Balfour
Beatty Workplace and used by the National Health Service (NHS) in Scotland. The
hospital opened in 2003 at a cost of GBP£190 million. It has some 900 inpatient beds
and the biggest maternity and reproductive health department in Scotland. It is also the
teaching hospital linked to the University of Edinburgh. It is the centre for many
specialist services including cardiac surgery, and kidney, liver, pancreas and bone
marrow transplants. The RIE is home to the Scottish Liver Transplant Unit. The hospital
is a centre of medical excellence and the trial of the DETIS has been allowed on the
grounds of its potential for producing a reduction in spread of Healthcare Acquired
Infections (HAI) such as C-Dif and MRSA, an issue high on the agenda of all
healthcare professionals.
Figure 7
Arial view of the new Royal Infirmary Edinburgh. Building
provided by Consort healthcare and Balfour Beatty workplace.
389 The drainage system in the RIE is very complex and extensive. A rough estimate gives
a total of approximately 650 vertical stacks in the whole building. The stacks
themselves are relatively straightforward in design and are easily testable individually
however the scale of the installation and the complexity of the horizontal collection
drain network are significant.
The approach taken in the proposed trial is to isolate a small area of the building for
investigation. The context for the choice of trial area is simply to ‘Protect Patients’
therefore the stacks identified for the trial cover two significant ward areas in the
hospital. The purpose of the trial is to prove that the system can work in a busy hospital
building so a small area will prove this easily. The ‘small area’ is in fact bigger than the
entire installation at the RBS building in Glasgow.
The application of the iterative and incremental methodology to this installation will be
invaluable since the complexity of installation provides many challenges which require
verifiable innovative solutions.
The extent of the challenges to be overcome can not be dealt with in this paper however
the following summarises some of the issues;
1.
The base of stacks are not all accessible.
2.
Many of the collection drains are greater than 150 mm which will cause severe
degradation of our signal and make measurements impossible using the
methodology used in Glasgow.
3.
Access to ward areas must be limited and planned.
Implementing the methodology to this installation produces an elegant solution to the
problems associated with working with the collection drain system. Extensive
simulations and the availability of plant room space produced the proposed installation
methodology shown below in Figure 7.
The advantages of this system are;






All testing and installation done in plant room, no need for costly works to drill
holes in floor slabs and fire stop.
Only one pressure transducer required, therefore fewer inputs are required on
data acquisition system and fewer transducers required.
The installation will be quicker and require less building operator input.
Only one stack is tested at a given time thereby simplifying the test
methodology and identification of traps.
The system is upgradable i.e. any number of stacks can be added without
changing the defect free data base of the whole installation.
The calibration required is reduced since the apparent delay observed in the
RBS building will be reduced.
390 The installation will be challenging however the simulation and laboratory tools
available provide considerable backup in the event of encountering problems on site.
The holistic approach taken to all these installations has proved effective and the
implementation of all the elements described in section 2 above in the conceptual
framework have been required to reach this point in the research successfully.
One Pressure Transient Generator and pressure transducer required Roof level P
75 mm pipe manifold Plant room Wards Floor 2 Wards Floor 1 Ground floor level Basement ( below ground level)
Figure 8
Proposed layout of installation for RIE showing
innovative manifold system.
391 4.
Conclusions
The derivation of solutions to complex problems is the main function of the engineer,
and the engineering academic is no exception. A collegiate approach, through
interactive research group to wider academic community and industrial collaboration is
vital to producing solution to real issues in the real world. The approach to the
development of the DETIS has included all these elements and is success is due to the
commitment of all involved to work together to produce a technology with far reaching
benefits for all.
On a practical level the use of the iterative and incremental conceptual framework
outlined in this paper has been a powerful tool in producing a robust methodology for
assessing the state of the seal in a building drainage system. The water trap seal, for all
its vulnerabilities, is still the main protection against ingress of foul odours, gases and
potentially harmful infectious material. The technology produced by this research has
gone a considerable way in highlighting this significant fact and has raised the issues
with those in healthcare for whom protection against vulnerabilities is more important
than for most.
5.
References
[1]
Larman,C & Basili, V.R "Iterative and Incremental Development: A Brief
History," Computer, vol. 36, no. 6, pp. 47-56, June 2003.
[2]
Kelly DA, Swaffield JA, Campbell DP, Gormley M, Jack LB ‘A transient –
based technique to locate depleted appliance trap seals within building drainage
systems. BHR 10th International conference on Pressure Surges, Edinburgh,
May 2008.
[3]
Kelly DA, Swaffield JA , Jack LB, Campbell DP, Gormley M ‘Pressure
transient identification of depleted appliance trap seals: a pressure pulse
technique.’ Building Services Engineering Research & Technology 29, 2 (2008)
pp. 165 - 181
[4]
Kelly DA (2008) ‘Reducing the risk of infection spread via the building
drainage system through identification of depleted trap seals by temporal
analysis of reflected transients. 34th CIBW62 Symposium on Water Supply and
Drainage for Buildings, Hong Kong
[5]
Swaffield, J, A. (2007) ‘Influence of unsteady friction on trap seal depletion’,
CIBW62 Symposium on Water Supply and Drainage for Buildings, Brno, Czech
Rep.
392 [6]
Beattie, R.K (2007) ‘Derivation of an empirical frequency dependent friction
factor for transient response analysis of water trap seals in building drainage
systems’ Unpublished MSc thesis Heriot Watt University, Edinburgh.
[7]
Swaffield, J.A. & Boldy A.P. (1993) Pressure Surge in Pipe and Duct Systems,
Ashgate, UK.
6.
Presentation of Authors
Dr. Michael Gormley is a lecturer in Architectural Engineering
and member of the Drainage Research Group at Heriot - Watt
University. His research interests are pressure transient
modelling and suppression in drainage systems and solid
transport in above ground drainage systems.
Charles Hartley is currently a member of the Drainage Research Group at Heriot -Watt
University. He is a Research Associate working exclusively on the commercialisation of
DETIS. Charles is a Radar and electronics specialist with extensive experience gained
in the Royal Navy for 12 years.
393 VII.3
Reducing the risk of cross-contamination
from the building drainage system using the
reflected wave technique to identify depleted
trap seals
D.A. Kelly
d.a.kelly@hw.ac.uk
School of the Built Environment, Heriot-Watt University, Edinburgh, United Kingdom.
Abstract
The appliance trap seal remains the primary defence against cross-contamination from
the foul air present within the building drainage system. Prevention of trap seal failure
depends upon both good design, to limit the air pressure transients propagated within
the system, and good maintenance. However, current maintenance regimes rely solely
on visual inspections to detect trap seal failure which is time consuming and often
impractical to implement in large complex buildings.
This paper documents the development of a novel approach to system maintenance. By
employing the well-known mechanisms governing pressure transient propagation,
applicable to any fluid carrying system, the reflected wave technique offers a systematic
and non-destructive method of remotely monitoring trap seal status by measuring the
system response to an applied low-amplitude sinusoidal pressure wave. The reflection
induced by the depleted trap seal is shown to sufficiently alter the system response such
that, when compared to a previously obtained defect free baseline, its detection and
location can be identified.
Automatic system diagnosis is provided by the
trap condition evaluator (TRACER) program which incorporates a time series change
detection algorithm to automatically calculate the return time of the trap induced
reflection. The TRACER program is shown to accurately pinpoint the location of a
depleted trap seal provided that the dampening influence of the junction effect (which
can delay the observed reflection return time) is taken into account.
An extensive series of laboratory experiments and field trials were carried out to collect
transient pressure data which, together with results from an existing mathematical
model (AIRNET) capable of simulating real system operation, have been used to test
394 and validate the proposed technique and to formulate a practical methodology which
may be applied to any building drainage system.
Keywords
Building drainage system, SARS, cross-contamination, trap seal depletion, transients
1
Introduction
The appliance trap seal remains the primary defence against cross-contamination from
the foul air present within the building drainage system. Once only considered to be a
nuisance due to its unpleasant odour, it is now well understood that the foul air within
the drainage network can facilitate the spread of infection and disease by mediating the
movement of pathogenic microorganisms within airborne aerosols (Gebra et al., 1975).
A great number of viruses (i.e. adenoviruses, astrovirus, enteroviruses, hepatovirus,
norovirus, reoviruses and rotavirus) and bacteria (i.e. Escherichia coli (E. coli),
Legionella pneumophila, Salmonella and Shigella) passed in the excreta of infected
individuals have been found not only to exist within the building drainage system but
are also amenable to airborne transmission within aerosolised water particles (Feachem
et al., 1983). Trap seal depletion provides a potential route for these harmful pathogens
to be transmitted from the drainage system into the building.
The consequence of trap seal depletion gained international attention in 2003 when it
was identified by the World Health Organisation as a transmission vector in the spread
of severe acute respiratory syndrome (SARS) at the Amoy Gardens housing complex in
Hong Kong. Home to approximately 20,000 residents and consisting of 19 tower
blocks, the outbreak of SARS at Amoy Gardens saw a total of 321 reported cases of
infection (almost 20% of all cases in Hong Kong), resulting in 42 deaths (WHO, 2003).
It was discovered that the virus was spread in airborne aerosols which entered the
building via depleted trap seals. Had a proper maintenance regime been in place at
Amoy Gardens then it is reasonable to assume that this community outbreak could have
been avoided or, at least, could have been considerably reduced.
The aim of the research presented in this paper is to develop and validate a systematic
method of remotely monitoring trap seal status as a regular and routine preventative tool
against the transmission of infection and disease from the building drainage system into
habitable space. Provision of such a methodology would thereby provide an automatic
and periodic system maintenance regime.
395 2
Identification of depleted trap seals
The severity of the SARS outbreak at Amoy Gardens highlights the importance of
adequate system maintenance so that, when trap seal depletion does occur, it can be
quickly identified and rectified to ensure that cross-contamination is minimised. There
is, therefore, a continuing and urgent need to develop a novel technique which offers a
holistic solution to system maintenance allowing the condition of all trap seals to be
determined quickly and easily.
2.1
Current approach
The current approach to system maintenance is inadequate as it relies solely on visual
inspection of the trap seal which is not only labour-intensive but also impractical and
effectively impossible to sustain in large buildings.
2.2
Proposed method
The reflected wave technique is based on the time analysis of the first reflected wave
induced by a depleted trap seal in response to an applied control transient. The location
of the depleted trap seal is determined by the time taken for the applied transient to
propagate to the depleted trap and for its reflection to be returned. The reflected wave
technique offers, for the first time, a systematic method of remotely monitoring trap seal
status as a regular and routine preventative tool against the transmission of infection and
disease from the building drainage system. Provision of such a methodology would
thereby provide an automatic and periodic system maintenance regime.
3
The reflected wave technique
The proposed method is fundamentally based on the mechanisms of pressure transient
propagation and, in particular, the reflection and transmission of pressure waves at
system boundaries, which are common to all full-bore fluid carrying conduits.
Pressure transients normally occur within the drainage system following rapid changes
in flow conditions as a result of appliance discharge. An instantaneous change in flow
conditions, V, generates a pressure transient, p, given by the Joukowsky formula:
 p    c V
(1)
where  is the fluid density, and c is the wave propagation speed in the system fluid – in
this case air - which is defined as:
c
p
(2)

where p is the absolute pressure and  is the ratio of specific heat (Swaffield and Boldy,
1993). The significance of the negative sign in Equation (1) means that pressure
increases with a decrease of velocity, and vice versa.
396 These pressure transients are generally considered a problem within the building
drainage system because they are themselves a potential cause of trap seal depletion.
However, when a pressure transient propagates though the drainage system, every
system boundary it encounters will induce a characteristic reflection, Table 1, which
alters the shape of the system transient response. Important information about the
features of the system boundaries is, therefore, carried within the system response
allowing their identification within the system. From Table 1 it can be seen that the
reflection induced by a trap seal is dependant upon the status of the trap. A fully primed
trap seal will generate a +1 reflection which, in response to a positive pressure transient,
appears as a pressure rise (having the same magnitude and sign as the incident
transient) in the system response, while a depleted trap seal will generate -1 reflection
which appears as a pressure drop (having the same magnitude, but opposite sign, as the
incident transient).
Table 1 Reflection and transmission coefficients for some of the most common
system boundaries found in the building drainage system
Boundary
condition
Influence
on
transient
Reflection
incident coefficient,
CR
Closed-ended
Generates a positive
pipe (i.e. an reflection which is
AAV or fully equal in magnitude CR = +1
primed
trap and sign to the
seal).
incident transient.
Open-ended
pipe (i.e. an
open
stack
termination or
a depleted trap
seal).
Generates a negative
reflection which is
equal in magnitude CR = -1
but opposite in sign to
the incident transient.
Pipe junction
The incident transient
is both reflected and
transmitted,
the
proportion of which is
determined by the
area ratio of the
connecting pipes.
A1 A2

 ... 
c
c2
CR  1
A1 A2

 ... 
c1 c2
Transmission
coefficient,
CT
CT = 0
CT = 0
An
cn
An
cn
CT 
2 A1
c1
A
A1 A2

 ...  n
c1 c2
cn
The reflected wave technique uses this change in reflection coefficient to detect and
locate a depleted trap seal, the presence of which is indicated by the arrival of a negative
reflection in the system response. The location of the depleted trap can be determined
from the expression (Swaffield, 2005):
XD 
c tD
2
(3)
397 where X D is the distance to the depleted trap seal from the measurement point, t D is
the pipe period of the depleted trap (i.e. the time taken for the pressure transient to travel
from the monitoring point to the trap and for the induced reflection to be returned to the
measurement point), and c is the wave propagation speed in air.
Precision in locating the depleted trap seal strictly depends on the accurate evaluation of
the arrival times of both the incident transient wave and the subsequent trap induced
reflection at the measurement point. Due to the complexity of normal building drainage
systems and the consequently large number of different system boundaries contained
within them, the number of reflections and re-reflections that occur within the system in
response to a propagated transient may be high. The successful operation of the
reflected wave technique, therefore, relies on the ability to distinguish the reflection
returned from a depleted trap seal from the reflections generated by the normal features
that exist within the system (e.g. junctions and open terminations). This is achieved by
obtaining a system response under normal defect free conditions – which will include
the effect of all reflections induced by normal system boundaries – to be used as a
comparison to subsequent system test responses. A change in trap seal status would
cause a deviation from the defect free baseline at a time equal to tD which can then be
used to derive the location of the trap, XD, from Equation 3.
Development of the reflected wave technique (Swaffield, 2005; Kelly et al., 2006;
2008a; 2008b; 2008c; Kelly, 2007; 2008; 2009) has focused on establishing a
monitoring method which is accurate and reliable, yet uncomplicated and non-invasive.
This research has developed a suitable transient entry device which uses a 10Hz
sinusoidal pressure transient which ensures the test does not pose a problem to system
integrity as this type of transient has previously been shown to be non-destructive to
connected trap seals (Kelly, 2007).
4
Automatic detection and location of a depleted trap seal
To provide the necessary level of protection against cross-contamination it is essential
that the building drainage system is monitored regularly and systematically. This can
only be achieved by providing automatic test scheduling and data analysis which is not
reliant on user intervention or interrogation.
This required the development of the TRACER program (a time series change detection
indicator) which automatically analyses the system test response and compares it with
the defect free baseline to determine trap status. The presence of a depleted trap is
automatically detected and its location identified by calculating the absolute difference
between the defect free baseline and the system test response.
4.1
4.1.1
Automatic determination of trap status
Test set-up
Before the TRACER program can be used for the automatic detection and location of
depleted trap seals, some important system information is first required.
398 As discussed in Section 3, a defect free system response, to which all subsequent system
test responses can be compared, must first be obtained. Ensuring that all trap seals are
fully primed, the system response to the applied incident transient is recorded at one or
more measurement points within the system. A total of N defect free system responses
are recorded from which the average defect free baseline, PjDF , is calculated:
Pj ,1  Pj ,2  ...  Pj ,N
1 N
(4)
Pj ,i 

N i 1
N
where j is the measurement point. To ensure that all system responses are exactly
aligned before calculating PjDF , the transient wave arrival time, t a , is set to time zero,
PjDF 
Pressure Pressure t 0 , see Figure 1. The method adopted here uses a sinusoidal transient with an initial
positive wave and so ta is taken as the time when the pressure first rises above
atmospheric pressure. If a sinusoidal transient with an initial negative wave was used,
then ta would be taken as the time when the pressure first falls below atmospheric
pressure. To allow direct comparison with the defect free baseline ta is set to t0 for all
subsequent system test responses.
80
0
80
t0 0.0
0
ta 0.1
0.2
0.3
time -80
t0 0.0
-80
0.1
0.2
0.3
time
When Pj first > atmospheric pressure, set ta = t0
Figure 3 The transient wave arrival time for the system response recorded at the
measurment point is set to zero
To allow discrimination between natural variations in the measured system test response
(e.g. due to signal noise), a system threshold value, h, is determined which provides the
expected signal variation for the defect free system. The threshold level is set as the
maximum discrepancy observed between the defect free baseline and each of the N
defect free system responses used to derive it:
h  max PjDF  Pj ,i
i N
(5)
i 1
All values below h are ignored while those that exceed this level are indicative of a
variation from the normal defect free baseline.
The final set of required information is the pipe period, tP, of each connected trap seal.
This can be obtained from the system geometry by measuring the distance to each trap
from the measurement point, or by using the reflected wave technique as a probe to
determine the reflection return time for each trap by systematically removing the water
seal one trap at a time.
399 Figure 2 presents a summary of the necessary steps for the collection and assimilation
of the system information required by the TRACER program.
Record N defect free pressure responses Set ta = t0 when Pj first > atmospheric pressure Calculate defect free baseline, PjDF
, using Equation 4 Calculate threshold value, h, using Equation 5
Establish pipe period for each connected trap seal Figure 4 System information required by the TRACER program for the automatic
detection and location of depleted trap seals
4.1.2
Test procedure
Having acquired the required system information, the TRACER program is now ready
to perform automatic detection and location of depleted trap seals. To determine if a
depleted trap induced reflection has occurred, and to estimate its time of arrival, a time
series change detection test is carried out by comparing the measured system test
response , PjM , with the defect free baseline, PjDF . The absolute difference between
PjDF and PjM is used to determine the goodness-of-fit between the two system
responses.
(6)
D t  P jDF  P jM
where Dt is the absolute difference at a time, t. For every sample of data, the change in
system response is then analysed. Starting at time t0, if Dt exceeds the threshold value,
h, then the discrepancy between the two traces is outwith the expected variation for a
defect free system response, the alarm is issued and the first time of change tD is
recorded as the reflection return time.
The TRACER program also includes a simple calibration test which is designed to
distinguish between a reliable system response and an unreliable system response. The
calibration test analyses Dt over the first section of the system response in which no trap
induced reflection would occur. If Dt > h during the calibration period then the system
response is considered unreliable and the system is retested. Otherwise, if Dt < h
during the calibration period then the system response is considered to be reliable and
the data analysis proceeds.
Figure 3 shows an example output from the TRACER program. A change from the
defect free baseline has been detected at 0.066 seconds. Automatic cross reference with
the known trap pipe periods has identified the cause as depleted trap T12.
400 Pressure (mm water gauge)
-100
0
0.0
0.1
0.2
0.3
Dt > h
1
100
Defect free baseline System test response -1
Time (seconds)
Is Dt > h over calibration period?
Is Dt > h during test period?
Depleted trap location?
NO, trace is reliable.
YES, at tD = 0.066 seconds.
T12.
Figure 5 An example output from the TRACER program recorded during the
Glasgow site tests
5
Validation of the reflected wave technique
To assess the practical application of the technique and accuracy of the TRACER
program, data were collected from three sets of field trials. The first field trial was
carried out in an unoccupied residential building in Dundee just prior to it being
demolished; the second field trial was conducted within the Arrol building at HeriotWatt University; and the final field trial was performed in an office building in
Glasgow. These tests will be referred to as the Dundee, Arrol and Glasgow field trials.
In each case, the test set-up procedure outlined in Section 4 was followed in order to
obtain the relevant system information required by the TRACER program. Figure 4
compares the predicted depleted trap location with the true depleted trap location.
401 60
Dundee
Arrol
Glasgow
50
Junctions with highest branch to stack area ratio Increasing number of junctions 30
X D
predicted (m)
40
20
Increasing number of junctions 10
0
0
10
20
30
40
50
60
true
D (m) X
Figure 6 Comparison of the true and predicted depleted trap locations recorded
during the Dundee, Arrol and Glasgow field trials
It can be seen that in most cases, the distance to the predicted depleted trap location is
greater than that for the true trap location and that this error increases with increasing
distance to the trap. This overestimation of trap location was found to be caused by a
delay in the observed return time of the trap induced reflection. Further inspection of
the data showed that this observed delay was related to the number and type of junctions
traversed by the propagating transient. As the number of junctions traversed increased,
and as the branch to stack area ratio, Abranch/Astack, increased, the observed delay was
also found to increase. This is clearly demonstrated in Figure 4.
5.1
Quantifying the junction effect
This junction effect is caused by the transient reflection and transmission process which
takes place at each junction. When the propagating transient arrives at a junction, part
of it will be reflected back along the index pipe while the remainder will be transmitted
forwards equally into the receiving pipes. The reflection and transmission coefficient
equations, Table 1, determine what proportion of the transient is reflected and what
proportion is transmitted based on the area ratios of the connecting pipes. The overall
result is that the propagating transient is separated into many smaller reflected and
transmitted parts which effectively “erodes” the incident transient as it propagates
further into the system. The observed delay in the return time of the trap induced
reflection occurs when the magnitude of the transient leading edge is reduced to such a
point that it fails to make an immediate impact on the measured system response.
402 The magnitude of the transient’s leading edge, pl, is thus proportional to the number of
junctions traversed, n, and the junction transmission coefficient, CT , which itself is
dependant upon the branch to stack area ratio:

 p l   p I  CT  CT  ...  CT
1
2
n

where pI is the initial incident pressure transient.
identical transmission coefficients, this becomes:

 p l   p I  C Tjunction
For identical junctions, with

n
For a three pipe junction consisting of identical pipes (e.g. all 100 mm diameter) the
reflection coefficient, CR, is -0.33 and the transmission coefficient, CT, is +0.67.
Therefore, if two junctions exist between the measurement point and the depleted trap
then the leading edge magnitude of the returning reflection, having passed four
junctions in total, will be 20% of the incident transient, while if ten junctions exist
between the measurement point and the depleted trap then, having passed twenty
junctions in total, the returning reflection will be less than 2% of the incident transient.
The junction effect is reduced when the branch to stack area ratio is reduced. For a
three pipe junction of unequal pipes (e.g. when the stack is 100 mm diameter and the
side branch is 32 mm diameter) then CR is now -0.05 and CT is +0.95. Therefore, for
the two junction example, passing four junctions in total, the returning reflection will be
slightly higher than 80% of the incident transient, while for the ten junction example,
passing twenty junctions in total, the returning reflection will be around 60% of the
incident transient.
5.2
Evaluation of the reflected wave technique once the junction effect is
included
Up to now, the reflected wave technique has relied on Equation 3 to predict the
reflection return time, t D . However, this takes no account of the junction effect and this
has shown to result in an overestimation of the true depleted trap location. Two
alternative approaches to predicting t D have been investigated:

Method A uses the AIRNET model to predict the simulated trap pipe period
t DAIRNET from the system response of the simulated system. The simulated trap
distance, X DAIRNET , and the base value of the wave propagation speed, cbase, are
known as they are both user defined (i.e. cbase = 343 m/s).

Method B uses the reflected wave technique as a probe to determine the
perceived trap pipe period t Dprobe . The method requires all depleted trap seal
cases to be sequentially simulated experimentally. Neither X Dprobe or c need to
be known prior to using this method as t Dprobe is measured directly from the
physical system.
403 Figure 5 compares the true and predicted depleted trap locations, shown previously in
Figure 4, now that the junction effect has been taken into account when predicting t D .
Both the AIRNET and PROBE methods show excellent correlation between the true
and predicted trap locations. All traps have been successfully detected and located to
within 7% of their true location.
60
50
predicted (m)
40
Dundee (AIRNET)
Arrol (AIRNET)
Glasgow (AIRNET)
Dundee (PROBE)
Arrol (PROBE)
Glasgow (PROBE)
X D
30
20
10
0
0
10
20
30
40
X
50
60
true
D (m) Figure 7 Comparison of the true and predicted depleted trap locations recorded
during the Dundee, Arrol and Glasgow field trials when using AIRNET and the
PROBE method to predict tD
6
Conclusion
This paper has provided an overview of an innovative method of systematically
monitoring the building drainage system against the threat of cross-contamination of
disease by remotely monitoring trap seal status. This was approached through the
development of the reflected wave technique as a tool for detecting and locating
depleted trap seals and by the collection of transient data sets for the testing and
validation of the technique.
A trap condition evaluator (TRACER) program was developed and implemented into
the data collection package to carry out automatic detection and location of depleted
trap seals. The TRACER program used a time series change detection algorithm to
estimate the arrival of a depleted trap induced reflection by computing the absolute
difference between the observed test data and a previously obtained defect free baseline.
404 The junction effect has been shown to influence the observed return time of the trap
induced reflection. However, by making allowance for the junction effect in the
prediction of reflection return time, it is shown that all depleted trap seals tested during
these extensive field trials can be pinpointed with excellent accuracy.
The reflected wave technique, used in conjunction with the TRACER program, offers
for the first time a practical and effective approach to maintaining and monitoring the
building drainage system against the threat of cross-contamination. This body of
research will continue and the reflected wave technique will be developed into a fully
integrated and automated maintenance regime. Further research is aimed at extending
the technique for application within hospital buildings to help prevent the secondary
spread of infection (Gormley, 2009).
7
References
Feachem, R.G., Bradley, D.G., Garelick, H. and Mara, D.D. (1983). "Sanitation and
disease: Health aspects of excreta and wastewater management," Wiley and Sons.
Gebra, C.P., Wallis, C. and Melnick, J.L. (1975). "Microbiological hazards of
household toilets: droplet production and the fate of residual organisms." Applied
Environmental Microbiology, 30, 229-237.
Gormley, M. (2009). “From desktop to plant room: developments of an innovative
system for mapping and assessing trap seal vulnerabilities in building drainage systems
– lessons from the field.” Proc. from the CIBW62 35th International Symposium on
Water Supply and Drainage for Buildings, Dusseldorf, Germany, 7-9 September 2009.
Kelly D.A. (2007). “Identification of depleted appliance trap seals within the building
drainage and ventilation system – A transient based technique.” Proceedings from the
CIBW62 33rd International Symposium on Water Supply and Drainage for Buildings,
Brno, Czech Republic, 19-21 September 2007.
Kelly D.A. (2008). “Reducing the risk of infection spread from the building drainage
system through identification of depleted appliance trap seals using the reflected wave
technique.” Proceedings from the CIBW62 34th International Symposium on Water
Supply and Drainage for Buildings, Hong Kong, China, 8-10 September 2008.
Kelly D.A. (2009). “Controlling the risk of cross-contamination from the building
drainage system using the reflected wave technique to identify depleted water trap
seals.” PhD Thesis, Heriot-Watt University, Edinburgh, UK.
405 Kelly D.A., Swaffield J.A., Jack L.B., Campbell D.P. and Gormley M. (2006).
“Transient defect identification in building drainage and vent systems – an application
of pressure transient theory and practice.” Proceedings from the CIBW62 32nd
International Symposium on Water Supply and Drainage for Buildings, Taipei, Taiwan,
17-20 September 2006.
Kelly D.A., Swaffield J.A., Campbell D.P., Gormley M. and Jack L.B. (2008a).
“A transient-based technique to locate depleted appliance trap seals within the building
drainage system”, 10th International Conference on Pressure Surges, BHR Group Ltd.,
Edinburgh, UK, 14-16 May 2008.
Kelly D.A., Swaffield J.A., Jack L.B., Campbell D.P. and Gormley M. (2008b).
“Pressure transient identification of depleted appliance trap seals: A pressure pulse
technique.” Building Services Engineering Research & Technology, 29 (2), 165-181.
Kelly D.A., Swaffield J.A., Jack L.B., Campbell D.P. and Gormley M. (2008c).
“Pressure transient identification of depleted appliance trap seals: A sinusoidal wave
technique.” Building Services Engineering Research & Technology, 29 (3), 219-232.
Swaffield, J.A. (2005). "Transient identification of defective trap seals," Building
Research and Information, 33(3), 245-256.
Swaffield, J.A. and Boldy, A.P. (1993). Pressure surge in pie and duct systems. Ashgate
Publishing Limited, England.
World Health Organisation. (2003a). Final Report: Amoy Gardens, WHO
Environmental Investigation.
8
Presentation of Author
David is a Research Associate in the Drainage Research Group at
Heriot-Watt University in Edinburgh. His research interests include
the measurement and simulation of air pressure transient propagation
within the building drainage system and the control of crosscontamination caused by trap seal depletion.
406 VII.4
Test Method of Trap Performance for
Induced Siphonage
K. Sakaue (1), H. Kuriyama (2), H. Iizuka (3), M. Kamata (4)
(1) sakaue@ isc.meiji.ac.jp
School of Science and Technology, Meiji University, Japan
(2)kuriyama.hana@nikken.co.jp
Nikken Sekkei Ltd., Japan
(3) iizukah@nikken.co.jp
Nikken Sekkei Ltd., Japan
(4) ft101743@kanagawa-u.ac.jp
Faculty of Engineering, Kanagawa University, Japan
Abstract
The drainage system is designed in such a way that drain traps are not prone to seal
break. Seal break is caused by seal loss phenomena, and induced siphonage plays an
important role. A reliable drainage system cannot be designed without first defining
trap performance for induced siphonage. However, no effective methods have been
established to test the performance of traps.
Based on the premise that seal loss phenomena are produced as response of seal water
to pneumatic pressure in drain, we conducted free vibration test, single sine wave
response test, and √2 times natural frequency response test using a pressure generating
device with two types each of fixture traps, floor drain traps, and WC to elucidate their
characteristics and strengths. The results clearly indicated that methods defining
minimum pressure required to cause seal break was most appropriate for testing traps
performance to withstand pneumatic pressure. It was also found that natural
407 frequencies obtained by the free vibration test, and max. response magnifications by the
single sine wave response test were relevant and applicable.
Keywords
Drainage system, trap, induced siphonage, test method, single sine wave
1. Introduction
The first consideration in designing any drainage system including vent system is to
make it in such a way that it does not cause seal break. So-called seal break phenomena
include such conditions as induced siphonage, self-siphonage, and evaporation. Of
these, however, one that is pertinent to the entire system would be induced siphonage,
which is characterized by seal loss caused by pneumatic pressure (referred to as
pressure below) affecting the trap seal. In other words, induced siphonage is a dynamic
vibrational response phenomenon caused by the flow of water in the drainage pipe.
To prevent seal break caused by this induced siphonage, it is necessary to establish a
way to predict pressure resulting from discharge load and entire system configuration
such as piping, pipe diameter, type of vent system, as well as to clarify anti-pressure
performance of trap (referred to as trap performance below). Theoretically it can be
achieved by establishing permissible pressure based on trap performance, and by
defining permissible flow rate against it for each system type. The more clearly we
define the performance of each trap, the more flexibly we can design the section of
piping in which the trap is placed.
In the current designing scheme, trap performance is only defined by minimum seal
depth with no pressure prediction made, and at the very best rough permissible flow rate
for each system type is defined. Of all plumbing codes in the world, only the former
DIN1986 in Germany and SHASE-S-218 in Japan stipulate permissible seal loss. As
for permissible residual seal depth, only SHASE-S-218 has stipulations.
In the ever-evolving world of plumbing system, designing of drainage system,
extremely ill-defined compared to that of water supply system and hot water supply
system, has been unable to respond to the trend of rationalization. Though the selfsealing traps without requiring seal water have been developed in Great Britain, water
seal trap is still the mainstream in the world. It seems imperative that we clearly define
trap performance to improve designing methods of drainage system.
Based on the above considerations, the authors have tested various methods of defining
trap performance [1-4]. In this study, as continuation of the ongoing research, the
authors have examined testing methods using a simple device generating single sine
waves with two types each of fixture traps, floor drain traps, and WC.
408 2. Experimental apparatus and measuring equipment
2.1 Experimental apparatus
The experimental apparatus used in this study consists of a pressure generator (pistons,
a frequency variable device, various chambers), drainage pipes, and an analyzer (Figure
1). It has a triple-piston structure with cranks adjustable at 8 steps between 15 ~ 50mm
in increments of 5mm. Frequencies are variable in increments of 0.033Hz within the
range of 0.166 ~ 4.5Hz. It is also equipped with a small blower as a bias pressure
device, which reproduces the steady pressure component. Amplitudes were adjusted by
changing the water contents of vertical cylindrical chamber and pressure adjustment
chambers, which in turn changed air volume in the chambers. Only one of the three
pistons on the apparatus was operated this experiment.
2.2 Test traps
Basic configuration parameters and cross sectional shapes of test traps are shown in
Table 1 and Figure 2. The P trap and S trap for fixtures, the bell trap and contrary bell
trap for floor drains are made of transparent resin, and seal conditions inside can be
observed from outside (Figure 2). Actual models of two types of WC, supper water
saving WC (WC-6L) and water saving WC (WC-8L) were used.
409 2.3 Measuring equipment
We used a diffuse semiconductor type pressure sensor for measuring pressure, and a
capacitive water level sensor for seal water level. The water level sensor had been
modified so that it could be inserted into the inlet leg of a trap. Both of them had
adequate response frequency of more than about 10kHz.
9 8 7 12
1
13
2
11 10 3
6 4
3
4
3
4
5
1; Changing device of frequency
device
3; Chamber with piston
chamber
2; Blower for bias pressure
4; Pressure control valve
5; Mixing
6; Pressure control chamber 7; Test trap 8; Water level sensor
9; Amplifier 10; Pressure sensor 11; Amplifier 12; Data roger
13; PC
Figure 1 - Test apparatus
410 Table 1 – Basic parameters of test raps
Volume of seal
Seal depth
Ratio of leg’s
sectional
area
1)
F [-]
P trap
150
60
1.00
S trap
150
60
1.00
550
53
1.41
410
50
1.26
WC-6L
2,470
55
-2)
WC-8L
1,370
67
-2)
Test traps
For
Fixtures
Bell trap
For floor
Contrary
drains
trap
WC
bell
Note 1) F = (mean sectional area of inlet leg) / (mean sectional area of outlet leg) of a
trap
2) These weren't confirmed because an internal form was complex.
31 92
112
30 40 30
P trap
30
40
30
S trap
Bell trap
14 50 53 10 60 60 88
45
33 43 Contrary bell trap
Figure 2 - Test traps
3. Free vibration test
3.1 Purpose
Natural frequency and damping ratio are the two main basic response characteristics in
forced vibration phenomena. Having damping ratio of 0.4 ~ 0.6, seal fluctuation is
known to cause weak damping oscillation [5, 6]. Considering the effects on trap
performance, we used only natural frequencies as a parameter. Though there are several
411 methods of measuring natural frequencies, we adopted a method based on power
spectrum of response of water level in this study.
3.2 Experimental method
We first caused half-full seal to free-vibrate, and measured free-vibration wave patterns
of seal. Then we obtained the density distribution of power spectrum of response of
water level by giving FFT treatment to the wave patterns. The frequency at which the
power spectrum density became maximum value was determined to be the natural
frequency f0.
3.3 Results and discussion
As shown in Figure 3, power spectrum density distributions of each test trap were as
follows:
P trap: 1.95Hz, S trap: 1.93Hz, Bell trap: 2.43Hz, Contrary bell trap: 2.57Hz.
These values corresponded with the values obtained from the calculate equation of
natural frequency. Although the density distribution of WC generally forms two peaks
as that of WC-6L [7], the second predominant frequency of WC-8L was extremely low
forming only one peak. If the most predominant frequency is regarded as natural
frequency, the peaks of density distribution for WC-6L and WC-8L were 1.61Hz and
1.46Hz respectively. From this, the range of natural frequencies when test traps were
half-full was found to be to be 1.4 ~ 2.6Hz.
P trap Bell trap WC‐6L F : 1 95 Hz
F : 2 43 Hz
F : 1 61
2
9
6
3
6
Power spectrum 12
0
0.0
2.0
4.0
6.0
0.0
2.0
4.0
6.0
0.0
2.0
4.0
6.0
S trap 9
6
2
F
3
6
Power spectrum 12
Contrary Bell trap WC‐8L F : 2 43 Hz
F : 1 46
1 93 H
7
0
0.0
2.0
4.0
Frequency [Hz] 6.0
0.0
2.0
4.0
Frequency [Hz] 6.0
0.0
2.0
4.0
Frequency [Hz] Figure 3 - Power spectrum of water level and natural frequency f0
412 6.0
4. Single sine wave response test
4.1 Purpose
This experiment was intended to define trap performance by obtaining response
characteristics (maximum response magnification) when frequencies of sine wave
pneumatic pressure were fluctuated with its amplitudes kept constant.
4.2 Experimental method
Single sine waves mainly consisting of atmospheric pressure with various frequencies
and constant amplitude were applied to half-full seal. Then the response magnification
curve as a function of frequency was obtained by based on calculation of the static seal
level (X) of trap seal and the ratios of response fluctuations (Xt) against single sine
wave pressure (response magnification M=Xt/X) at each frequency were obtained.
4.3 Results and discussion
Response magnification curves for each test trap are shown in Figure 4. Response
magnification curves corresponded with density distribution curves in the free vibration
test, and frequencies at which response magnification M reached maximum roughly
corresponded with natural frequency f0. However, with WC-8L, response magnification
comparable to natural frequency f0 was noted at the second predominant frequency
(2.4Hz)
M
Bell trap P trap ifi ti
6
5
4
3
2
1
0
M = 4.8 (1.9 Hz) 2.0
4.0
6.0
M
0.0
6
5
4
3
2
1
0
S trap ifi ti
Response
Response
Vibration characteristics of test trap seal seemed to manifest more evidently in the
single sine wave response experiment than in the free vibration experiment.
Experimental method is simpler with only single sine waves of constant amplitudes
used. As a result an experimental apparatus is not required to generate a wide range of
pressure load, and the experimental apparatus itself can be reduced to small size.
M = 5.0 (2.0 Hz) 0.0
2.0
4.0
Frequency [Hz] 0.0
2.0
4.0
WC‐6L 6.0 0.0
2.0
0.0
2.0
4.0
Frequency [Hz] 6.0 0.0
2.0
4.0
Frequency [Hz] Figure 4 - Response magnification M of test raps
413 6.0
WC‐8L Contrary bell trap 6.0
4.0
6.0
Vibration characteristics of test trap seal seemed to manifest more evidently in the
single sine wave response experiment than in the free vibration experiment.
Experimental method is simpler with only single sine waves of constant amplitudes
used. As a result an experimental apparatus is not required to generate a wide range of
pressure load, and the experimental apparatus itself can be reduced to small size.
5. Seal break test
5.1 Purpose
This experiment is intended to define trap performance by measuring amplitudes at the
time of instantaneous seal break when frequencies and amplitudes of single sine waves
are changed. Instantaneous seal break refers to the conditions when air bubbles pass
through top dip of trap.
5.2 Experimental method
Single sine waves consisting of amplitudes of atmospheric pressure with a constant
frequency were applied to half-full seal, and the minimum amplitude leading to
instantaneous seal break within 10 seconds (referred to as seal break pressure) at the
given frequency was measured. Then the seal break characteristics curve was obtained
as a function of frequency and seal break pressure.
5.3 Results and discussion
Seal break characteristics curves (seal break pressure curves) for each test trap are
shown in Figure 5. The seal break characteristics curves roughly corresponded with
density distribution curves obtained from the free vibration test and response
magnification curves from single sine wave response test. The frequency where seal
break pressure was minimum also corresponded with the natural frequency f0. However,
the minimum value of seal break pressure was extremely small in WC. The reason for
this seems to be that WC was made of opaque ceramic, and that instantaneous seal
break was not detected accurately as the conditions of seal could not be observed
directly.
Based on dynamic oscillation, seal break test is geared to focus on seal break, the most
important element in this study, and seal break pressure provides convenient data to
evaluate trap performance. However, as mentioned above, seal break test has a
drawback in that instantaneous seal break cannot be detected accurately if traps are
made of opaque materials.
414 6. √2 times natural frequency response test
6.1 Purpose
In the field of oscillation technology, frequency √2 times the natural frequency f0 of
experimental subject has been used as eigenvalue of force. This experiment is
Frequency [Hz] 0.0
2.0
Frequency [Hz] Frequency [Hz] 4.0
6.0 0.0
2.0
4.0
6.0 0.0
2.0
0
P trap Seal break pressure [Pa] -200
-400
4.0
6.0
WC‐6L Bell trap Psmin.= Psmin.= Psmin.= ‐196 Pa ‐216 Pa ‐183 Pa -600
-800
0.0
2.0
6.0 0.0
4.0
2.0
4.0
6.0 0.0
2.0
4.0
6.0
0
-200
-400
S trap Contrary bell trap WC‐8L Psmin.= Psmin.= Psmin.= ‐186 Pa ‐204 Pa ‐152 Pa -600
-800
Psmin : Minimum seal break pressure Figure 5 – Seal break curve of test raps
intended to define trap performance by studying response levels of seal to single sine
wave with √2 times natural frequency f0.
6.2 Experimental method
Single sine waves mainly consisting of atmospheric pressure with various amplitudes
and constant frequency were applied to half-full seal for 30 seconds, and seal
fluctuations and seal loss Hloss were measured. Pressure waves with natural frequency f
and √2 times f0 were applied.
6.3 Results and discussion
The relationship between applied pressure and maximum seal fluctuation for each trap
is shown in Figure 6. The input pressures (pressures waves inputted into the pressure
generator) were retained at a constant level until the amplitude of applied pressure
reached approximately ±150Pa, but the larger the amplitude, the more differences in
output pressures (pressures actually measured in the pipe of test apparatus) were
actually measured in traps (Table 2, Table 3). This phenomenon seems to have more to
do with seal volume than with the size of cross-sectional areas of trap legs. It can be
assumed that input pressure was lessened by resonance phenomenon when frequency f0
is applied, but on the other hand, air inside the apparatus became compressed before
415 applied pressure had any influence on seal fluctuation as seal volume increased when √2
times f0 frequency was applied. As far as seal loss is concerned, no evaluation of trap
performance was possible as seal loss remained zero regardless of amplitude of applied
pressure at √2 times f0 frequency (Figure 7).
The experiment using √2 times natural frequency f0 proved to be unproductive as it
required cumbersome procedures and considerable time, and it was difficult to adjust
amplitude of applied pressure because of compression of air inside the apparatus. It was
also found difficult to obtain clear indices by which performance of each trap can be
evaluated and ranked.
120
P trap Bell trap WC‐6L Maximum seal water level [mm] 90
60
30
0
0
-200
-400
120
-600
0
-200
S trap -400
-600
0
-500
Contrary bell trap -1000
WC‐8L 90
60
7
30
0
0
-200
-400
Pressure [Pa] -600
0
-200
-400
Pressure [Pa] -600
0
-500
Pressure [Pa] Figure 6 – Relations ship between Pressure and max. water level of test raps
416 -1000
Table 2 Input pressure and output pressure (f0)
Input pressure [Pa]
Test traps
150
200
250
350
400
450
550
650
750
850
Output pressure [Pa]
P trap
225
229
287
352
425
--
--
--
--
--
S trap
184
237
252
359
424
--
--
--
--
--
Bell trap
164
206
220
337
371
--
--
--
--
--
Contrary
tarp
bell 169
235
250
336
378
--
--
--
--
--
WC-6L
246
262
318
400
480
405
469
545
564
637
WC-8L
105
120
174
262
310
283
355
418
442
456
Table 3 Input pressure and output pressure (√2・f0)
Input pressure [Pa]
Test traps
150
200
250
350
400
450
550
650
750
850
Output pressure [Pa]
P trap
167
226
288
323
453
--
--
--
--
--
S trap
214
201
313
367
445
--
--
--
--
--
Bell trap
222
255
333
431
497
--
--
--
--
--
Contrary
tarp
bell 179
223
308
408
490
--
--
--
--
--
WC-6L
213
238
314
383
479
511
631
765
941
1.057
WC-8L
187
241
319
422
502
523
600
695
816
900
417 P trap f0
√2・f0
0
-200
-400
-600
Seal loss [mm] Seal loss [mm] 40
35
30
25
20
15
10
5
0
40
35
30
25
20
15
10
5
0
Bell trap f0
√2・f0
0
-200
-400
-600
Pressure [Pa] Pressure [Pa] Figure 7 – An example of output pressure against input pressure (f0, √2・f0)
7. Conclusion
√2 times natural frequency response test was found inappropriate for testing trap
performance. Considering the ease of experimental method and the size of
experimental apparatus, minimum seal break pressures obtained by the seal break test
seems to provide most appropriate criteria for evaluating trap performance. In addition,
it is desirable to include natural frequencies and maximum response magnifications
obtained by the free vibration test and the single sine wave response test respectively.
An issue that needs to be addressed in the future is how to apply the seal break test to
opaque traps. One way to achieve this would be not to judge instantaneous seal break
by visually, but to use a water level indicator, and regard seal break as having occurred
when water level reaches near seal depth 0mm when pressure is applied.
In this study performance evaluation was only made in terms of negative pressure,
where seal break is most likely to occur, as we used amplitudes of pressure waves based
on atmospheric pressure. Therefore it is necessary to develop a method of performance
evaluation in terms of positive pressure by applying bias inside the apparatus.
418 8. References
1. Sakaue, K., Kamata,M., Kuroda, K., Wang Y. (2001). Studies on Dynamic
Characteristics of Trap Seal, Proceedings of CIB W062 International Symposium (pp.
E4-1-4-5)
2. Sakaue, K., Kamata, M., Zyang, Y. (2007). A study on the Test method of Trap
performance, Proceedings of CIB W062 International Symposium (pp.321-332)
3. Kuriyama H., Sakaue, K., Yanagisawa, Y., Kamata, M., Sudo, H. (2006), Studies on
the Test method of Trap performance (Part 11), Technical papers of Annual meeting,
SHASE (pp.793-796).
4. Kuriyama, H., Sakaue, K., Yanagisawa Y., Kamata, M., Iizuka, H. (2008). A study
on the Test method of Trap performance using Simple Test Apparatus, Proceedings
of CIB W062 International Symposium (pp.252-263)
5. Sakaue, K., Shinohara, T., Kaizuka, M. (1977). A Study on the Dynamic
Characteristics of the Trap Seal, Proceedings of CIB W062 International Symposium.
6. Sakaue, K., Shinohara, T., Kaizuka, M. (1982). A Study on the Dynamic
Characteristics of the Seal in Deformed Traps, Proceedings of CIB W062
International Symposium.
7. Tomonari, H., Wang, Y., Kamata, M., Sakaue, K. (2002). Numerical Studies of Seal
Movement in Traps, Transactions of SHASE (pp.87-96)
419 9. Presen
ntation off Authors
Dr. Eng.) iss a professsor at Dep
partment off
Kyosuke Sakaue (D
Architectuure, of Schoool of Sciennce & Techhnology, and a head off
New Plum
mbing Systeem Institutee, Meiji Unniversity. His
H fields off
specializaation includde water envvironment, building su
urvices andd
plumbing system. He
H is curreently engagged in the studies off
d
syystems, trapps, toilets, stainless stteel piping,
siphonic drainage
water saviing systemss, maintenannce, history of plumbin
ng system.
Hana Kuriyama
K
(M
M. Eng.)
Nikkenn Sekkei Ltdd.
She is engaged
e
in Planning
P
annd Air cond
ditioning sysstem.
Motoyaasu Kamataa (Dr. Eng.)
A profeessor at Kannagawa Uniiversity
He is reesearching Building
B
Seervices and energy conservation.
Hiroshii Iizuka (P.E
E.Jp)
Nikkenn Sekkei Ltdd.
He is enngaged in planning
p
of plumbing
p
sy
ystems and air conditiooning
system. He also works as a coonsultant.
42
20 Session VIII: Miscellaneous
VIII.1
Analysis of corrosion mechanism of sprinkler
piping and corrosion protection
Toshihiro Yamate a),Saburo Murakawa b)
(a) Technical Research Laboratory, Takenaka Corporation
1-5-1 Ohtsuka, Inzai City, Chiba, 270-1395, Japan
(b) Graduate School of Engineering, Hiroshima University
1-4-1 Kagamiyama, Higashi-Hiroshima City, 739-8527, Japan
Abstract
The corrosion of the pre-action sprinkler 100A straight piping which was galvanized
steel pipe caused the water leakage. The piping was used for 1.5 years. When the valve
operation was examined, the testing water was remained in the piping system.
Therefore, the localized corrosion caused in the soaking part. As the localized corrosion
part, it penetrated through the piping with a very fast corrosion rate of 2.67mm/year.
Piping in the vicinity of the corrosion part was cut out, and the electric potential in the
health part (zinc plating layer remaining as a part) and the localized corrosion part
(outcrop of the carbon steel as a basic metal) were measured. As a result, the reversal
electric potential between the zinc and the carbon steel was confirmed. It made to
atmosphere near the oxygen partial pressure in the sprinkler piping, and the galvanic
current was measured under the composition of the corrosion battery of which the
health part was a cathode and the corrosion part was an anode. A galvanic current of the
level corresponded to an actual corrosion rate was observed. It was clarified that the
reversal electric potential between the zinc and the carbon steel was a cause of the
localized corrosion. The water quality and the corrosion products were analyzed, and
the factors of which the electric potential of zinc was made to a noble side were
examined. As another example, it leaked from the screw in the air layer of the upper part
of the wet sprinkler piping after one year usage (corrosion rate: about 1.75mm/year). It
421 was clarified that the corrosion caused under the reversal electric potential between the
zinc and the carbon steel similar to the pre-action sprinkler piping. The pressure in both
piping of pre-action sprinkler piping and wet sprinkler piping was high. It is thought
that the height of dissolved oxygen concentration of water became a big driving force of
the localized corrosion for that. It was suggested for preventing corrosion that the water
in pre-action sprinkler piping should be pulling out completely, and that the airs in preaction sprinkler piping and wet sprinkler piping should be substituted for nitrogen.
Keywords
Sprinkler Piping,Galvanized Steel Pipe,the Reverse of Electrical Potential between
the Zinc and the Carbon Steel,Hyperbaric Air, Substitution with Nitrogen
1. Introduction
The pre-action sprinkler system is mainly used in cold regions where there is a danger
of freezing in the pipes, but in recent years it has been used outside of cold regions, with
the objective of avoiding accidental sprinkling of water due to faulty operation of the
wet-pipe sprinkler system. In this case, the pipes are filled with air, so it has been
thought that there is no problem with corrosion. However, recent surveys have shown
that water leakage events have occurred due to corrosion in the pipes of pre-action
sprinkler systems. Water leakage due to corrosion in pre-action sprinkler systems has
occurred in the earliest cases one-and-a-half years after start of use, and most cases have
occurred in the short period of three to four years after start of use. Also, in the wetpipe sprinkler system, the pipes are filled with water but there almost no replacement of
the water, so there have been almost no reports of water leakage due to corrosion.
However, water leakage due to corrosion has occurred in the sprinkler pipes in a
warehouse within only one year after start of use. This type of corrosion incident has a
major impact of the fire extinguishing function during an emergency, so determining its
causes and countermeasures is extremely important for maintenance of buildings. In
this report, the results of analyzing the occurrence of corrosion in a pre-action sprinkler
system (case A) and a wet-pipe sprinkler system (case B) are described. Further,
countermeasures are proposed based on the analysis results.
422 2. Analysis of corrosion case A in a pre-action system
2.1 Process of occurrence of corrosion and the piping system
Water leakage occurred from a 100A horizontal pipe one year and six months from start
of use, and from an 80A horizontal pipe three years and seven months from start of use.
The average corrosion rates calculated from the time elapsed and the thickness of the
pipe were 2.67mm/year and 0.98mm/year respectively. As the average corrosion rates
from the results1) of actual surveys of corrosion in architectural equipment piping and in
freshwater are about 0.1mm/year2), the corrosion rates in this case are extremely high.
A diagram of the sprinkler piping system is shown in Fig. 1. A fire pump fills each
header and the deluge valves on the first floor with water from the fire tank installed in
the basement, and compressed air at 4105Pa pressure fills the secondary side piping
from the deluge valves onwards. The water leakage due to corrosion occurred in each
case in the secondary side piping system of header-2, where water that had not been
discharged had remained in the pipe from the time of the valve test.
Leakage part by corrosion
1F
Sprinkler head
Deluge valve
Header 2
Header 1
Air piping
Header 3
Water piping
Makeup water pipe for
tap water
B1F
Fire pump
Fire tank
Figure 1: Schematic diagram of pre-action sprinkler system in case A
2.2 Analysis of the corrosion phenomenon
2.2.1 Observation of the sample pipes and analysis of the pipe material and corrosion
products
The pipe (80A, pipe thickness 3.5mm, horizontal pipe) at the water leak location that
occurred three years and seven months after start of use was cut out, the inner surface of
the pipe was observed, and the pipe material and the corrosion products were analyzed.
Figure 2 shows the state of the inside surface of the pipe, and Fig. 3 shows the state of
the inside surface and the cross-section of the corroded part. The inner surface of the
pipe was separated vertically in the horizontal direction, white corrosion products had
formed over the whole undersurface side, and corrosion tubercles were dispersed within
423 the corrosion products. Localized corrosion had occurred below the corrosion tubercles.
On the upper surface there were few white corrosion products, but the gray metal under
layer was visible. It is considered that the there was an air layer on the upper surface
side, and water remained in the undersurface side. As shown in Figs. 2 and 3, the
localized corrosion consisted of pit holes that were circular in plan and bowl shaped in
cross-section, with perforations in the bottom of the pit holes. The corrosion rate at the
perforations was calculated to be 0.98mm/year, as stated previously. As shown in Fig.
4, corrosion products were sampled from the surface layer of the corrosion tubercle (the
outer surface side and the inner surface side of the corrosion tubercle), the inside and the
outside of the corrosion pit, and the zinc plating layer, and analyzed using energy
dispersive X-ray spectroscopy (EDX) and X-ray diffractometry (XRD). Also, the crosssection was analyzed using an electron probe X-ray micro analyzer (EPMA). The EDX
and XRD analysis results are shown in Table 1. According to Table 1, in the intact part
mainly zinc oxide ZnO, and zinc hydroxide Zn(OH)2 occurs. Zinc oxide has the
properties of a semiconductor, and from the behavior of its passive film the potential of
the zinc is said to become more noble3), 4). Figure 5 shows the analysis results of the
corrosion pit cross-section using EPMA. Within the corrosion pit, areas with
concentrations of Cl and S (thought to be due to the sulfuric acid group SO42-) were
found, so it is considered that hydrochloric acid HCl and sulfuric acid H2SO4 were
generated by the hydrolytic reaction as shown below, which reduced the pH within the
pit hole and promoted corrosion.
FeCl2 + 2H2O → Fe(OH)2 + 2HCl
FeSO4 + 2H2O → Fe(OH)2 + H2SO4
The analysis results of the pipe material showed that the content of P (0.029wt%) and S
(0.024wt%) were each within the specified range of JIS G 3452.
Figure 2: Inside corrosion status of sample pipe Figure 3: Inside surface and cross section of perforation part by localized corrosion
424 B(Outside)
Tubercle
E
D
Corrosion product
A
C(Inside)
Zinc plating layer
Carbon steel(The basic metal)
Figure 4: Sampling point of corrosion products in cross section of localized corrosion
Table 1 - Analysis of corrosion products by EDX and XRD
Sampling point (In figure-4)
Detected main elements (EDX)
Identified compounds (XRD)
Zn>O C, Fe, Al, Mg
ZnO, Zn(OH)2
Zn, O, Fe>C, Mn, Cl, S>Si, Al
ZnO, Zn(OH)2, Fe3O4, Fe,
FeO(OH),  -FeO(OH)
Sample A
(Surface of zinc plating layer)
Sample B
(Outside of surface of tubercle)
-
Sample C
Fe, O>Zn>C>S, Si, Al
(Inside of surface of tubercle)
Fe3O4,
Sample D
Fe, O>Zn>C>Si, Al
(Inside of tubercle)
FeO(OH),

O, Fe C, Cl, S, Si, Cu, Zn, Cr, Al
 -FeO(OH), 
(Inside of corrosion pit)
Compound where possibility of existence is suggested in Sample D: σ-FeCr
425 -
-FeO(OH), FeCO3,
ZnO, Zn(OH)2
Sample E
 -FeO(OH), 
Fe,
-FeO(OH), ZnO
Figure 5: EPMA analysis of cross section in perforation part by localized corrosion
2.2.2 Water quality
Water quality analysis was carried out for the water that remained in the pipes and the
re-supplied water. Table 2 shows the water quality analysis results for the water that
remained in the pipes, the re-supplied water (source water), and Inzai City tap water
which was used in the tests for comparison. The characteristics of the re-supplied water
compared with the average values of tap water within Japan5) are that the electrical
conductivity, the hardness, the M alkalinity, chloride ions, sulfate ions, nitrate ions,
soluble evaporation residues, and evaporation residues are high, and the saturation index
is positive (+). The water quality of the water remaining in the pipes tended to have
increased iron and zinc as a result of corrosion, reduced hardness due to precipitation
reactions on the surface of the pipe or in the water, reduced nitrate ions and increased
ammonium ions due to reduction reactions, etc. Also, the increase in KMnO4
consumption is considered to be due to the activity of microbes, or contamination with
organic matter, or oxidation of Fe2+ (Fe2+  Fe3+) eluted from the pipe. Of these, it is
considered that the reason the concentration of nitrate ions NO3-, bicarbonate ions
HCO3- (M alkalinity  1.22) were high is the potential of the zinc was made noble4), 6), 7).
426 Table 2 - Analysis for makeup water and residual water in sprinkler piping, and
tap water of Inzai City, Chiba, Japan
Sample
Unit
Makeup
water
(Tap water)
Residual water in
sprinkler piping
Tap water of Inzai
City
Analysis item
Turbidity
NTU (×0.7)
<1
123
<1
Color
HU (×0.7)
<2
15
<2
7.2
8.5
7
pH
Conductivity
mS/m
77.0
68.2
27.9
M-alkalinity (as CaCO3)
mg/L
300
280
43
Ca hardness (as CaCO3)
mg/L
228
10
57
Total hardness (as CaCO3)
mg/L
394
32
81
Cl-
mg/L
47.3
37.7
25.8
SO42-
mg/L
45.1
36.3
47.2
NH4+
mg/L
0.01
1.71
0.02
NO2-
mg/L
<0.01
0.12
<0.01
NO3-
mg/L
29.2
0.62
13.1
Soluble residue
mg/L
508
491
199
Total residue
mg/L
535
688
204
Soluble SiO2
mg/L
17
6
21
Total SiO2
mg/L
17
11
21
KMnO4 consumption
mg/L
2
59
2
Fe
mg/L
<0.1
53.3
<0.1
Zn
mg/L
0.05
22.7
<0.05
Cu
mg/L
<0.05
<0.05
<0.05
Mn
mg/L
<0.01
0.21
<0.01
Cr
mg/L
<0.05
<0.05
<0.05
0.2
0.1
-1.4
Saturation Index (at 25C)
2.2.3 Inference of the causes of the corrosion
427 In Fig. 2 it can be seen that the carbon steel base metal is corroded locally over a wide
area in the intact part (area fully covered by zinc plating layer), so this indicates that at
least the sacrificial anode action of the zinc with respect to the carbon steel was not
effective. Also, the localized corrosion had a circular shape in plan, which suggests that
the intact portion had a strong cathodic action with respect to the corroded portion.
Furthermore, in Table 1 it can be seen that oxidized zinc is detected in the intact part, in
Table 2 it can be seen that the concentration of hydrogen carbonate ion is high, and the
concentration of nitrate ion is high in the re-supplied water. Therefore, it is considered
that the localized corrosion was caused by a galvanic couple action (the flow of an
electric current due to a electrochemical reaction) caused by the reversal of the potential
between the zinc (intact portion) and the carbon steel (corroded portion). The area ratio
of the cathode (intact part = zinc) and the anode (corroded portion = carbon steel) is
high, so the current density in the anode portion is increased greatly, and it is considered
that this promoted localized corrosion. The reason that water leakage due to corrosion
occurred only in the secondary side piping of the header-2 system is because water
remained within the pipes of this system.
2.3 Tests to reproduce the corrosion phenomena
2.3.1 Test method
Four test samples were cut from the intact portion (zinc) and the corroded portion
(carbon steel) of the sample pipe, lead wires were soldered to their outside surfaces, and
electrodes were made in which only the inside surface of the pipe was exposed, and all
other parts were covered with resin, as shown in Fig. 6. Likewise, three each electrodes
were made by exposing the zinc plating on the outside surface of the pipe and exposing
the carbon steel base metal on the inside surface of the pipe by using wet-type #400
sandpaper. For both the intact portion electrodes and the corroded portion electrodes,
the natural potential was measured by immersing in the water remaining in the pipe
under atmospheric pressure, room temperature, and static conditions using an
electrometer with a saturated mercury chloride (I) electrode(Saturated Calomel
Electrode=SCE) as the reference electrode. Both the electrodes of the intact portions
and the corroded portions were immersed in the remaining water, the re-supplied water,
and the Inzai City tap water, and the natural potential was measured for 49 days under
atmospheric pressure, room temperature, and static conditions in the same way. Also,
to investigate the reversal of potential between the zinc and the carbon steel by the
galvanic couple action, a cathode electrode was made from the intact portion and an
anode electrode was made from the corroded portion, and the galvanic current was
measured by a non-resistance ammeter, as shown in Fig. 7. The pressure within the
sprinkler pipes is 4105Pa, or about four times atmospheric pressure, so it is considered
that the concentration of dissolved oxygen in the water remaining in the pipes was
428 higher than that under atmospheric pressure (saturated concentration 8.11mg/L at
25C). Therefore, the top surface of the test cell was covered with film, and oxygen
was input into the gaseous phase, so that the galvanic current was measured in a pure
oxygen atmosphere (equivalent to atmospheric pressure 4.82105Pa). Further, in order
to investigate the effect of deoxidation as a countermeasure, the gaseous phase was
replaced with pure nitrogen and the behavior of the galvanic current was investigated.
Figure 6: Electrode for electric potential measurement
429 Non-coating
Non-coating
Coating
Coating
Coating for the
surrounding
Coating except
localized
corrosion
Zinc plating
Layer
remaining
(Cathode)
Back
Coating
(Back and side)
The position
of localized
corrosion
Coating
(Back and side)
Side
Corrosion part
(Anode)
10mm ø
3.5mm t
3.5mm t
The front of the The back and the side
The front of the The back and the side
electrode
of the electrode
electrode
of the electrode
Anode electrode (Corrosion part)
Cathode electrode (Zinc plating layer remaining)
33mm ø
Nonresistance ammeter
A
Residual water
in sprinkler piping
The position of
localized corrosion
Cathode electrode
(Zinc plaiting
layer remaining)
10㎜
Anode electrode
(Corrosion part)
Figure 7: Electrode for galvanic current measurement, and measuring method of
galvanic current
2.3.2 Test results and analysis
1) Natural potential
Figure 8 shows the measured results for the natural potential of each electrode. In the
water remaining in the pipe, the potential of the zinc of the intact portion of the sample
pipe (portion with the zinc coating remaining) and the surface of the zinc coating on the
outside surface of the pipe became more noble with time, and the potential of the carbon
steel of the corroded portion of the sample pipe and the base metal (the carbon steel
base) became more active compared with the start of immersion. It is considered that
the reason the potential of the intact portion of the sample pipe is more noble than the
surface of the zinc coating of the outside of the pipe, and the potential of the corroded
portion is more noble compared with the potential of the carbon steel parent metal is
430 because of polarization (the phenomenon in which the natural potential is changed by
the current entering and leaving the surface of the metal) due to the corrosion. Also, the
potential of the intact portion of the sample pipe was already more noble than the
potential of the corroded portion at the start of immersion, and a potential difference of
320mV was measured on the 49th day. In other words, reversal of the potential of the
zinc (intact portion) and the carbon steel (corroded portion) was observed. It is said that
to prevent corrosion due to contact of dissimilar metals the potential difference should
be less than 50mV8), so it is considered that a potential difference of 320mV is
sufficient to drive a galvanic coupling action.
0
~
~
The potential of the zinc plating of the outer surface of the pipe in the water that
remained in the pipe was the most noble, and was about -450mV (vs. SCE) on the 49th
day, so a reversal of the potential with the carbon steel parent metal (about -690mV)
occurred. There was a slow trend for the potential to become more noble with time in
the re-supplied water, and on the 49th day about -780mV was measured. In the Inzai
City tap water, the potential temporarily became more noble, and again became more
active. It is considered that the difference in the behavior of the potential of the zinc is
due to the difference in the type and properties of the oxide films formed on the surface
of the metal, so it is considered that it is affected by the water quality as stated
previously.
Atmospheric
pressure,Room temperature:ab.21℃
大気解放
室温(
約21℃)
Static
state
Potential(mV vs. SCE)
静止状態
滞留水中
(Zn)Zinc layer remaining
健全部
(Zn)
part in sprinkler piping
Residual Water in
sprinkler piping
(Fe)Corrosion part
Zn 滞留水中
腐食部
sprinkler piping
(Fe)
-500
Makeup
water 印西市水
Inzai city tap water
炭素鋼 basic metal of
Fe(The
(母材)
sample pipe)
補給水
滞留水
Distilled
water
滞留水
補給水
Makeup
亜鉛
Zn(The
outer surface of
(管外面)
sample pipe)
water
0 1
3
5
7
Time (day)
~
~
印西市水
Inzai
city tap water
-1000
in
49
Figure 8: Change over time in potential of sample pipe in pre-action sprinkler system
2) Galvanic current
Figures 9 and 10 show the measurement results for the galvanic current. In the galvanic
couple (area ratio of intact portion / corroded portion: 10.9) formed by the intact portion
and the corroded portion, a galvanic current (the value after stabilization) of about
0.0325mA was measured with the intact portion as cathode and the corroded portion as
431 anode, as shown in Fig. 9. From the form of the corrosion pit shown in Fig. 3, it can be
assumed that the diameter of the anode surface is about 10mm, so a corrosion rate of
0.48mm/year can be calculated from the current density of the galvanic current. In the
case of a pure oxygen atmosphere as the gaseous phase, the galvanic current increased
by a factor of about 6 (0.075mA/0.012mA) compared with an atmosphere of air, as
shown in Fig. 10, or a greater increase than the oxygen partial pressure increase (a
factor of 4.7). From this it is inferred that within the actual sprinkler pipes (4105Pa)
the galvanic current was a factor of 3.95 (pressure within the pipe / atmospheric
pressure) or greater, so the corrosion rate was calculated to be 0.483.95 =
1.896mm/year. This value is of the same order as the localized corrosion rates that
occurred.
Also, when the gaseous phase was changed back to air again from pure oxygen, the
galvanic current only reduced down to about double that before replacement with
oxygen. It is considered that this is because when the atmosphere was pure oxygen the
concentration of oxygen dissolved in the water increased, and as oxidation of the film
on the intact portion proceeded, it became more noble. Further, when the gaseous phase
was replaced with nitrogen the galvanic current gradually reduced to less than 0.3A as
the test water became deoxygenated. This is because as the concentration of dissolved
oxygen reduces, the following cathode reduction reaction is greatly reduced, so this
indicates that replacement of the gas within the pipes with nitrogen is effective as a
countermeasure against this corrosion phenomenon.
Galvanic current (mA)
[Dissolved oxygen cathode reduction reaction]
0.1
1
O 2  H 2 O  2e  2OH 
2
Atmospheric pressure
Cathode/anode ratio:10.9
Room temperature:21.6℃
Static state
Residual water in sprinkler piping
0.05
0
0
60
Time (min.)
120
Figure 9: Change over time in galvanic current in galvanic couple of zinc layer
remaining part/corrosion part
432 Galvanic current (mA)
0.08
Room temperature
Static state
Residual water in sprinkler piping
0.06
Air
O2
Air
N2
0.04
0.02
0.00
0
500
1000
Figure 10: Change in galvanic current in galvanic couple of zinc layer remaining
part/corrosion part according to kind of gas in vapor phase part
3. Analysis of corrosion case B in a wet-pipe system
3.1 Process of occurrence of corrosion and the piping system
Figure 11 shows a diagram of the sprinkler piping system. One year after start of use
water leaked from the screwed part of the branch between the vertical main piping and
the horizontal piping, and one year and five months after start of use water leaked from
a location of the same specification at the same height. Assuming the thickness at the
screw portion was reduced by cutting the screw thread, and that the screw portion was
penetrated and water seeped out9), and assuming the pipe thickness at the penetration
location was about 1.75mm (32A), it is estimated that the average corrosion rate was
1.75mm/year or more. The water for fire extinguishing is rainwater that is filtered,
sterilized with chlorine, and supplied as miscellaneous water, and when the quantity of
miscellaneous water is insufficient, the miscellaneous water tank is replenished with tap
water. The pressure within the pipes is 9.63105Pa, and the internal surfaces of the fire
extinguishing water tank are concrete with a mortar finish.
433 40A
GL
32A
Tee
Leakage
(Details in leakage part by corrosion)
21.7m
ab.30m
Leakage part by corrosion
Rain
Sprinkler head
Fire pump
Fire tank
Filter
Non-potable Rain receiving
water tank tank
Figure 11: Schematic diagram of wet-pipe sprinkler system in case B
3.2 Analysis of corrosion phenomena
3.2.1 Observation of the sample pipes and analysis of the pipe material and corrosion
products
The plan shape of the corrosion tubercles in the leakage area were irregular, but within
them circular localized corrosion occurred. Figure 12 shows the state inside a pipe at the
same height as the water leakage and a part with no water leakage at a lower height
(horizontal pipes in both cases). In the pipe at the same height as the water leak,
corrosion tubercles can be seen everywhere, and localized corrosion has progressed
similar to the water leakage location. Apart from this location, white products that
appear to be the corrosion products of zinc have formed to a uniform thickness. Also,
the distribution of corrosion tubercles reduces towards the end of the pipe, and near the
sprinkler heads corrosion tubercles were not seen. On the other hand, in the lower pipe
where there was no water leak, white products were formed on all surfaces, but the
thickness of the white products layer was thinner compared with the pipe where the
water leak occurred, and no corrosion tubercles were seen. In the pipes at a height
higher than the location of the leak, there was a slight change of color to gray, but no
corrosion tubercles or white products distributed in the intact portion were seen. This
suggested that there was no water above the location of the water leak, and subsequently
it was confirmed that there was no water by using the ultrasonic longitudinal wave
method. In other words, it is considered that in the part above the water leak all the air
was not bled from within the pipes, so there was a large quantity of compressed air
within the pipes.
Table 3 shows the results of analysis of the corrosion products at the location of the
water leak and at a location with no water leak. At the intact parts of the water leak
location and the location with no leak, zinc oxide ZnO and zinc hydroxide Zn(OH)2
occurs, and the possible presence of calcium carbonate is indicated. Also, the EPMA
434 surface analysis results of the corrosion pit cross-section showed areas of concentrations
of S and Cl in the bottom of the corrosion pit and in the interior, which suggests the
generation of hydrochloric acid HCl and sulfuric acid H2SO4 by a hydrolytic reaction.
According to the analysis results of the pipe material, the content of P and S were within
the range specified by JIS G 3452.
Table 3 - EDX analysis and XRD analysis of corrosion products
Sampling point
Detected main elements (EDX)
Identified compounds (XRD)
Zinc layer remaining part in leakage
piping
Zn, O>Fe, C, Cl, S
ZnO, Zn(OH)2, (CaCO3)
Corrosion part in leakage piping
Fe, O>Zn, C, S
Zinc layer remaining
nonleakage piping
part
in
Fe3O4, ZnO, (  -FeOOH,

-
FeOOH)
Zn, O>Fe, C, Cl, Ca, Mg, Si, S
ZnO, Zn(OH)2, (CaCO3)
( ): Compounds where possibility of existence are suggested in sample pipes.
3.2.2 Water quality
The results of the analysis of the water quality of the water in the fire tank are shown in
Table 4. The electrical conductivity and the concentration of dissolved chlorides are
within the range of the water quality of normal tap water5), so it is inferred that the
proportion of tap water was high. Also, the pH and M alkalinity were high, and it is
considered that this was due to the effect of elution of alkaline compounds from the
surface of the fire tank mortar finish.
Table 4 - Analysis for water in fire tank
Sample
Water in fire tank
Unit
Analysis item
pH
9.0
Conductivity
mS/m
30.5
M-alkalinity (as CaCO3)
mg/L
60
435 Ca hardness (as CaCO3)
mg/L
30
Total hardness (as CaCO3)
mg/L
33
Cl-
mg/L
30
SO42-
mg/L
41.8
NO3-
mg/L
9.7
Na+
mg/L
35.1
Soluble SiO2
mg/L
15
Fe
mg/L
<0.05
Zn
mg/L
0.03
3.2.3 Measurement of the sprinkler pipe potential
Two each test specimens were cut from the pipe at the height of the water leak (water
leak system) and from the pipe at a height lower than the water leak (no water leak
system). The pipe external surfaces and the cut surfaces were coated with paint, and
only the pipe internal surface (the intact part where the zinc plating remained) was
exposed, and lead wires were connected to the pipe outside surface. The sample pipes
were immersed in water taken from the fire tank (Table 4), and the natural potential of
the pipe internal surface was measured for 13 days using a silver-silver chloride
electrode (Ag / AgCl / saturated KCl: displays a potential that is 42mV noble at 25C
compared with an SCE standard) as the reference electrode. The positions of
measurement of the potential are shown in Fig. 12. For comparison, the natural
potential of carbon steel (SS400, polished with wet-type #400 sandpaper) was also
measured. Measurement results of natural potential are shown in Fig. 13. In the water
leak system, the potential of the whole pipe internal surface was more noble than the
potential of the carbon steel, and the potential difference with the carbon steel was 80 to
240mV. In the no water leak system, the potential of the pipe internal surface when
initially immersed was more active than the carbon steel, but it became more noble with
time, and after 13 days it reversed by 40 to 120mV. Also, the potential of the parts
where much white products were found in both the water leak system and the no water
leak system tended to be more noble than the parts with few white products. From this
it is suggested that at the location of the water leak the potential of the zinc and the
carbon steel can be easily reversed, and the white products on the surface of the pipe
cause the potential to become more noble.
436 Figure 12: Inside corrosion status of sample pipes
-400
-450
Potential (mV vs. Ag/AgCl)
Potential (mV vs. Ag/AgCl)
-400
■A-1 □A-
-500
-550
-600
-650
-700
×Fe:soaked independent
-750
-450
◆B-1 ◇B-2
-500
-550
-600
-650
-700
-750
×Fe:soaked independent
-800
-800
0
100
200
300
0
400
-400
-450
-450
Potential (mV vs. Ag/AgCl)
Potential (mV vs. Ag/AgCl)
-400
-500
-550
-650
-700
×Fe:soaked independent
-750
200
300
400
●B-3 ○B-4
-500
-550
Pipe of nonleakage piping
▲A-3 △A-4
-600
100
Soaking time (hr)
Upper surface of nonleakage piping
Soaking time (hr)
Upper surface of leakage piping
-800
-600
-650
-700
-750
×Fe:soaked
-800
0
100
200
300
400
0
Soaking time (hr)
Under surface of leakage piping
100
200
300
400
Soaking time (hr)
Under surface of nonleakage piping
Figure 13: Change over time in potential of sample pipes in wet-pipe sprinkler
system
437 3.2.4 Inference of the cause of the corrosion
From the fact that localized corrosion occurred in the carbon steel base metal despite the
zinc plating layer remaining up to the side of the corroded part, and the potential of the
intact part where the zinc remained in both the water leak area and the areas with no water leak
were more noble than the potential of the carbon steel, it is inferred that reversal of the potential
between the carbon steel and the zinc occurred. The zinc plating layer was perforated by the
corrosion pits that occurred on the surface of the zinc layer, so the zinc did not act as a
sacrificial anode for the carbon steel base metal that was locally exposed, but conversely
corrosion of the carbon steel was promoted as the cathode. In other words, the potential of the
zinc layer was made more noble by the high pressure 9.63105Pa within the pipes, which
caused pitting corrosion and perforation of the zinc layer, so locally the carbon steel base layer
was exposed. Next, it is considered that localized corrosion was caused by the galvanic couple
action due to the reversal of the potential of the exposed carbon steel base metal and the zinc. It
is also considered that the high concentration of hydroxide ions in the water also promoted the
ennobling of the potential of the zinc4), 6), 7). An outline of these corrosion mechanisms is shown
in Fig. 14. Also, the corrosion in all cases occurred close to the water line, and there was
virtually no corrosion near the pipe terminals or near the headers, so it is considered that the
potential of the zinc layer was particularly noble near the water line. Normally in water the
concentration of dissolved oxygen becomes uniform with the passage of time due to diffusion of
the dissolved oxygen (concentration diffusion), but in this case it is inferred that a difference in
dissolved oxygen concentration was produced between the area near the water line and below it.
H20
Anode reaction
Zn→Zn2++2e
Cathode reaction
O2+H2O+2e→2OH-
Cathode
Cathode
Zn
Anode
Fe
① Localized corrosion in zinc plating layer
Corrosion products
H20
Cathode
Cathode
Making noble of the electric
potential by generation such as
ZnO ZnO on surface of zinc plating
Zn
The reverse of electrical potential
between the zinc and the carbon steel
Anode
Fe
② Penetration to carbon steel of basic metal
H20
Zn
Fe
Anode reaction
Cathode Fe→Fe 2++2e
Cathode
Fe 2+
Fe 2+
Anode
Anode
Cathode reaction
O2+H2O+2e→2OH-
③ perforation in galvanized steel pipe
Flow of current
Figure 14: Localized corrosion mechanism in galvanized steel pipe for sprinkler
piping
438 4. Conclusions
Water leakage occurred within a short period of time in pre-action sprinkler pipes and in
wet-pipe sprinkler pipes, in which there are virtually no corrosion expected on
theoretical and practical grounds, and. The results of analysis of each corrosion case
confirmed that reversal of potential between the zinc and the carbon steel was the cause
of the corrosion in both the pre-action sprinkler pipes and the wet-pipe sprinkler pipes.
The following are the main analysis results.
1) In both the pre-action sprinkler pipes and in the wet-pipe sprinkler pipes, the
potential of the intact part of the internal surface of the pipes (the part where the zinc
plating layer remained) was more noble than the corroded part (the carbon steel). The
potential difference in each case was greater than the 50mV which is considered to be
necessary for galvanic couple action.
2) In the galvanic cell formed between the intact part/corroded part (area ratio: 10:9)
in the pre-action sprinkler pipe, a galvanic current was measured in the same order as
that equivalent to the actual corrosion rate.
3) Zinc oxide was detected on the surface of the pipes in both the pre-action sprinkler
pipes and the wet-pipe sprinkler pipes, which suggests ennobling of the potential.
4) In both the pre-action sprinkler pipes and the wet-pipe sprinkler pipes, oxygen was
dissolved into the water from an air layer under high pressure, so the dissolved oxygen
concentration increased which promoted ennobling the potential of the zinc layer. As a
result, pitting corrosion was induced in the zinc layer, the potential reversed between the
exposed carbon steel base metal and the zinc plating layer, localized corrosion occurred
in the carbon steel base metal due to the galvanic couple action, and in a short period of
time the pipes were perforated.
439 5. Countermeasures
Figure 15 shows proposed countermeasure methods based on the results of the analysis
of the causes of the corrosion. The corrosion countermeasures are broadly divided into
environment measures to improve the corrosion environment on the internal surfaces of
the pipes, and material selection to improve the corrosion resistance of the material.
Environment measures include removing the water from the pipes in the pre-action
sprinkler system, or filling with nitrogen in order to eliminate oxygen. For the wet-pipe
sprinkler system countermeasures include eliminating the air layer using an air purge
valve, or the like, replacing the gaseous phase with nitrogen, or adding a corrosion
inhibitor to the water. Regarding material selection, if carbon steel pipe is used,
because the pressure is high there is the possibility of formation of an oxygen
concentration cell and the occurrence of localized corrosion.
Installation of drain valves
Drain
Pre-action method
Deoxidation
Environmental
treatment
Wet method
Corrosion prevention
measures in
sprinkler piping system
Deaeration
Installation of air purge valve
Deoxidation
Filling of nitrogen
Corrosion
inhibitor
Corrosion resistance
metal
Material selection
Pre-action method
・Wet method
Filling of nitrogen
Lined steel pipes
Plastic pipes
Nitrite
Molybdate
Stainless steel pipes
Copper pipes
Unplasticized polyvinyl chloride
lined steel pipes
Polyethylene powder lined steel pipes
High impact resistance polyvinyl chloride
pipes
Unplasticized polyvinyl chloride pipes
Cross-linked polyethylene pipes
Polyethylene pipes
Polybutene pipes
Figure 15: Corrosion prevention measures in sprinkler piping system
440 6. References
1.
Yamate, Toshihiro (1991). An Investigation on the State of Piping Corrosion in
Existing Buildings, Journal of Architecture, Planning and Environmental
Engineering, No. 426, (pp. 67-81), August.
2.
Matsushima, Iwao (2002). Practical Knowledge for Corrosion Prevention, Ohmsha,
Ltd., (p. 94).
3.
Gilbert, P. T. (1952). The Nature of Zinc Corrosion Production, Journal of the
Electrochemical Society, 99, (pp. 16-21).
4.
Fujii, Tetsuo (2001). Corrosion Prevention Science, Kogyo Chosakai Publishing,
Inc., pp. (126-127).
5.
Takazaki, Shinichi (2001). Freshwater Quality and Corrosion Indices in
Architectural Equipment, Japan Society of Corrosion Engineering, 25th Technical
Seminar, (pp. 13-22).
6.
Hoxeng, R. F., Prutton, C. F. (1949). Electrochemical Behavior of Zinc and Steel in
Aqueous Media, Corrosion, 5, (pp. 330-338).
7.
Matsugawa, Yasuki, Miyashita, Mamoru, Okada, Katsuyuki, Miyata, Yoshikazu,
Asakura, Shukuji (1998). Research into Potential Reversal Phenomena between
Zinc and Steel (Part 3) The Effect of Various Anions, 45th Materials and
Environment Discussion Meeting, B-203, (pp. 175-178).
8.
Wranglen, G., Yoshizawa, Shiro, Yamakawa, Koji, Katagiri, Akira joint translation
(1973). Introduction to Corrosion Prevention in Metals, Kagaku-dojin Publishing
Co., Ltd., (p. 87).
9.
Yamate, Toshihiro (2004). Corrosion Diagnosis and Monitoring of Architectural
Equipment Piping, Zairyou-to-Kankyo 2004, (pp. 215-218), April.
7.
Presentation of Author
Toshihiro Yamate is an Assistant to General Manager in the
Environmental Engineering Section of the Takenaka Research &
Development Institute at Takenaka Corporation. His area of expertise is
research into the prevention of corrosion in building equipment piping.
Recently he has been engaged in the study of water treatment for
equipment and sanitation.
441 VIII.2
Study of sanitary equipments installed on
light-weight partitions
(1) M.C. Lee, Dr. (2) R.Z. Wang, Dr. (3) C.L. Cheng, Dr. (4) Y.C. Yu, Mr. (5) Z.Y. Shih,
Ms.
(1) MCJL@mail.ntust.edu.tw, (2) rzwang@ncree.org.tw, (3) CCL@mail.ntust.edu.tw
(1) National Taichung Institute of Technology, Department of Interior Design; 129,
Sec.3, Sanmin Road, Taichung 404, Taiwan.
(3)(4) National Taiwan University of Science and Technology, Department of
Architecture; 43, Sec.4, Keelung Road, Taipei 106, Taiwan.
(2)(5) National Center for Research on Earthquake Engineering; 200, Sec. 3, HsinHai
Road, Taipei 106, Taiwan.
Abstract
Sanitary equipment installed on walls are supported by the shear force of bolts or joint
and the friction between bolts and the wall material. Light-weight partition walls are
frequently used in building interiors because of the quick delivery time. The downside
of light-weight partitions is they have a reduced dead load capacity due to its structure.
As a result of the change in structure, the friction between bolts and the board material
is reduced, making installation of sanitary equipment difficult. Therefore, this study
tries to analyze the structural relationship between the sanitary equipment and the lightweight partition wall by computer numerical analysis, and also submits 8 solutions to
install the equipment on the light-weight partition wall. Finally, we verify the structural
behavior of equipment, joints and light-weight partitions by experiment, and offer
solutions to safety support loading in different conditions.
Keywords
Sanitary equipments, Light-weight partition wall, Numerical analysis, Safety, Install
solution
442 1. Introduction
A string of restroom related incidents happened around 1990’s in Taiwan, many
sanitary equipments cracked or fell down due to incorrect use or improper installation
that hurt users, such as shown in Fig.1. After those events happened, many safety
standards were developed for protection by standardizing equipment load bearing,
installation methods, and behavior suggestion, but those standards were established by
equipment installed on concrete walls or brick walls that were the most solid
construction types in Taiwan. The sanitary equipment installed on concrete walls or
brick walls are supported by the shear force of bolts or joints and the friction between
bolts and the wall material. Because of the parcel force between bolts and wall material,
the sanitary equipment can be installed on the wall and bore the loading with stable
conditions, as shown in Fig. 2. The added load of sanitary equipment in toilet is 218.6kg
and in basin is 113.4kg via Chinese National Standard[1] (CNS) to provide the toilet
safety.
Fig.1 Cracked basin hurts user at the society event around 1990’s in Taiwan.
Light-weight partition walls are frequently used in building interiors in these coming
years because of the quick delivery time and light weight. The downside of light-weight
partitions is they have a reduced dead load capacity due to its wall structure. The
structure of light-weight partition is used C-type steel as the frame, covers the fire proof
board in both sides. Fire insulation is installed in the inner light-weight partition wall for
fire protection and sound proofing, shown as Fig.3. As a result of the change in wall
443 structure, the friction between bolts and wall material is reduced, making installation of
sanitary equipment difficult. But the added load also follows CNS for toilet safety.
Fig.2 Parcel forces between bolts and wall material support sanitary equipment
and added loading. [2]
This study tries to analyze the structural relationship between the sanitary equipment
and the light-weight partition wall by computer numerical analysis and experiment
certification to find the best installation solution of sanitary equipment on the lightweight partition wall.
Fig.3 Structure of light-weight partition wall
444 2. Light-weight partitions and bolts
2.1 Light-weight partitions
There are many different light-weight partition systems used in the building plan, and
we classify them in four types; frame system, boards system, blocks system, and others,
as Table 1. The frame system is the popular system used in the site plan via our
investigation, so this study focuses at the frame system in light-weight partition.
Table 1 Light-weight system classification
Types
Frame system
Board system
Block system
Other system
Introduction
The structure is used C-type steel as the frame, covers the fire proof boards.
Overlapping the precast concrete boards with tubular holes as walls.
Overlapping blocks ( brick, stone, etc..) as walls.
Different system above.
The sanitary equipment installed on a frame system in light-weight partitions maybe
supported by the cover boards. The loading support by boards is the important condition
in this study. So we have to investigate the board material characteristic and refer the
references to analyze the stress and strain. Based on our investigation and reference [3],
three kinds of boards are popular in different building type, such as Calcium silicate
board, Gypsum board, and Fiber cement board, and the proportion is also different in
different buildings, as shown in Fig 4.
Non residentail Building
Residentail Building
Fiber cement
board, 10%
Fiber cement
board, 15%
Gypsum
board, 30%
Calcium
silicate
board, 55%
Gypsum board
Gypsum
board, 65%
Calcium silicate board
Fiber cement board
Fig. 4 Popular in different building type and the proportion
445 Calcium
silicate
board, 25%
2.2 Bolts
For solid walls, the sanitary equipment is installed on the solid wall by anchor bolts, and
it provides the support force by parcel force of the wall material. The support force of
one anchor bolt (over D12mm), Fig. 5, installed into concrete wall is normally over
100kg when installed correctly by product testing at the factory. Light-weight partition
walls have a hollow space inside so there is not any parcel force from the wall material.
The hollow anchor bolt was developed to support the equipment on the light-weight
partition wall. The hollow bolts are made of S304 steel, and the normal force supported
is 300kg in D12mm by product testing at the factory, shown as Fig. 6.
Fig. 6 Hollow anchor bolt
Fig. 5 Anchor bolt(used on solid
wall)
The big question by the installation worker, designer and user is, Is the hollow anchor
bolt installed on the light-weight partition wall safe? Therefore, we try to use computer
numerical analysis and experiments to prove the support force between hollow bolts and
cover boards of light-weight partition wall are sufficient.
3. Computer numerical analysis
Calculate the basin support on the wall by its self weight (W) in 20kg and added load
(P) in 113.4kg by CNS, as shown in Fig. 7. The equations are listed as Eq.1 and Eq.2
and the normal force (F) of the calculation is 1938N in one bolt.
V  P W
Eq.1
446 Take moment at A point (Fig. 7)
F
P  L W  X
h
Eq.2
Fig. 7 Structure analysis of basin
Eq. 1 and Eq. 2 are based on the same assessments to calculate the basin support
moment on the wall, as shown in Fig. 7. The equations are listed as Eq.3 and the
moment (M) of the calculation is 1678.3N-m at B point.
447 M  P  L W  X
Eq.3
Fig. 8 Two unit models and boundary conditions for numerical analysis
We make two unit models and control the boundary conditions as a light-weight
partition in the computer numerical analysis program, ANSYS. These two unit models
are only board supported equipment, 30cm distance between two C-type steel columns,
and two boards support the equipment at a 45cm distance that includes one C-type steel
column in the middle, shown in Fig. 8
448 Fig. 9 Two examples in single board and two boards of ANSYS results
The process of ANSYS analysis shows the force distribution, and we can understand the
broken board area and the scale of support force in different materials in 9mm
thickness, as shown in Fig. 9. The support force in Gypsum board is 36kg, in Calcium
silicate board is 70kg, and in Fiber cement board is 120kg, shown in Fig. 10. The results
show that only fiber cement board can support over 113.4kg to satisfy CNS for safety.
So we have to think some component methods or independent structure system to
support the sanitary equipment.
120
Suppor t laoding, kg
120
100
70
80
60
36
40
20
0
Gypsum boar d
Calc ium silic at e boar d
Fiber c ement boar d
Fig. 10 Support force in different board material in 9mm thickness
449 4. Solution and experiment certification
Using hollow anchor bolts to install sanitary equipment on light-weight partition walls
can be done by CNS standards, but only through use of fiber cement board over 9mm
thickness. There are also some install methods developed on site in Taiwan, and we
classify them in two parts. One is called “dry method” and “wet method”. The “dry
method” consists of adding reinforcement members inside the light-weight partition to
provide more support force. The “wet method” consists of grouted light weight
reinforcement concrete (LWRC) inside a unit frame within the light-weight partition.
The LWRC is occurs where the sanitary equipment is located, as shown in Fig. 11.
Concept of dry method
Concept of wet method
Fig.11 Concepts in two part of install methods on site
To find the ideal installation method, we consider 8 solutions that are classified by the
levels of outlook influence. We also consider the following possible solutions; structure
reinforcement, material complex, and equipment structure as listed in Table 2. After
solution submission, we do the experiments to certify the support force in these
solutions.
450 Table 2 Considerations of possible solutions
Influence
Members
Solution methods
A. Hollow anchor bolts
Non impact of
B. Extend installed area
Boards
outlook and structure
C. Unity toilet system
Non impact of structure
Frame bracing
D. Steel sheet fixed in column
E. Reinforcement fixed in column
F. Reinforcement and sheet fixed
Impact of structure
Reinforcement
in column
G. LWRC
Impact of outlook
Equipment structure
H. Plus frame to support equipment
Table 3 List of experiment models and orders
Thickness
of boards
A.
Non-
symmetric
symmetric
9mm Order 1
4
6mm Order 2
3
D.
E.
B.
F.
5
6
7
8
11
G.
13
The experiment was conducted in a big steel frame construction structure that is
simulated as a real house. There are three elevations in this construction, and we design
4m as a set to install three different material boards, as shown in Fig. 12. A total of 10
sets with different solution methods were conducted in this experiment that include 4
sets of hollow anchor bolts to test the symmetry influence of boards installed on the
frame. Therefore, 8 solution methods and 10 model sets were conducted in this
experiment and the order was considered the experiment action area and the boards
were removed and installed, as listed in Table 3. The 8 solution concepts and
experiment processes are listed in Table 4.
451 The big steel frame construction
Three different material boards as one set
Fig. 12 Experiment field and activated conditions
Table 4 Eight solution concepts and experiment process
Solution concepts
Experiment process
Hollow anchor bolts (Support by board)
Almost broken by block shear
Extend area to decrease each bolt loading
Broken by block shear
452 Steel sheet fixed on column
Support by self-drilling screw in column
Reinforcement fixed in column
Bending broken in reinforcement
Solution concepts
Experiment process
Reinforcement and sheet fixed
in column
Support by self-drilling screw fixed
in column
453 Light weight reinforcement concrete
Support by parcel force of the wall
Independent structure system
Support table or feet
Unity toilet system
5. Results and discussion
Based on the experiment results, we can separate the discussion into two parts, one is
the support possibility with boards by hollow anchor bolts, and the other is support
possibility with component methods. We discuss them as follows:
A. Support possibility with boards by hollow anchor bolts
(1) Only fiber cement board can support over 100kg, therefore, sanitary equipment
installed on boards without any component methods is not recommended.
(2) The support loading on the boards installed on the frame with symmetry is 20%
larger then it is on the boards that is installed on the frame with non-symmetry.
454 (3) The support loading on the boards that cross a column is 20% larger then it on
the boards that do not cross any column.
(4) The support loading in the board with 9mm thickness is 60% to 100% larger
than the board with 6mm thickness, depending on the difference in material.
(5) Expanding the installation area can decrease loading on each bolt, but this
method would be broken by block shear, therefore, is not recommended.
B. Support possibilities with component methods
(1) The support loading with a steel sheet fixed on a column is larger then 113.4kg,
that satisfies CNS, but the steel sheet thickness should not be too thick because
of weight, around 1mm to 2mm is ideal.
(2) Reinforcement fixed in a column is easy broken by buckling, and the support
loading is not satisfied CNS 113.4kg.
(3) The support loading with reinforcement and sheet fixed on a column is better
than with reinforcement fixed in column, because of sheet flexure.
(4) Grouted 1:1:3 (weight ratio) light weight reinforcement concrete as mortar into
light-weight partition to fill the hollow part as a solid wall, and it can use the
typical install method to support 300kg over 20 min.
(5) Independent system supports loading by its structure, therefore the support
loading has no relationship with the light-weight partition.
(6) The experiment results of component methods are shown as Table 5, the support
loading in solution is satisfied CNS 113.4kg except reinforcement fixed in
column.
Table 5 Experiment results in component solution methods
Location
Solution methods
Outside
Different
material
Inside
Same
material
Steel sheet fixed in column
Reinforcement
column
fixed
Reinforcement and
fixed in column
Conditions
Support loading
6mm Calcium silicate board
123kg
6mm fiber cement board
134kg
9mm Calcium silicate board
204kg
9mm fiber cement board
216kg
in Column distance in 30cm
97kg
Column distance in 45cm
82kg
sheet Column distance in 30cm
174kg
Column distance in 45cm
214kg
455 Light weight reinforcement
concrete
Water : Cement : Sand
300kg
1: 1:3
Over 20min.
6. Conclusion
The safety of sanitary equipment is a very important issue for the user. Many safety
standards were developed for protection by standardizing equipment load bearing on
concrete walls or brick walls. Some solutions have been developed recently to install
the sanitary equipment on light-weight partitions, but no studies have been done
regarding safety. This study focuses on support loading by different boards with hollow
anchor bolts and component methods recommend by this study. The results show that
only fiber cement board can support over 100kg, therefore, sanitary equipment installed
on boards without any component methods is not recommended. Results of the
component methods are shown as Table 5, the support loading in the solution satisfies
the CNS 113.4kg requirement except for reinforcement fixed in columns, and the
component members have to connect with the frame of light-weight partitions as the
same structure.
7. Reference
4. Chinese National Standard (CNS), 3220-3 r2061-3
5. M.C. Lee, C.L. Cheng, 2001, The Research of Crack Damage in Ceramic Sanitary
in Building, International Symposium of Plumbing System in Asia, 2001.06.22,
Taipei, Taiwan
6. Jui-Hsing Chien, 1998, Study of Evaluation Model of Light-Weight Partition
System in Building Construction, M.C. Thesis, NTUST, Taipei, Taiwan
8. Presentation of Author
Meng-Chieh Lee is a Ph.D in Architecture and an Assistant
Professor at National Taichung Institute of Technology,
Department of Interior Design. His major is water plumbing
system, sanitary equipment and new technology development,
interior environment comfort, and energy saving.
456 VIII.3
A Study on Performance Evaluation for
Toilet Systems with an Uroflowmeter
(1)Yuta Takahashi
(2)Masayuki Otsuka
(3)Hironori Yamasaki
(1) m0843011@kanto-gakuin.ac.jp
Graduate Student, Graduate School of Engineering, Kanto Gakuin University
(2) dmotsuka@kanto-gakuin.ac.jp
Department of Architecture College of Engineering, Kanto Gakuin University,
Dr. Eng
(3) hironori.yamasaki@toto.co.jp
TOTO Ltd.
Abstract
In Japan, along with a rapidly aging society, eating habits are becoming more and
more westernized and illnesses and diseases such as diabetes are on the increase. In
hospitals, the management and diagnosis of patients’ health conditions by the amount
of urine excreted (urine output) and the condition of urine excretion (urine flow)
provide vital indicators. However, after measuring urine outputs and urine flows,
nurses need to discard the measured urine and this creates more burdens for the nurses.
To improve the situation, a toilet system which automatically measures urine outputs
and urine flows was developed. This study aims to examine in an experimental manner
the impact of pipe pressure variation caused by effluence on the accuracy and
reliability of urine output/flow measurement when the toilet system is installed to a
drainage system. To be more precise, this study discusses the above point based on two
field experiments which were carried out using a high-rise experimental drainage
tower and in an actual hospital.
457 Keywords
toilet, urine output, urine flow, drainage load, pressure variation
1.Introduction
In Japan, eating habits are becoming more and more westernized, so are lifestyle-related
diseases and illnesses themselves. Furthermore, along with the arrival of a rapidly
growing aging society, adult diseases and illnesses, such as enlarged prostate, are on the
increase. In addition, in hospitals, not only urinalysis but also the examination of the
amount of urine excreted (urine output) and the condition of urine excretion (urine flow)
play roles as vital indicators in managing and diagnosing patients’ health conditions.
The problem is that discarding urine properly after measuring it for its output and flow
is a burden to nurses. It has also been reported that taking a sample and transporting it is
a burden to patients. In order to improve such circumstances, a toilet system which
automatically measures urine outputs and urine flows (“urine measuring toilet system”
hereafter) was developed and an elementary discussion was made in the previous
paper1).
This study aims to examine in an experimental manner how drainage load-induced pipe
pressure variations may affect the accuracy and reliability of measurement of urine
outputs and urine flows when the urine measuring toilet system is installed to an actual
drainage system. To be more specific, the study is about examining the performance of
the urine measuring toilet system by implementing two experiments; a basic experiment
using a high-rise experimental drainage tower, and a field experiment in an actual
hospital.
2. Overview of the experiments and methods used
2.1 Urine measuring toilet system
Shown in Fig. 1 is the principle of how the urine measuring toilet system works. Initial
measuring water level A (the level of water reserved in the toilet bowl) is set lower than
overflow water level B. As urine is discharged into the toilet bowl, the reserved water
level rises, so does the water level in the measuring pipe comprising the Z pipe and the
lead pipe to form a U shape. The water level sensor detects the rise of water level in the
measuring pipe, and urine flow changes are detected in relation to time variation with
reference to the predetermined analytical curve (output-water level correlation).
458 Overflow water level
Initial measuring water level
Rise of the reserved water level =
rise of the in-pipe water level
Top height of the measuring pipe
This zone is the
measurement range
Measuring pipe
C
B
A
Seal depth
50 mm
Water level sensor
Z pipe
Lead pipe
Drainpipe
Fig. 1 Measurement principle
Overview of the basic experiment
(1)Test drainage stack system
The experiment was carried out using the environmental construction simulation tower
at Kanto Gakuin University. Fig. 2 shows two drainage systems which were used for the
experiment. One is a high-rise drainage stack system (“drainage stack [1]” hereafter),
which is connected to the house drain (3200[mm] apart between the drainage stack [1]
and the house drain). The other drainage stack system (“drainage stack [2]” hereafter) is
also connected to the house drain, thus enabling a combined flow to the house drain
from both drainage stacks.
459 Bell mouth
Sensor
Item to measure
W: heat-wire
anemometer
Top vent end with W
P: pressure
sensor
Horizontal fixture branch on
each floor with P (small type)
Drainage stack [1]
W
P
125A×125A
LT: fitting for joint section
125A×125A
P
Drainage load flow rate [L/s] P
Drainage stack [1]
Drainage stack [2]
0.5
0.5
1.0
1.0
1.5
1.5
2.0
2.0
2.5
2.5
P
7F
6F
5F
Drainage stack [2]
100A
75A
3F
3F
3000
P
2F
600
3250
IC
LL
Primary house drain
330
0
rain
se d
Hou
Urine measuring
toilet system
IC
P
LT
00
115 r125A
te
e
m
d ia
LL
1F
GL
0
320
50
t1/1
dien
Gra
Fig. 2 Test drainage Stack
systems
460 3000
2F
3750
3000
P
3000
Drainage load
600
4F
3000
P
4F
1F
GL
25200
LL: long elbow
8F
75A
3000
100A×125A
P
3000
IC: increaser
Drainage load
3000
Pipe diameter
9F
3000
Pipe component
100A
Unit[㎜]
To implement the experiment, a nominal diameter of 100A was applied to both drainage
stacks and 125A to the house drain, plus the gradient of the house drain was set to
1/150. Two urine measuring toilet systems were also installed; one on the 7th floor,
which was the floor beneath where drainage was applied, and where the pipe pressure
was within the negative range, and the other one on the 2nd floor, which was a lower
floor of the experimental system where combined flows from both drainage stacks most
likely generated positive pressure.
(2) Method of applying drainage load
Drainage load was applied consistent with SHASE-S 2182), and constant flow load of
0.5 - 2.5[L/s] was applied from the top floors of drainage stacks [1] and [2]; the 9th floor
and the 4th floor respectively. First, drainage load was applied from drainage stack [1],
and when the drainage flow became steady, further drainage load was applied from
drainage stack [2]. Fig. 3 exemplifies how the pipe pressure fluctuates when drainage
load is applied as described above. This is a variation of pressure recorded on the 2nd
floor when drainage load of 2.5[L/s] was applied from both of drainage stacks [1] and
[2]. Incidentally, artificial urine was applied, using the method described in 2.4, at point
(1) and (2) in the graph where the pipe pressure fluctuates significantly.
(3) Items to measure and measuring methods
The variation of vent pipe velocity W and the pressure variation P of each horizontal
fixture branch, which is connected to the drainage stack system, were measured. A
criterion value of ±400[Pa] or below (pipe pressure variation) was set, and
measurements were made with reference to this criterion value. Urine outputs and flows
were measured and analytical waveforms and urine output W, plus the maximum urine
flow qmax were used to confirm measurement accuracy.
Combined drainage from systems [1] and [2] Pip e p ressu re (2 F) [Pa]
400
200
Drainage from system [1]
0
-200
(1)Artificial urination
(2)Artificial urination -400
0
20
40
60
80
100 120 140
Measurement time [s]
160
180
200
Fig. 3 Pipe pressure variation (in the case of 2.5[L/s] and
461 2.5[L/s])
2.2 Over view of the field experiment
(1)Test drainage stack system
The experiment was carried out using a 19-storey high-rise hospital building in Tokyo.
Shown in Fig. 4 is the drainage stack system which was used for the experiment. This
drainage system employs a two-pipe system using a diameter of 150A for the drainage
stack and 125A for the vent stack. The horizontal vent branch pipe with a diameter of
65A is drawn from the horizontal fixture branch, and its end is connected to the vent
stack. Two tests were carried out using multiple sanitary fixtures connected to the
horizontal fixture branch; 1) horizontal fixture branch combined drainage test and 2)
drainage stack load test.
Vent stack125A
3900
Vent pipe 150A
3900
PH
150A
Horizontal vent
branch pipe 65A
3900
19F
18F
3900
Fig. 5 Horizontal
fixture branch 100A
Drainage stack
76200
3900
17F
3900
10F
9F
3900
Loop vent pipe 65A
8F
4500
Drainage stack offs 200A
To the vent pipes of lower floors
Unit[mm]
7F
th
* The drainage stack on the 7 floor and below is offsetted on nd
the 2 floor, connected to the house drain (200A), and st
connected to the catch basin outside from the 1 floor. Fig. 4 Drainage stack system
462 (2)Method of applying drainage load
In the field experiment, drainage load was applied from the 18th floor, the plan view of
which is shown in Fig. 5, and drainage patterns, as shown in Table 1, were used.
Drainage intervals among the sanitary fixtures were determined in such a way that pipe
water level variations would overlap at the measurement position (H1 in Fig. 5), and
combined drainage was created.
Table 1 Drainage
D rainage
P attern
1
No.1
No.2
No.3
No.4
No.5
No.6**
W ater C loset (W C )
2 3 4 5 6 7
○
○
○ ○ ○ ○
○ ○ ○ ○
○
○
○
○
○
8
○
○
○
○
○
○ ○ ○
○ ○ ○
1
2
Labatory (L)
3 4 5 6
W ashbasin (W B )
7 8 1 2 3 4 5
○ ○
○ ○ ○
○
○ ○
○
○ ○ ○ ○ ○
○ ○ ○ ○ ○
○ ○ ○ ○ ○
* Sanitary fixture nos. refer to Fig. th
th
**No.6: when drained from 17 and 18 H : Water level 、P : Pipe pressure
Urine measuring toilet system
Horizontal vent branch pipe 65A
WC3P2
WB1
Drainage stack
H1
WC4 P3
L1
L2
L3
L4
40A
40A
L5
L6
WC7 P4
L7
WC1P1
WC2
WB2 P5
WB3
2600
WC5
L8
A
WC6P6
WB4
WB5
22000
25400
Fig. 5 Layout of the horizontal fixture branch system
463 WC8
H2
Vent stack
800
Horizontal fixture branch 100A
Gradient 1/100
Unit [mm]
(3) Items to measure and measuring methods
In test 1), pipe water level variations on one end of the horizontal fixture branch (H1 and
H2) and the pipe pressure of the urine measuring toilet system (P1 - P4) were measured.
Artificial urine was discharged into four toilets; WC1, WC3, WC4, and WC7, as shown
in Fig. 5. The pipe pressure variation at the time of draining from the fixtures was
measured in advance, and artificial urine was applied when the variation value was at
the maximum or minimum (Pmax・Pmin).
In test 2), drainage load was applied from the 18th floor, and measurements were made
along the horizontal fixture branch, which was installed on the 17th floor where a
negative pressure zone was created right beneath the drainage-applied floor. Artificial
urine was discharged into two toilets; WC1 and WC3 (P1 and P2). Also, drainage load
was applied from the 17th floor and the 18th floor, and pipe pressure variations were
measured along the horizontal fixture branches on the 9th floor and the 10th floor, where
a great positive pressure was predicted to occur just before the offset plumbing, as in
Fig. 4. The toilets into which artificial urine was discharged were WC6 on the 9th floor
and WC1 and WC4 on the 10th floor. In applying artificial urine outputs, three urination
patterns were used, as described in 2.4; typical outputs for normal and enlarged prostate
conditions: 200mL and 400mL respectively, and 50mL with neurogenic bladder
condition.
2.3 Method of applying artificial urine
As shown in Fig. 6, waveforms of actual patients’ urine outputs and urine flows can be
grouped into three urination patterns; (1) normal, as the standard, and according to the
clinical conditions; (2) enlarged prostate and (3) neurogenic bladder. This classification
was also used in this study when clean water was put into the toilet bowl according to
urination patterns (1), (2) and (3) and the possibility of accurate measurement was
examined.
In tests (1) and (2), four different volumes of urine excreted; 200mL, 400mL, 600mL and
800mL, were used, and to simulate the flow of urine, clean water was first saved in a
container with a pluggable hole (Fig. 7) and then released at a constant rate as the plug
was removed. In test (3), the total volume of urine excreted (clean water); 50mL, was
measured and applied in small amounts using a dropper.
464 U rine flow
U rine output
250
50
200
Urine flow [mL/s]
40
35
qmax
30
25
150
W
20
100
qmax W
15
qmax 10
50
Urine output [mL] a
45
W 5
0
0
0
10 20 30 40 50
T ime [s]
(1) Normal
0
10 20 30 40 50 0
T ime [s]
(2) Enlarged prostate
10 20 30 40 50
T ime [s]
(3) Neurogenic bladder Fig. 6 Artificial urine application patterns
※ Reference: Urological Diagnostics for Routine Care, InterMedical Co., Ltd. Container with a hole
Support
Artificial urine
(2) Cross‐section view
(1) Outline view
Fig. 7 Method of applying artificial urine
3. Experiments results and discussion
3.1 Results of the basic experiment
(1)Pipe pressure distributions and load-induced analytical waveforms
Fig. 8 exemplifies pipe pressure distributions by floor according to the application of
either single drainage or combined drainage. When a single drainage load of 2.5[L/s]
was applied, the minimum pressure Psmin was -438[Pa] on the 7th floor where the urine
measuring toilet system was installed, and when a combined drainage load of 2.5[L/s] +
465 2.5[L/s] was applied, the maximum pressure Psmax was about 345[Pa] on the 2nd floor.
Artificial urine was applied based on these pressure distributions.
Drainage flow[L/s] 1.0
Symbol
1.5
2.0
2.5
Height Floor [m] [F] 25 Drainage flow[L/s] 1.0
Symbol
Height Floor [m] [F] 9F 25 9F 8F toilet system
toilet system
20 7F 7F 6F 15 345.3Pa Urine measuring 5F 10 toilet system
6F Urine measuring 5F 10 4F 5
2.5
Urine measuring 20 15 2.0
‐438.1Pa
Urine measuring 8F 1.5
toilet system
4F -400
-200
0
200
400
P ipe pressure [P a]
600
5
-600
-400
-200
0
200
400
P ipe pressure [P a]
(1) Max. pressure ([1]+[2])
(2) Min. pressure ([1] only) Fig. 8 Pipe pressure distributions
Shown in Fig. 9 are urine flow and output waveforms by urination pattern which were
recorded when there was no drainage load and when variations occurred to the pipe
pressure. As shown in the graphs, when variations occur to the pipe pressure, the
maximum-minimum pressure of both urine flow and output waveforms almost reach
±200Pa and ±400Pa. Comparing to when there was no drainage load, urine flow
waveforms change drastically when the pipe pressure was 400Pa, especially, with the
“normal” and “enlarged prostate” application urination patterns. This suggests that the
positive pressure which was generated in the pipe disturbed the surface of seal water in
the toilet bowl significantly, thus distorting the urine flow waveforms. In contrast, the
pipe pressure of around -400Pa does not cause much distortion to the urine flow
waveforms, and the waveforms become smaller in comparison to when there is no
drainage load. This suggests that the pipe pressure becomes negative and draws the seal
water in the toilet bowl into the drainage pipe, thus making measurements smaller than
when there is no drainage load.
Furthermore, the adequacy of measurement accuracy has already confirmed in the
previous paper, at a pipe pressure of ±200Pa, and although, in this study, some
measurement errors were found with the urine flow waveforms, the urine output
waveforms were measured with few errors compared to when there was no load applied.
466 Hence, the results found in this study using each urination pattern demonstrate the
results acquired in the previous study.
400
W
30
300
20
200
10
100
0
0
0
10
20
30
T ime [s]
40
A pprox.-400P a
A pprox.-200P a
50
Urine output [mL] 500
qmax 40
N o load
A pprox.200P a
Urine flow [mL/s]
50
A pprox.400P a
Urine output [mL] Urine flow [mL/s]
N o load
500
qmax 40
400
30
300
W 20
200
10
100
0
0
0
50
10
20
30
T ime [s]
40
50
W
20
200
10
100
0
0
10
20
10
30
T ime [s]
40
80
60
4
40
2
20
0
0
10
20
30
T ime [s]
40
300
200
10
100
0
0
10
20
30
T ime [s]
40
50
10
100
8
80
W 6
60
qmax 4
40
2
20
0
0
0
50
W qmax 20
0
100
6
0
400
30
Enlarged qmax W
8
500
40
50
Urine output [mL] 300
50
10
20
30
T ime [s]
40
Urine output [mL] 30
Urine flow [mL/s]
400
qmax Urine flow [mL/s]
40
0
Urine flow [mL/s]
Urine output [mL] 500
Urine output [mL] Urine flow [mL/s]
Normal 50
50
Neurogenic (2) Negative pressure (single (1) Positive pressure (combined Fig. 9 Urinalysis waveforms by
* Urine volume: 200mL (“normal” and “enlarged (2) Measurement results of pipe pressure and urine volume W
Fig. 10 shows the minimum pipe pressure Pmin measured on the 7th floor where the
urine measuring toilet system was installed, the maximum pipe pressure Pmax
measured on the 2nd floor, and urine output W measured when different amounts of
artificial urine was applied. The results are shown by urine urination pattern; (1)
normal, (2) enlarged prostate, and (3) neurogenic bladder. As in 3.1, the adequacy of
measurement accuracy was ensured by the results when the normal and enlarged
prostate urination patterns were applied and the negative pressure was around -200[Pa].
As for the positive pressure, it was measured fine within an acceptable error range
between applied urine outputs and measured pressure values when the normal and
enlarged prostate urination patterns were applied. When the neurogenic bladder
urination pattern was applied, however, errors between applied urine outputs and
467 measured values of both positive and negative pressure were significant, recording the
maximum error of 25[mL].
Urine output [mL]
1000
200mL
400mL
800mL
600
400
Errors
200
0
-600
-400
-200
1000
200mL
0
200
Pipe pressure [Pa]
400
600
Pmax(2F)
Pmin(7F) Urine output [mL]
600mL
800
(1) Normal
400mL
600mL
800mL
800
600
Errors
400
200
0
-600
-400
Pmin(7F) -200
0
200
Pipe pressure [Pa]
400
Pmax(2F)
600
100
Pmax(2F)
200
Urine output [mL]
(2) Enlarged prostate
100
50mL
75
Errors
50
25
0
-200
Errors
-100
Pmin(7F) 0
Pipe pressure [Pa]
(3) Neurogenic bladder
Fig. 10 Max./min. pressure in relation to urine output
measurement
When the neurogenic bladder urination pattern is applied, the measurement accuracy
with urine output W is not sufficient, and the reason for that is thought to be because a
measured waveform is digitally filtered to remove all noise information except a
specified frequency and is adjusted with a moving average, and this process decreases
the measurement accuracy, in principle, when the neurogenic bladder urination pattern
is applied in small amounts at short intervals.
468 (3) Measurement results of the pipe pressure and the maximum urine flow qmax
Similarly to Fig. 10, Fig. 11 shows the maximum urine flow qmax in relation to the
maximum/minimum pipe pressure (Pmax/Pmin) by artificial urination pattern. When
the normal and enlarged prostate urination patterns were applied, the maximum urine
flow was measured, showing values which were more or less consistent below ±200Pa.
When the neurogenic bladder urination pattern was applied, measured values became
irregular, and this is very similar to what was recorded with the urine output
measurement. As for the urine flow variation, however, it can be referred to Fig. 9 in
which the measurement results of positive/negative pipe pressure are shown, and the
combined use of the measurement results of urine outputs and urine flows would enable
the determination of patients.
Maximum urine flow
qmax [mL/s]
80
60
40
20
200mL
0
-600
-400
400mL
-200
Pmin(7F)
600mL
800mL
0
200
Pipe pressure [Pa]
400
600
Pmax(2F) (1) Normal Maximum urine flow
qmax [mL/s]
80
200mL
400mL
600mL
800mL
60
40
20
0
-600
-400
-200
0
200
Pipe pressure [Pa]
400
Pmin(7F)
600
Pmax(2F) Maximum urine flow
qmax [mL/s]
(2) Enlarged prostate
20
50mL
15
10
5
0
-200
-100
Pmin(7F)
0
Pipe pressure [Pa]
100
200
Pmax(2F) (3) Neurogenic bladder Fig. 11 Max./min. pressure in relation to maximum
urine flows
469 3.2
Results of the field experiment
(1) Pipe pressure
1) Horizontal fixture branch combined drainage test
Shown in Fig. 12 are the maximum and minimum values (Pmax and Pmin) of pipe
pressure variation which was measured at P2 (Fig. 5) along the pipe which was
connected to the urine measuring toilet system. When drainage load was applied from
multiple sanitary fixtures (No.3 and No.4), No.3 recorded +90[Pa] and No.4 recorded
+290[Pa], both falling within the criterion value of ±400[Pa].
2) Drainage stack load test
Fig. 13 shows the distributions of Pmax and Pmin in the No. 5 case; drainage load was
applied from the 18th floor, and in the No.6 case; drainage load was applied from the
17th floor as well as from the 18th floor. In the No.5 case, on the 17th floor, which is the
floor right beneath where drainage load is applied, Pmax and Pmin values fall below
±100Pa. Also, it was confirmed that in the No.6 case, Pmax and Pmin values measured
on the 9th and 10th floors still remained below ±200Pa even when drainage load was
applied from the 17th and 18th floors.
Pipe pressure [Pa]
800
600
400
SHASE‐S 218 Line
200
0
-200
-400
No.1
No.2
Fig. 12
No.3
Max./min. pipe pressure (Pmin/Pmax) by drainage
pattern * Horizontal axes: drainage pattern nos. in Table 1
470 No.4
Pmin (No.5)
Pmin (No.6)
Pmax (No.5)
Pmax (No.6)
Floor
19
17
Drainage
15
load
13
11
9
7
-250
-200
-150
-100
-50
0
50
Pipe pressure[Pa]
100
150
200
250
Fig. 13 Stack pressure distributions (No.5 and No.6)
(2)
Load-induced analytical waveforms and measurement errors
Fig. 14 shows load-induced analytical waveforms in the no-load case and the No.9 case
using the normal urination pattern (200mL). The urine outputs W and the urine flow
waveforms are very similar in both cases, and the measurement accuracy was therefore
confirmed under this condition.
50
500
N o Load-U rine flow
400
N o.9-U rine flow
Urine output [mL]
Urine flow [mL/s]
40
N o Load-U rine output
30
300
N o.9-U rine output
20
200
10
100
0
0
0
10
20
T ime [s]
30
40
50
Fig. 14 Load-induced analytical waveforms – normal 200mL (No.9)
Moreover, Fig. 15 shows all urine outputs W and maximum urine flows qmax that were
applied and measurement errors which were found in the field experiment. An error
tolerance line of ±10% is drawn in each graph to check the measurement accuracy. The
471 urine outputs W fall within the range of ±10%, but urine flows qmax are irregular,
creating the maximum error of approx. 10[mL/s].
Urine ou tp ut W (m easured ) [m L]
M ax Urine flow qmax (m easured) [m L/s]
±10% error 500
Normal
Enlarged prostate
400
Neurogenic bladder
300
200
100
±10% error 50
Normal
Enlarged prostate
40
Neurogenic bladder
30
20
10
0
0
0
100
200
300
400
Urine output W (actual) [mL]
0
500
10
20
30
40
50
Max Urine flow qmax (actual) [mL/s]
(2) Max. urine flow qmax (1) Urine output W Fig. 15 Actual artificial urine amounts applied and measurement errors (field
experiment)
4. Conclusion
Through the basic experiment implemented on the high-rise experimental drainage
tower and the field experiment carried out in an actual hospital, the performance of the
urine measuring toilet system was examined and the following findings were
established:
(1) Urine flows and outputs were measured by applying artificial urine in three different
urination patterns, which simulates actual patients’ urination patterns; normal,
enlarged prostate, and neurogenic bladder. The results ensure that the measurement
accuracy of the urine measuring toilet system is more than adequate for measuring
urine flows and outputs with “normal” and “enlarged prostate” conditions, providing
that the pipe pressure variation is within ±200Pa.
(2) The urine measuring toilet system is used on drainage systems with two-pipe (loop
ventilation) system in hospitals and other facilities, and it has been confirmed that the
measurement accuracy of the urine measuring toilet system is more than adequate
when the allowable pressure difference is within ±100Pa, which is stipulated by
SHASE-S 206.
472 References
1.Yuichi FURUTA, Study on the Performance Evaluation of the Water-closet system
equipped with Uroflowmetry analyzer,AIJ J. Technol. Des. Vol. 14, No.27, 187192, Jun., 2008.
2.Yuichi FURUTA, Study on the Performance Evaluation of the Water-closet system
equipped with Uroflowmetry analyzer.~Confirmation of the performance with
uroflowmetry, under the pipe pressure fluctuations which is given by actual drainage
load.~ Technical Papers of Annual Meeting The Society of Heating, AirConditioning and Sanitary Engineers of Japan 2008 pp.21~24
3.Testing Method of Flow Capacity for Drainage System in Apartment Houses,
SHASE-S218- 2008
4.SHASE-S206 “Standard for Plumbing and Sanitary System” (2000)
Presentation of Authors
Yuta Takahashi is a graduate student of the Otsuka laboratory, Kanto
Gakuin University. He is a member of AIJ(Architectural Institute of
Japan)and SHASE(Society of Heating, Air-Conditioning and
Sanitary Engineers of Japan. His current research interests are the
drainage performance of drainage stack systems with consideration of
combined drainage in the house drain and the symbiosis housing.
Masayuki Otsuka is a professor at the Department of Architecture,
Kanto Gakuin University. He is a member of AIJ(Architectural
Institute of Japan)and SHASE(Society of Heating, AirConditioning and Sanitary Engineers of Japan, academic
director). His current research interests are the performances of
plumbing systems, drainage systems with drainage piping systems for
SI(Support and Infill)housing.
Hironori Yamasaki is a Senior Engineer, Restroom Producs
Research Dept., Sanitary Ware Product Research Sect., TOTO
LTD.. He is a member of NBS (Neurogenic Bladder Society),
JSME (Japan Society of Mecanical Engineering) , AIJ
(Architectural Institute of Japan) and SHASE (Society of Heating,
Air-Conditioning and Sanitary Engineers of Japan). His current
research interest is the Physical Monitoring for Medical and Health
Care.
473 VIII.4
The possibility of wet sludge utilization for
greenhouse effect gas emission reduction.
Yoshiharu Asano
yasanok@shinshu-u.ac.jp
The department of Architecture, faculty of engineering, Shinshu University.
Nagano, Japan.
Keywords
CO2 emission; wet biomass; methane fermentation; categorization of biomass
Abstract
SHASE (The Society of Air-Conditioning and Sanitary Engineers of Japan) set up the
committee in 2007 for investigating committee the utilization and application of wet
sludge”, and decided to arrange the various utilizations of wet sludge. The wet sludge is
recognized as a biomass resource in Japan. There is a lot of sludge utilization trial,
which is not so popular. The sewage sludge, farm animal excrement treated sludge, and
leftover/dead stock foods have not been utilized enough for the civic life. The
committee collected the case samples of sludge utilization, and categorized by field and
progress level of them.
Industrials, which generated the sludge, were classified into three fields, those are the
livestock business including fish processing, sewage/garbage exhausted in a city life,
and dead stock foods. The utilizations of sludge are classified into three patterns, which
are experimental phase, demonstration phase and practical phase. Methane fermentation
process is particular in each case.
When building a manufacturing plant, in which the sludge is generated, the utilization
facility of fermentation gas will be designed, for examples, thermal application and
474 electric power generation. The new buildings in the city should be set for the utilization
facilities of fermentation gas changed from self rejected sludge and garbage. These
plans will serve a useful purpose of greenhouse effect gas emission reduction.
1. The purpose and background of research
A few organizations in Japan are working on environmental protection based on the
Kyoto Protocol, and biomass was identified as a new energy source, which is part of a
comprehensive strategy in formulating a basic plan for promoting the formation of a
recycling-oriented society. When thinking about global warming, a crucial technology
for reducing carbon dioxide emissions is necessary, which is to make the best use of
biomass with characteristic carbon-free.
Despite the wet biomass feedstock originated livestock excreta, sewage sludge and food
scrap are plentiful, utilization are still expected to expand very quickly because of
carbon-free technology. In this paper case examples should be investigated and
classified as the basis on which to consider the possibility of the introduction of plant
planning, constructions and operations.
2. Classification
2.1. The classification method
Table 1, 2 & 3 show the results of classification. A kind of industry and the emission
source of wet biomass are the first group, and the treatment process of wet biomass and
attainment level of it are the second group.
Biomass is categorized into 2 groups, one of which is with high humidity and water
content, and other is dry and with low moisture content. This study deals the wet
biomass with high moisture content.
Biomass feedstock is dotted at various places. In this study wet biomass, which is
livestock excrement and sewage sludge, garbage, food waste, industrial wastewater are
separated for sources generated in each industry. Industry are categorized the food
industry and the life system, the agriculture/livestock systems.
There are many treatment objects or installations, which are for carbonized matters and
methane fermentation, composting, esterification, ethanol fuel manufacture, methanol
fuel manufacture.
Progress is classified as 3 levels. The first is experimental phase of the project level in
laboratory. The second is the empirical level, some of which are under subside of
475 NEDO (New Energy and Industrial Technology Development Organization of Japan) or
MAFF (Japan Ministry of Agriculture, Forestry and Fisheries). The third is the practical
application level.
2.2. The result of classification
There are a lot of utilization of progressive technologies and facilities of treatment for
the biomass derived by the raw materials from the animal excreta, food waste and
sewage sludge. However, there is not much utilization of the biomass derived by the
raw materials from fishery processing residuum, septic tank of human waste and sewage
sludge of agricultural community.
In particular, there are more technologies of the treatment the biomass derived by the
raw material from the waste of food processing technology than from another organic
matters.
3. Investigation of EWH
3.1. Outline of Investigation
12 plants were selected as typical examples of the practical level or the empirical level
based on Table 1, 2 & 3. The Contents were (1) business and object of treatment, (2)
planning of process plan and process flow, (3) capacity of processing facilities, (4)
actual performance of process, (5) the amount of carbon dioxide reduction. This data
was collected by a questionnaire, the website of each facility and the public information
sheet of offices.
3.2. The results of the survey
Examples of the facility to process sewage sludge biomass are shown in Table 4. The
other examples of treatment facilities for wet biomass are shown in Table 5 . In general,
wet biomass, has the characteristics of being carbon neutral. Then, the amount of carbon
dioxide reduction in each case is shown in Table 4 & 5. Ten of the 12 facilities were for
the methane fermentation. Others are the facilities that sewage sludge is carbonized, and
the facilities that manufacture the food waste as raw material feed.
By the methane fermentation, in addition to methane gas, composting fertilizer or liquid
fertilizer is manufacturing, and finally residue is occurred after the fermentation. The
amount of methane emissions and composting emissions and residue generated after
fermentation are different depending on the type of biomass feedstock. The methane gas
generated is used as fuel for power generation and boiler. The heat generated is used for
heating the methane fermentation tank in the facility and used as air conditioning and
476 hot water in the office. Methane gas is part of the city gas supply or used as vehicle fuel.
In this case, methane gas is required for desulfurization and purification. Also, some of
the methane gas generated is used for combustion.
4. Consideration
It is clear that methane gas after fermentation of wet biomass was often used. Future
challenges for effective utilization that the biogas could be burned with surplus, are
necessary to consider the reduction of volume of the fermentation residue.
It is possible to lose weight of waste to be discharged after treatment of wet biomass. By
using biogas as fuel, building users have the advantage of reducing carbon dioxide
emissions. From the results of this survey, the capacity of each facility will have been
understood in order to calculate the amount of carbon dioxide reduction. It is important
to introduce a facility for methane fermentation on site of building.
5. References
(1) SHASE: http://www.env.go.jp/
(2) NEDO: http://www.chuden.co.jp/
(3) MAFF: http://www.nafgano-toshi-gas.co.jp/
6. Presentation of Author
Yoshiharu Asano is a professor of Shinshu University, and department dean of architecture. He
teaches and conducts research in the department of architecture and building engineering. He
specializes in building equipment, water supply and drainage in buildings and specially
controlled installations.
477 Table 1-a Classification of wet sludge from agriculture/livestock industry
industry
objects
treatment
experimental level (contents)
carbonization
micro cogeneration system(2007)
agriculture/
livestock
business/
fisheries
industry
methane fermentation
livestock
excreta
on-the-premise methane fermentation
equipment(2007)
composting
fodder production
carbonization
seafood
processing
methane fermentation
low temperature gasification by
chlorella(-)
gasification
industry
objects
treatment
empirical level
practical level
carbonization
livestock
excreta
agriculture/livestock
business/fisheries
industry
methane fermentation
composting
fodder production
carbonization
seafood
methane fermentation
processing
gasification
Table 1-b Classification of wet sludge from agriculture
478 No.5
Table 2-a Classification of wet sludge from life system
industry
objects
treatment
experimental level (contents)
carbonization
low pressure wet condition oxidation system(2002)
power generation by micro gas turbine(2007)
methane fermentation
phosphoric acid fuel cell by using sewage digestion gas(2006)
hydrothermal gasification(2002)
gasification power plant by internal circulating fluidized-bed
type gasifier(2007)
gasification
hydrogen fermentation by anerobic microflora(2005)
sewage sludge
mechanism of hydrogen generation by bacterial
fermentation(2005)
fluidify
composting
alkaline catalyzed oil reaction
pretreatment technology for processing sewage sludge into fuel
for steam heating(2006)
ketone reaction of water soluble organic(2005)
fodder production
other
life system
carbonization
residential
septic tank
methane fermentation
composting
carbonization
agricultural
commune
methane fermentation
composting
carbonization
hydrogen-methane two-stage fermentation(2005)
phosphoric acid fuel cell and generation system(2005)
methane fermentation
food scraps
hydrogen fermentation by anerobic microflora(2005)
thermofilic methane fermentation (2005)
composting
fodder production
479 Table 2-b Classification of wet sludge from life system
industry
objects
treatment
empirical level
carbonization
No.1
methane fermentation
sewage
sludge
practical level
No.2, No.3,
No.4
gasification
fluidify
composting
fodder production
other
life
system
carbonization
residential
septic tank
methane fermentation
composting
carbonization
agricultural
methane fermentation
commune
composting
carbonization
methane fermentation
food scraps
composting
fodder production
480 No.6, No.7
Table 3-a Classification of wet sludge from food industry
industry
objects
treatment
experimental level (contents)
carbonization
methane fermentation
foods
industrial
drainage
hydrogen fermentation and methane fermentation
combination(2005)
glucose - air battery(2005)
fluidify
composting
fodder production
carbonization
dewatering technology in oil(2006)
hydrogen fermentation and methane fermentation
combination(2005)
methane fermentation
hydrogen generation by photosynthetic
pigment(2005)
biogas cogeneration system(2005)
foods
industry
gasification
gasification process by supercritical water(2003)
composting
utilization of gas by composting
fodder production
alcohol production by genetically-modified
bacteria(2002)
leftover food
ethanol production
acid digestion(2007)
glycation(2005)
bio-ethanol production by yeast(2005)
methane - methanol conversion technology(2005)
methanol production
bio battery by bacteria(2005)
alcohol production
enzymatic saccharification process(202)
acetone-butanol fermentation(2005)
esterification diesel fuel
esterification of vegetable oil(2005)
481 Table 3-b Classification of wet sludge from food industry
industry
objects
treatment
empirical level
practical level
carbonization
methane fermentation
foods
industrial fluidify
drainage
composting
No.8
fodder production
carbonization
methane fermentation
foods
industry
No.10, No.11
No.9
gasification
composting
leftover
food
fodder production
ethanol production
methanol production
alcohol production
esterification diesel fuel
482 No.12
Table 4 Examples of the facility to process of sewage sludge
inlet flow
(m3/y)
location
No.1
Tobu sludge
plant
Tokyo
metropolitan
No.2
Nagaoka chuo
sewage plant
Niigata pref.
33,032,360
61,767
87,884 (m3/y)
(digester chamber)
No.3
Higasinada
sewage plant
Kobe city
57,673,291
157,577
317170 (m3/y)
(digester chamber)
No.4
Morigasaki
sewage plant
Tokyo
metropolitan
444,099,160
1,216,710
_
treatment
amount of
emergence
usage
amount of
used energy
CO2 reduction (tCO2/y)
carbonization
charcoal 7,700
(t/y)
fuel of
thermal
power plant
64,476,720
(MJ)
37,000
city gas
496,000
(m3/y)
18,451,200
(MJ)
1,300
1,200
4,772
No.1
No.2
methane
fermentation
methane gas
1,672,260(m3/y)
input
7,486,200
20,450
99,000 (t/y) (carbide
(dehydration (dehydration
furnace)
sludge)
sludge)
No.3
methane
fermentation
methane gas
3,22,418 (m3/y)
vehicle fuel
99,981
(m3/y)
3,719,293
(MJ)
No.4
methane
fermentation
methane gas
12,980,000
(m3/y)
fuel of
power plant
21,180
(kWh/y)
76,248,000
(MJ)
483 through-put
(m3/d)
plant
Table 5-a Examples of the facility to process of wet sludge
treatment
methane gas
production
methane
fermentation
160-200
(m3/d)
methane
fermentation
1,152,821
(m3/d)
methane
fermentation
228,599
(m3/y)
50 (m3/d)
methane
fermentation
500-700
(m3/d)
amount of
energy used
compost
fertilizer
production
liquid fertilizer
production
CO2 reduction
(t-CO2/y)
single generator
(37kW)
150-170
(kWh/d)
2 (t/d)
9 (t/d)
34.8
No.6
fuel of boiler/ cogeneration system
boiler
(40.4Nm3/d)
generator
(1,700kW)
1,682,991
(kWh/y)
166 (t/y)
202 (t/y)
_
No.7
fuel of cogeneration system
double generators
(200kW)
40,776
(kWh/y)
243 (t/y)
977 (t/y)
25.8
_
10 (t/y)
0
19-270
(design value)
source
amount of
waste
capacity of
treatment
livestock excreta
13 (t/d)
13 (t/d)
garbage
0.2 (t/d)
1 (t/d)
garbage and
livestock excreta
16,477 (t/d)
80 (t/d)
livestock excreta
14,384 (t/y)
74 (t/d)
garbage
786 (t/y)
3 (t/d)
Soybean plant
soy broth
_
usage
facilities for
methane gas
No.5
fuel of power
generation
plant
No.5
Kuzumaki
biomass plant
No.6
Hida city biomass
center
No.7
Yamaga city
biomass plant
No.8
boiler-1
(31Nm3/h)
No.8
fuel of boiler
boiler-2
(12Nm3/h)
484 Table 5-b Examples of the facility to process of wet sludge
treatment
methane
gas
production
160 (t/d)
methane
fermentation
1,300,00
(m3/y)
19.6 (t/d)
methane
fermentation
494,040
(m3/y)
(design
value)
578,400
(m3/y)
(design
value)
plant
source
No.9
Sludge
treatment
plant
sludge
14,400
(t/y)
No.10
Soft drink
manufacturing
residuum of
coffee and
tea
8,234 (t/y)
No.11
Distilled
spirit
production
factory
strained less
11,940
(m3/y)
(design
value)
40 (t/d)
methane
fermentation
No.12
Feed
production
leftovers of
school lunch
317 (t/y)
1 (t/d)
change of
animal feed
_
usage
Facilities for
methane gas
amount of
energy
used
compost
fertilizer
production
liquid
fertilizer
production
CO2
reduction
(t-CO2/y)
No.9
fuel of
generation
4 generators
(1,900 kW)
2,000,000
(kWh/y)
3,500 (t/y)
1,300 (t/y)
1,100
No.10
fuel of
boiler and
generation
generator
(6 kW)
boilers
(1,400 kg/h)
1,410,000
(MJ/y)
1,800 (t/y)
1,100
No.11
fuel of
boiler
boiler
(750 kg/h)
_
_
_
_
No.12
_
_
_
_
245 (t/y)
_
485 capacity of
treatment
amount of
waste
VIII.5
Formulation of a synthetic greywater to
evaluate the performances of on-site
greywater recycling technologies
Fanny Hourlier1,2 (fanny.hourlier@cstb.fr), Anthony Massé2, Pascal Jaouen2, Abdel
Lakel1, Claire Gérente3, Catherine Faur4, Pierre Le Cloirec5,6
1
CSTB, 11 rue Henri Picherit, BP 82341, 44323 Nantes Cedex 03, France
2
GEPEA UMR CNRS 6144, CRTT, 37 boulevard de l'Université, BP 406, 44602 SaintNazaire Cedex, France
3
EMN, GEPEA UMR CNRS 6144, 4, rue Alfred Kastler, BP 20722, 44307 Nantes
Cedex 3, France
4
Université de Montpellier 2, CIRAD UMR CNRS 016 Génie des Procédés Eau –
Bioproduits, 2 place Eugène Bataillon, 34095 Montpellier Cedex 5, France
5
ENSCR, UMR CNRS 6226, Avenue du Général Leclerc, CS 50837, 35708 Rennes
Cedex 7, France
6
Université européenne de Bretagne
Abstract
Greywater recycling on-site appears today as one of the main ways to preserve water
resources in urban areas. Greywater are collected inside buildings through a dedicated
collection system, then treated in-situ. Recycled water can be used to flush toilets or
water gardens. The aim of this study was to reconstitute a greywater in order to carry
out reproducible experiments on different greywater recycling technologies. The
synthetic greywater (SGW) developed in this study consists in a mixture of septic
486 effluent to provide indicators of faecal contamination, and of technical grade chemical
products to mimic organic pollution. The physico-chemical and microbiological
characteristics of this SGW were tested (just after production and after a week of
storage at room temperature) and compared to those of real greywaters. Then, the SGW
was used to test a direct membrane nanofiltration system. The performances were
compared to the fluxes and retentions obtained on real greywater. The results were
really similar: evolutions flow were comparable, and permeate quality was high for both
effluents. These results validate the use of the synthetic greywater to evaluate
membrane efficiency or other technologies to treat greywaters.
Keywords
Greywater recycling, synthetic and real greywaters, membrane nanofiltration
Introduction
Water shortage has become a global issue in recent years: it concerns arid regions as
much as urban areas with very dense population. The water recycling in-situ appears as
one of the main ways of preserving water resources in urban areas. But in this case, it is
interesting to not recover all domestic wastewaters but only greywaters (GW), which
come from bathrooms, laundry facilities, and even from dishwashers and kitchen sinks,
but these latter are often left aside because they are putrescible. Within a household,
GW reuse can provide sufficient amounts of recycled reclaimed water to reduce potable
water consumption by 29 to 47 % (Lazarova et al., 2001). GW are collected inside
buildings through a dedicated collection system and are then treated in-situ for reuse. As
the recycled water has to be acceptable by the final user on both sanitary and aesthetic
aspects, GW have to be treated and disinfected (Winward et al., 2008). There are
innumerable examples of on-site GW recycling processes. For instance, direct
nanofiltration (NF) of wastewaters from public showers in a sports centre in Israel has
been mentioned by Ramon et al. (2004). A polyamide tubular membrane (molecular
weight cut-off of about 200 Da), at a pressure ranging between 6 and 10 bar, enabled
retentions of 93 % chemical oxygen demand, 83 % total organic carbon and 98 %
turbidity. But in some parts of the world, the authorities still need confirmation that
these technologies could be technically practicable, economically viable, and above all
completely safe. To evaluate and compare the performances of several recycling
processes, it is preferable to conduct experiments on a synthetic greywater (SGW), so
that the effluent is perfectly reproducible and representative of household GW. Very
few SGW have been reported in literature using elementary products actually found in
greywaters (Oschmann et al., 2005): most of authors have used a mixing of commercial
hygiene products and chemical substances to simulate the GW load. The formulations
of different SGW used in previous studies (Diaper et al., 2008; Fenner and
Komvuschara, 2005; Jefferson et al., 2001) are given in Table 1.
487 Table 1 – Synthetic greywater formulations from four previous studies
Fenner and Komvuschara,
Jefferson et al., 2001
2005
Commercial products
Chemical substances
Diaper et al., 2008
Secondary effluent
20 mL.L-1
E.coli culture 15 mL.L-1
Tertiary effluent 2.4 mL.L-1
H3BO3
1.4 mg.L-1
Amylodextrine 55 mg.L-1
Synthetic soap 64 mg.L-1
C3H6O3
28 mg.L-1
Dextrine
85 mg.L-1
Na2PO4
39 mg.L-1
K2SO4
4.5 mg.L-1
Na2SO4
35 mg.L-1
Na2CO3
55 mg.L-1
NaHCO3
25 mg.L-1
NaH2PO4
11.5 mg.L-1
Clay (Unimin)
50 mg.L-1
NH4Cl
75 mg.L-1
Deodorant
10 mg.L-1
Yeast extract
70 mg.L-1
Shampoo
720 mg.L-1
Laundry
150 mg.L-1
Sunscreen or moisturiser 15 or 10 mg.L-1
Toothpaste
32.5 mg.L-1
Vegetable Oil
7 mg.L-1
488 Shampoo
0.8 mL.L-1
Cooking oil
10 µL.L-1
The aim of this study is to reconstitute a SGW in order to carry out reproducible
experiments on different GW recycling technologies. As GW composition is highly
variable, this SGW was formulated so that the values of its parameters are comprised
between the extremum found in literature review and with direct greywaters analysis.
The SGW developed in this study consists in a mixture of septic effluent to provide
indicators of faecal contamination, and of technical grade chemical products to mimic
organic pollution. So, this SGW is reproducible in both time (no variation of the quality
of the products) and space (these products can be found all over the world). First, the
characteristics of this SGW are analysed and compared to those of real GW. Then, the
stability of the SGW as a function of time is tested. To do so, the SGW is produced and
stored during 13 days at room temperature, and its characteristics are evaluated 5 times
during the experiment. Lastly, the SGW is used to test a direct NF membrane system,
and the performances are compared to the fluxes and retentions obtained on real
greywater. A tubular NF membrane was chosen to carry out the tests because i) tubular
membranes can operate on a high load effluent without any pre-treatment, and ii) direct
membrane filtration is the least constraining membrane process when considering onsite recycling (Ramon et al. 2004). Moreover, NF seems to be the process that offers the
best compromise between solute retention (better than ultrafiltration) and energy
consumption (lower than reverse osmosis) (Trébouet et al. 1999).
Materials and methods
Synthetic and real greywaters
The composition of the synthetic greywater reconstituted in this study is given in Table
2. So, the SGW is mainly composed of technical quality chemical products to simulate
organic and inorganic pollution of greywaters from bathrooms (pollution due to human
body, body hygiene products and make-up related products). Septic effluent is added to
provide indicators of faecal contamination to simulate the presence of faecally
transmitted pathogens that are proven to be present in greywaters (Lazarova et al.,
2001). The addition of a few bacterial strains with a perfectly controlled population has
been considered. But the presence of a larger panel of micro-organisms is required to
reproduce microbiological interactions between greywater and greywater recycling
process. These interactions can consist in competition within a bioreactor, or biofilm
formation on a membrane surface. So, as such interactions can influence the recycling
scheme operating mode and efficiency, the choice was made to provide microbiological
load in the SGW by introducing a constant quantity of a very stable septic effluent.
489 The real greywaters, analysed and nanofiltered in this study, are collected in five
households located in north-west of France, in urban and rural areas, between March
2007 and September 2008. The samples are collected in the five households, directly in
the bath tubes, showers or wash basins. During the six campaigns of analysis that are
presented in this study, the samples taken from the five families are mixed, so that the
analysed real greywater is representative of an average family, composed of adults
(80 %), children under 15 years old (10 %) and babies under 2 (10 %).
Table 2 – Composition of the synthetic greywater
Product
Lactic acid
Bentonite
or
Cellulose
Sodium dodecyl sulfate
Glycerol
NaHCO3
Na2SO4
Septic effluent
CAS
number
Supplier
Purity Function
PSD*
Conc.
-1
1 2 3 4 5 6 7 (mg.L )
50-21-5 Panreac
> 85 % acid produced by skin x
1302-78-9 Riedel de Haën na
suspended solids
x
9004-34-6 Serva
> 90 %
151-21-3 Merck
> 85 % anionic surfactant
56-81-5 Panreac
99 % denaturant, solvent, x
moisturising agent
144-55-8 Panreac
> 99 % pH buffer
7757-82-6 Panreac
99 % viscosity
control
agent
microbiological load x
xx
x
100
100
x
x x 50
x x x x x x 200
xx
x 70
50
10
na : not available, * PSD: pollution simulated is due to: (1) human body (2) shampoo and shower gel (3)
soap (4) deodorant (5) tooth paste (6) shaving and moisturising cream (7) make-up and make-up remover
Physico-chemical, microbiological and particle size analyses
Conductivity and pH are determined by means of a Consort C862 apparatus following
ISO 7888:1985 and ISO 10523:2008 norms respectively. The experimental error is
± 2 %. Turbidity is evaluated in accordance with ISO 7027:1999 by a Lovibond
apparatus with ± 9 % uncertainty within a range of 1 to 2000 NTU. Suspended Solids
(SS) are measured by filtration through glass fibre filters (1.2 µm porosity) following
ISO 11923:1997. Experimental error is evaluated to ± 7 %. Chemical Oxygen
Demand (COD) is analysed by means of Spectroquant kits and Spectroquant Multy
spectrophotometer provided by Merck. The chemical reaction involved is similar to that
used in ISO 6060:1989. Biochemical Oxygen Demand for 5 days (BOD5) is measured
by respirometry with an OxiTop system from WTW incubated at 20 ± 0.1 °C during
5 days. The detection limit of this method is 0.5 mg O2.L-1. Dissolved Organic Carbon
(DOC) is analysed by a Shimadzu TOC 5000A apparatus following ISO 8245:1999
guidelines. The detection limit is 2 mg.L-1. The experimental errors on COD, BOD5 and
DOC are respectively worth 3 %, 4 % and 2 %.
490 Anionic surfactants (a-surfactants) are analysed with a kit method similar to ISO 78751:1996, which evaluates the concentration of Methylene Blue Active Substances
(MBAS). The kits (Spectroquant, 0.05 to 2 mg MBAS.L-1) and photometer
(Spectroquant Multy) are provided by Merck. The error on the measure is ± 4 %. Nitrate
and ammonium are measured by means of coloration analysis kits (Lovibond Water
Testing kits, Tintometer GmbH), and the coloration of samples is read with a
photometer (The Analyst, Orbeco) at 565 and 640 nm respectively. Nitrate are measured
in a 1 - 40 mg.L-1 NO3- range, and ammonium concentrations have to range between
0.03 and 1.3 mg.L-1 NH4+. Particle size distribution is determined by means of a LS 230
Coulter-Beckman granulometer equipped with a 15 mL microvolume cell, which allows
the access to the distribution of particles comprised between 0.4 and 2000 µm in
diameter.
Total coliforms, faecal coliforms and Enterococcus are isolated using two different
techniques: i) dilution and seeding method for raw greywater and membrane filtration
concentrates (adaptation of NF EN ISO 8199:2008), and ii) filtration on 0.2 µm poresize sterile cellulose nitrate filters for permeates (following NF EN ISO 9308-1:2000
and NF EN ISO 7899-2). In the two methods, the same agars and incubation durations
are used: total coliforms are incubated for 21 ± 3 hours at 36 ± 1 °C on lactose agar with
Tergitol 7 (100 mg.L-1) and triphenyl-2,3,5-tetrazolium chloride (25 mg.L-1). Total
coliforms count is given with an error of  0.4 log with a 80 % confidence level. Faecal
coliforms are incubated on the same agar during 21 ± 3 hours at 44 ± 2 °C. Only orange
or red colonies with a yellow hallow are considered as faecal coliforms. The results are
given with an error of  0.6 log. The incubation of Enterococcus takes place on Slanetz
and Bartley agar for 44 ± 4 hours at 36 ± 1 °C. Pink to dark red colonies slightly
bulging are counted as Enterococcus. The error on this method is  0.2 log.
Protocol for SGW stability study
The synthetic greywater is produced following the formulation given in Table 2 using
bentonite. The SGW is stored in a glass container sheltered from light, at room
temperature (comprised between 19 and 23 °C during the trials), on a rotating shaker at
60 rpm. Samples are collected by pumping the effluent, and several parameters are
analysed after 0, 1, 2, 3, 7 and 13 days.
491 Nanofiltration pilot plant and membrane
NF experiments are performed on a tubular membrane fitted onto a Microlab 40 plant
provided by VMA Industries (Figure 1). This pilot plant is operated at 35 bar. The flux
is measured by a Mettler PM4600 balance linked to a computer, which records the mass
( 0.1 g) of permeate filtrated during 1 or 2 min. Experimental error on permeation flux
measurement is lower than 2 %. The temperature is maintained at 25.0 ± 0.3 °C thanks
to a heat exchanger connected to a Mouvex RFA-30 cooling group. The tangential
velocity of the solution over the membrane surface is 2.5 m.s-1, so the Reynolds number
in the membrane module is equal to 3200 at 25 °C, meaning that the flow is turbulent.
The SGW used to test this process is formulated with cellulose, because it was observed
that at a high pressure, bentonite abrades the pilot plant tubing and pump.
The pilot plant is equipped with AFC80 tubular membrane made of a thin-film
composite polyamide/polyethersulfone provided by PCI Membrane Systems
(Basingstoke, UK). Its pure water flux at 25 °C at 35 bar is 54.8 L.h-1.m-2. This
membrane retains 80 % of a NaCl solution at 5 g.L-1 concentration at 20 bars. The
effective membrane surface is 0.031 m². The permeate is extracted while the
concentrate is recirculated into the feed tank so that the effluent concentration increases
progressively. During nanofiltration lab experiments, effluents are sampled when the
concentration factor, concentration factor (CF) reaches 87.5 %. This factor is given by
equation 1: CF = VP / Vi, where Vi is the initial volume of raw greywater and VP is the
final volume of permeate extracted. The retention of a pollutant (R) is calculated by
means of equation 2: R = 1- (CP / CR), where CP and CR are respectively the
concentrations of the pollutant considered in permeate and in the raw greywater,
analysed at steady-state.
Figure 1 – Simple flow diagram of the NF pilot plant Microlab 40
492 Results and discussion
Synthetic greywater composition
In Table 3, the properties of the SGW are compared to real greywater parameters
analysed in this study (RGWA), and in Table 4, they are compared to real greywater
form previous publications (RGWL). Three research publications give the composition
data of RGWL reported in Table 4: i) the first one (Jefferson et al., 2004) gives results of
a survey carried out on 102 individuals in England; ii) the second one (Eriksson et al.,
2002) is a review of numerous surveys concerning GW composition in several countries
from 1974 to 1999; iii) the third one (Friedler et al., 2004) reports data measured on 148
samples collected in Israel. The comparison between RGWL and RGWA shows that
numerous parameters have the same order of magnitude, such as BOD5 and DOC,
although the latter is higher in RGWA than in RGWL. Considering mean and standard
deviation values, the main differences between RGWA and RGWL can be observed in
three parameters: COD, total coliforms and Enterococcus loads are higher in RGWA. As
greywater composition depends on many factors, including the geographical position of
the households and their inhabitants’ culture, it is not surprising to observe significant
differences between RGWL, whose data come from numerous different locations, and
RGWA, which all come from the same area.
When comparing synthetic and real greywaters, it appears that COD is higher in SGW
than in RGW while, on the contrary, conductivity and BOD5 are lower in SGW than in
RGWL and RGWA. The differences observed in BOD5 are explained by the fact that
SGW lacks chemical products representing the pollution due to the human body,
especially sebum, which is quite difficult to mimic. A-surfactants are higher in SGW
than in RGWA: it is thus normal to observe an excess of COD in the presence of a high
content of a surfactants, as these latter are organic compounds. Microbiological loads of
RGWA and RGWL are disparate, but the addition of septic effluent to SGW has the
advantage of supplying micro-organisms so that indicators of faecal contamination are
comprised between those found in RGWL and RGWA. This is why total coliforms and
Enterococcus are more numerous in SGW than in RGWL, but less numerous than in
RGWA.
493 Table 3 – Average composition of the SGW and the RGWA
greywater,
Synthetic greywater, our study Real
(RGWA, 6 samples)
(SGW, 9 samples)
pH
our
m
σ
min
max
m
σ
min
max
6.76
0.30
6.29
7.29
7.28
0.41
6.46
7.84
Conductivity
µS.cm-1
188
18
159
212
377
36
331
434
Turbidity
NTU
24
16
4
42
53
19
26
75
SS
mg.L-1
72
14
41
87
59
19
23
80
COD
mg O2.L-1
454
33
391
505
253
43
176
323
BOD5
mg O2.L-1
65
6
58
75
110
23
85
155
DOC
mg.L-1
132
14
106
149
103
23
86
154
A-surfactants
mg MBAS.L-1 49.1
11.5
33.5
69.8
20.2
8.4
4.5
30.8
study
Total coliforms CFU / 100 mL 3.8 105 2.5 105 9.6 104 8.4 105 4.9 108 4.0 108 1.7 108 1.4 109
Faecal coliforms CFU / 100 mL 9.6 103 1.4 104 1.6 102 4.1 104 1.3 106 2.1 106 4.0 103 5.7 106
Enterococcus
CFU / 100 mL 2.7 103 2.6 103 5.3 101 8.2 103 2.2 105 4.2 105 8.0 103 1.2 106
m: mean, σ: standard deviation, min: minimum, max: maximum
494 Table 4 – Average composition of the SGW and the RGWL
Real greywater, literature (RGWL)
Synthetic greywater, our
study
Jefferson et Eriksson et
Frielder, 2004
(SGW, 9 samples)
al., 2004
al., 2002
pH
m
σ
min
max
m
σ
m
σ
m
σ
6.76
0.30
6.29
7.29
7.41
0.46
7.18
0.18
7.47
0.29
1286 425
101
109
Conductivity
µS.cm-1
188
18
159
212
166
84
Turbidity
NTU
24
16
4
42
93
33
SS
mg.L-1
72
14
41
87
108
57
174
134
100
145
COD
mg O2.L-1
454
33
391
505
339
107
364
225
451
289
BOD5
mg O2.L-1
65
6
58
75
142
42
240
177
146
54
DOC
mg.L-1
132
14
106
149
40
20
56
28
41
44
A-surfactants
mg MBAS.L-1 49.1
11.5
33.5
69.8
23.1
26.2
7.4
103
9.8103
2.0
103
6.0103
1.7
103
4.5103
Total coliforms CFU / 100 mL
3.8 10 2.5 10 9.6 10 8.4
5
5
4
105
2.6106
4.3
106
Faecal coliforms CFU / 100 mL 9.6103 1.4104
1.6
102
4.1
104
1.5103
1.9
103
CFU / 100 mL 2.7103 2.6103
5.3
101
8.2
103
1.8104
1.7
104
Enterococcus
3.1
106
5.7
106
m: mean, σ: standard deviation, min: minimum, max: maximum
Synthetic greywater stability
The pH of the SGW is stable during 13 days, with a minimum of 7.38 after 3 days, and
a maximum of 7.83 after 7 days. This means that, if degradation occurs in the SGW,
acids (mainly from lactic acid) and bases (mainly from sodium hydrogencarbonate) are
decomposed concurrently.
COD, BOD5 and DOC decrease simultaneously in course of time (Figure 2Figure ). In
13 days, COD decreases from 421 to 204 mg O2.L-1, while BOD5 is reduced from 90 to
25 mg O2.L-1. So, organic matters are oxidised chemically and biologically during the
495 trial, but DOC, BOD5 and DOC are constant for 24 hours: during the first day, the
evolution of these parameters is negligible.
When the SGW is stored at room temperature, microbiological activity is significant,
and are observed i) directly via the enumeration of faecal indicators (Figure 3), and
ii) indirectly, via the evolution of nitrates and ammonium ions (Figure 4). Enterococcus
population decreases continuously for 13 days: the total reduction is 2.8 log, and the
decline is accentuated during the second and the third day (1.4 log diminution). Total
and faecal coliforms behaviour is noticeably different. Total coliforms load is steady for
a week (average of 4.6 105 UFC / 100 mL ± 1.3 log) and then increases till the
population is multiplied by 100 after a 6 days period. Faecal coliforms population is
stable during 4 days (0.4 log decline). Then, the population is less numerous after 7
days of trial (1.1 log reduction in 4 days), but faecal coliforms are then multiplied by
more than 20 between seventh and thirteenth days.
500
250
400
200
300
150
200
100
100
50
-1
COD and BOD5 (mg O2.L )
The changes observed on nitrates and ammonium ions could be explained by the
microbiological evolution (Figure 4). NO3- and NH4+ concentrations in SGW are very
low during the whole trial (less than 2 mg N.L-1 for both parameters), but the small
changes in concentration could be explained by a nitrification/denitrification
phenomenon due to the presence of bacteria in both aerobic and anaerobic areas in the
container, which is not completely air-tight in order to be representative of a real
situation. It is very likely that SGW conductivity increase during the first and third days
is related to these successive transformations of nitrogen in the effluent.
-1
DOC (mg.L )
COD
BOD5
DOC
0
0
0
2
4
6
8
10
12
Time (days)
Figure 2 – COD, BOD5 and DOC evolution in course of time
496 log UFC / 100 mL
8
Enterococcus
6
Total coliforms
4
Faecal coliforms
2
0
0
2
4
6
8
10
12
Time (days)
Figure 3 – Evolution of faecal indicators load in course of time
-1
500
1,5
1,0
450
0,5
400
0,0
350
-
+
-1
NO3 and NH4 (mg N.L )
Conductivity (µS.cm )
550
2,0
0
2
4
6
8
10
NH4
+
NO3
-
Conductivity
12
Time (days)
Figure 4 – Evolution of nitrates and ammonium ions in SGW in course of time
Nanofiltration treatment of real and synthetic greywaters
The fluxes obtained during trials on the AFC80 with real and synthetic greywaters are
shown in Figure 5. Permeation fluxes in both effluents are very similar: the initial water
fluxes are comparable (70 vs. 60 L.h-1.m-2 on synthetic and real, respectively) and the
ratios between the fluxes measured and the initial water fluxes correspond pairwise.
These permeation fluxes, higher than 50 L.h-1.m-2 on both GW, are very satisfactory.
The filtration of GW induces very little fouling, and this is almost always reversible: the
final water flux, which occurs after reaching a concentration factor of 87.5 %, allows
the recovery of more than 94 % of the initial water flux of both effluents, meaning that
the cake layer is almost totally removed by a simple water rinsing.
The analyses of the effluents produced during filtration of SGW and RGWA on AFC80
at 35 bar are given in Table 5. The retentions are very satisfactory with, for both
effluents, a residual turbidity lower than 1 NTU, a maximum of 2 mg MBAS.L-1 of
497 a-surfactants and 2 mg O2.L-1 of BOD5 found in the permeate. Retentions and permeate
quality are similar for all pollutants except COD. This is probably due to a difference in
particle size distribution in the two effluents. The particle size analysis of real and
synthetic greywater showed no noticeable difference in distribution, so it seems that the
difference in particle size distribution is located in particles smaller than 0.4 µm in
diameter. The particle size distribution at a submicronic scale analysis should be studied
to confirm this conclusion.
-1
-2
Flux (L.h .m at 25 °C)
80
60
40
SGW
RGW S
20
0
Initial
water
flux
0%
< CF <
50 %
75 %
50 %
< CF <
< CF <
87.5 %
75 %
permeate
Flux
of
Final
water
flux
CF: concentration factor; SGW: synthetic greywater; RGWA: real greywater analysed in this study
Figure 5– Fluxes obtained during nanofiltration of SGW and RGWA on AFC80
From a health point of view, disinfection is ensured by the membrane on SGW and
RGWA: neither faecal coliforms nor Enterococcus are found in the greywater permeate.
The quality of the permeate produced by filtering SGW at 35 bar with AFC80 meets the
most stringent regulations regarding recycled water, such as the Japanese regulation:
less than 10 CFU / 100 mL of total and faecal coliforms, less than 10 mg O2.L-1 BOD5,
turbidity lower than 5 NTU and pH between 6 and 9 (Surendran and Wheatley, 1998).
Thus, the permeate produced by filtering greywaters on AFC80 at 35 bar seems to be
suitable for domestic use, such as toilet flushing and even clothes washing. These trials
showed similar results to the ones observed by Ramon et al. (2004) on a tubular
nanofiltration membrane used to treat real greywaters. This lead to the conclusion that
the SGW developed in this study efficiently simulates a real greywater. It can so be
used to evaluate and compare greywater recycling processes based on membrane
filtration: fluxes and retention rates obtained on real and synthetic greywaters are
similar.
498 Table 5 – Effluent analysis during nanofiltration of SGW and RGWA on AFC80
pH
Conductivity
µS/cm
Turbidity
NTU
COD
BOD5
DOC
A-surfactants
Real greywater, this study (RGWA)
CR
CP
R
7.2
7.5
163
9
382
67
4
<1
> 75.0 % 31
<1
>96.8 %
-1
464
77
83 %
258
< 25
>90.3 %
-1
63
2
97 %
115
1
100 %
149
47
69 %
91
39
57 %
50
7.9 103
2
ud
97 %
3.90 log
17
3.6 106
1
ud
92 %
6.55 log
2.5 103
ud
3.40 log
2.0 104
ud
4.30 log
mg O2.L
mg O2.L
-1
mg C.L
-1
mg MBAS.L
Faecal coliforms CFU / 100 mL
Enterococcus
Synthetic greywater (SGW)
CP
R
CR
6.3
6.9
CFU / 100 mL
CR: concentration in raw greywater, CP: concentration in permeate, R: retention, ud: undetected
Conclusion
A synthetic greywater was formulated and its composition was compared to that of real
greywaters from previous publications (RGWL) and analysed in this study (RGWA). As
greywater composition is highly variable, some noticeable differences were observed
between RGWL and RGWA, but SGW parameters were comparable to those of real
greywaters. The physico-chemical parameters of this SGW are stable for two days,
except for ammonium, nitrates and micro-organisms. The nitrogen variation is certainly
due to microbiological activity, so if the SGW has to be stored before use, it should be
stocked between 4 and 6 °C in order to limit the ageing of the SGW and to ensure the
effluent is representative of a fresh real greywater. Studying the real greywaters ageing
would allow a comparison between real and synthetic effluent evolution as a function of
time, and would confirm the similarity of the two effluents.
Next, the performance of a nanofiltration membrane (AFC80) at 35 bar was tested on
both synthetic and real greywaters. The results are almost identical, except for the COD
parameter. In both cases, fluxes were acceptable and user safety was assured: neither
faecal coliforms nor Enterococcus were found in permeates, which also contained very
low loads of organic matter. These results validate the use of the synthetic greywater
developed in this study to evaluate the performance of a process for greywater
nanofiltration. Thus, the validation of this synthetic effluent by comparison with real
greywater demonstrates that this new tool will be a real asset for future studies, for
example in evaluating greywater recycling processes on a reproducible effluent.
499 References
Diaper C., Toifl M., Storey M. (2008), Greywater technology testing protocol, CSIRO:
Water for a healthy country, National Research Flagship report series,
ISSN: 1835-095X
Eriksson E., Auffarth K., Henze M., Ledin A. (2002), Characteristics of grey
wastewater, Urban Water, Vol. 4, pp. 85-104
Fenner R.A., Komvuschara K. (2005), A new kinetic model for ultraviolet disinfection
of greywater, J. Environ. Eng., Vol. 131, pp. 850-864.
Friedler E. (2004), Quality of individual domestic greywater streams and its
implication for on-site treatment and reuse possibilities, Environ. Technol., Vol. 25,
pp. 997-1008.
Jefferson B., Burgess J.E., Pichon A., Judd S.J., Quarmby J. (2001), Nutrient addition
to enhance biological treatment of greywater, Wat. Res., Vol. 35, pp. 2702-2710
Jefferson B., Palmer A., Jeffrey P., Stuetz R., Judd S. (2004), Grey water
characterisation and its impact on the selection and operation of technologies for urban
reuse, Wat. Sci. Technol., Vol. 50, pp. 57-64.
Lazarova V., Hills S., Birks R. (2003), Using recycled water for non-potable, urban
uses: a review with particular reference to toilet flushing, Water Sci. Technol.: Water
Supply, Vol. 3, pp. 69-77.
Oschmann N., Nghiem L.D., Schäfer A.I. (2005), Fouling mechanisms of submerged
ultrafiltration membranes in greywater recycling, Desalination, Vol. 179, pp. 215-223.
Ramon G., Green M., Semiat R., Dosoretz C. (2004), Low strength graywater
characterization and treatment by direct membrane filtration, Desalination, Vol. 170,
pp. 241-250.
Surendran S., Wheatley A.D. (1998), Greywater reclamation for non-potable reuse, J.
CIWEM, Vol. 12, pp. 406-413.
Trébouet D., Schlumpf J.P., Jaouen P., Malériat J.P., Quéméneur F. (1999), Effect of
operating conditions on the nanofiltration of landfill leachates: pilot-scale studies,
Environ. Technol., Vol. 20, pp. 587-596.
Winward G.P., Avery L.M., Frazer-Williams R., Pidou M., Jeffrey P., Stephenson T.,
Jefferson B. (2008), A study of the microbial quality of grey water and an evaluation of
treatment technologies for reuse, Ecol. Eng. Vol. 32, pp. 187-197.
500 Presentation of Authors
Fanny Hourlier started in 2006 a PhD in GEPEA (laboratory of Process Engineering
for Environment and Agri-food sectors, CNRS Mixed Research Unit, www.gepea.fr)
with private financing of CSTB (Technical and Scientific Centre for Building). Her PhD
thesis is about “Greywater recycling and reuse by means of membrane processes inside
buildings”.
Dr. Anthony Massé studied “Submerged membrane bioreactors for urban wastewater
treatment” during his PhD. He has been associate professor of the University of Nantes
at GEPEA since 2006. His research interests are in structuration and abrasion
mechanisms of membrane deposit for complex mixtures, and submerged membranes for
aquaculture water treatment and recycling.
Prof. Pascal Jaouen (University of Nantes) is responsible for the team "Marine
bioprocesses and separations" at GEPEA. He has a long time experience with research
on membrane separation processes in marine biotechnology and aquaculture (e.g. water
treatment and recycling in aquaculture).
Dr. Abdel Lakel got is PhD in 1995 in Physics and chemistry, working on purification
process engineering. He is now a senior engineer in charge with air and water treatment
processes at CSTB.
Dr. Claire Gérente is a Process Engineering doctor and a research Engineer at the
School of Mines of Nantes. Her research is focused on biosorption of metallic ions and
arsenic from industrial or hyper diluted effluents.
Prof. Catherine Faur (University of Montpellier 2) is specialised in the engineering of
processes including solid/liquid interactions (adsorption, intensified processes coupling
adsorption and membrane filtration). She also works on elaborating activated carbons
and natural or synthetic polymers.
Prof. Pierre Le Cloirec is director of ENSCR (Superior National School of Chemistry
of Rennes). His work focuses on processes with fluid/solid interactions and complex or
hyper diluted mediums, with applications in air, water and wastewater treatment.
501 VIII.6
Materials in contact with drinking water
F. Derrien,
francois.derrien@cstb.fr
Dep. Hydraulics and Sanitary Equipment, CSTB, France
Abstract
Water distribution systems in buildings have an influence on water quality especially
when potable characteristics are required. The migration of substances from the
different materials constituting the pipes, fittings and other system components is due to
corrosion phenomena or any damage resulting of contact with water. The result is a
modification of organoleptic characteristics of water as well as an increase of metals or
organic compounds concentration. All around the world, different test methods have
been developed to measure these substances concentrations, and regulators have given
limiting values for parameters in order to protect the user's health. More and more
substances have been detected in drinking water during last years, and a great evolution
is expected in test methods and their accuracy, especially when introducing new
materials. Unification of evaluation methods and requirements is also highly desired.
This paper presents the state of the art, essentially in Europe. Possible existing piping
material degradation phenomena are reviewed with their consequences on water quality.
Corresponding regulatory requirements such as materials positive lists or standardised
test methods are presented. The evolution towards more accuracy, new analytical
methods and new requirements is discussed. The difficult unification work for
analytical methods and requirements towards more efficient and cost effective
evaluation of materials and products in contact with drinking water inside building
piping system is explained.
Keywords
Water treatment, scaling, calcium carbonate, bacterial growth, corrosion, magnetic field
502 1 Introduction
Materials constituting drinking water distribution systems are normally considered as
being without detrimental influence towards water. However, different reactions take
place at the interface with water and may impair the water quality. These reactions can
be corrosion, scaling, degradation of organic materials, biofilm developments and
bacterial growth. Faced to this problem, health authorities in different countries have
developed attestation of conformity systems. These systems are more or less complete
and based on materials positive lists and migration tests. Tendency towards
harmonisation of evaluation systems is required in the context of products market
globalisation worldwide. Evolution of the nature of controls and their accuracy implies
changes in quantity of controlled parameters and their threshold values.
2 Materials degradation phenomena
Depending on their nature, materials may suffer from different degradation phenomena.
Influent parameters are :
-
temperature in sanitary hot water (thermodynamic acceleration of
phenomena) or thermal shocks creating mechanical failures.
- addition of chemicals like disinfectants
- flow rate that may induce erosion problems.
Some problems in water distribution systems find their origin in poor design, bad
installation or lack of maintenance. In addition, some materials may favour the
proliferation of microorganisms.
All these phenomena have consequences on water quality. Organoleptic characteristics
like flavour, odour, colour and turbidity are modified.
Chemicals can be present in the distributed water by leaching of additives or coatings
used in the system materials, leaching of original material itself, reaction of materials
with chlorine or other direct additives, as well as biotransformation of leachates by
fungi, or bacteria in the system.
Periods of stagnation also contribute to increase chemicals trace level concentrations in
the stagnant water.
2.1 Corrosion
All metallic materials, when in contact with water are subject to corrosion. This process
is an electrochemical reaction between metal and water and its intensity depends on
temperature, concentration of different species in water, especially when chemicals like
disinfectants, most of them having an oxidising nature, are added. The resulting
damages are piping system destruction when pitting occurs and modification of water
composition by metal dissolution when uniform corrosion is the main process. As a
503 consequence very few metals are authorised for drinking water distribution. Copper,
stainless steel and galvanised steel are the main ones. However, other metals can be
dissolved in water like nickel from chromium plated valves, or lead from copper alloys
or brazed joints. In some cases, bacterial growth can be observed as a consequence of
corrosion due to the layer of products formed on the pipe surface which favours biofilm
development and offers nutriments to bacteria.
Figure 1 – Galvanised steel corrosion
2.2 Degradation of plastic materials
The interactions between polymers and water are the absorption of water by the
polymer and the migration of soluble species. The influent parameters are water
temperature, renewal of the water and the mode of contact. The CPVC, for example,
proves to be reactive with water while structure-properties relationship and the usual
utilisation did not lead to predict such a phenomenon. Indeed, immersed in hot water,
the CPVC is the seat of a chemical ageing leading to an osmotic damage depending on
production of chlorides whose origin is the polymer. Therefore, the absorption of water,
for example at 80°C, is close to 4% after 7000 hours of test. The kinetics of absorption
is probably correlated to the chlorinated species formation which participates to the
propagation of cracks. The main factors accelerating this process are the contact by
immersion and the temperature.
The kinetic study of the migration shows that some particular species : organo-tin
compounds, volatile organic compounds (chloroform, ethyl-hexanol, etc…), calcium
and chloride ions migrate from CPVC to water at least until 1000 hours.
Another example is degradation of HDPE in presence of chlorine based disinfectant.
The observed phenomenon is a longitudinal rupture of pipes. The kinetics shows a
maximum velocity when chlordioxyde is used as disinfectant and a lower velocity
corresponds to the use of bleaches. It is due to a greater diffusion of chlordioxyde in the
material than the diffusion of bleaches. This results for chlordioxyde in a total
consumption of anti-oxidant stabilizer in a depth of around 1 mm, leading to a
504 fragilisation and cracking, whilst for bleaches in a partial consumption of anti-oxidant
stabilizer.
Migration of monomer and additives such as plasticizers, antioxidants and application
solvents is also observed in pipes made of plastic materials.
Figure 2 - Lens-shaped hole (degradation by osmotic pressure) near water CPVC
interface
2.3 Degradation of other materials
This chapter refers to cementitious products (factory made or on site made), e.g. cement
mortar linings to metallic pipes, tanks, concrete pipes intended to be used for the
transport and storage of drinking water. Although covered by a liner, most of the time
concrete, after contact with drinking water, may alter it which results in changes in
odour, flavour, colour and turbidity.
Dissolution of admixture in water may also occur.
Other materials like sealing materials such as silicones and different rubbers are also
subject to impairment of water quality.
2.4 Consequences on water quality
The most perceptible water quality modification is taste and odour degradation. It can
be due to dissolved metals, dissolved additives of plastic materials, bacterial growth,
algae or fungi development or reactions between added chemicals and living organisms
505 in water or organic matters. Taste and odour degradation is difficult to measure and is
until now mostly carried out by standardised methods based on testing panels of
specialists.
Inorganic or organic matters dissolution in water is measured by traditional analytical
methods like GC/MS, ICP, and IR spectrometer.
3 Basis of assessment
3.1 Introduction
Different ways to assess the adequacy of materials or products for their use in contact
with drinking water exist. In many countries, no official system exists. In Europe
materials in contact have to fulfil chapter 10 of Drinking Water Directive : "Member
States shall take all measures necessary to ensure that no substances or materials for
new installations used in the preparation or distribution of water intended for human
consumption or impurities associated with such substances or materials for new
installations remain in water intended for human consumption in concentrations higher
than is necessary for the purpose of their use and do not, either directly or indirectly,
reduce the protection of human health provided for in this Directive". However, before
the publication of this directive, countries like France, Germany, The Netherlands,
United Kingdom and the USA had adopted systems going from positive lists of
authorized products to complete certification.
3.2 Positive lists
These lists gather materials or products authorized for contact with drinking water. They
are established by authorities, but they sometimes give no specific data on material
composition. They may be based on test results from migration or toxicological
experiments, but also on expert decisions. These lists do not exist in USA or many of
European countries. Updating these lists is most of the time a long procedure, and
creates barrier to trade or innovation.
3.3 Organoleptic testing
It is the first step in the European approval process, followed by cytotoxicity testing and
analytical screening. Most of the material rejections occur during orgnoleptic testing.
3.4 Migration tests
To constitute positive lists or in addition to them some countries have defined migration
tests. These migration tests are described in standards, and are based on a measurement
of concentration of expected substances in the water. They are carried out as stagnation
with test waters or stagnation with disinfection treatment water. Influent parameters are
time of stagnation, temperature, test specimen shape and surface to volume ratio.
506 Figure 3 – Test arrangement for migration test according to EN 12873-1
3.5 Test rigs
Particularly adapted for metals, test rigs have been developed for evaluation of material
quantity dissolved in water. More realistic than a single immersion of specimens in
water, because using pipe pieces and based on functioning regime simulating real
function in a building, these tests have the inconvenient of being more time consuming
(minimum six months), and not able to answer rapidly to the arrival on the market of
new products. According to EN 15664-1, the test can be used for three purposes :
-
Assess a material as a reference material for a category of materials using
the results of several investigations in different waters covering a broad
range of water compositions.
-
Asses a material for approval by way of comparative testing.
-
Obtain data on the interaction of local water with a material.
507 The test is based on a programme of alternating periods of once-through flow and
stagnation in a rig simulating the conditions in a domestic distribution system. The test
conditions are more relevant than conditions of continuous through-flow or sit and soak
tests and are applicable to all metallic materials in distribution systems. In domestic
drinking water installations, stagnation times of water considerably exceed the times of
through-flow. In most cases, metal release decreases with operation time. For some
alloying elements and impurities, however, an increase in their release can be observed.
Figure 4 – Test Rig according to EN 15664-1
3.6 Existing test standards
The most commonly used standards to asses the compatibility of materials in contact
with drinking waters are :
-
In the USA : ANSI/NSF Standard 61, health effect based, and AWWA
standards, performance based.
-
In Europe : EN standards like EN 1420 : Determination of odour and
flavour assessment of water in piping system, EN 12873 : Influence of
materials on water intended for human consumption – influence due to
migration, EN 13052 : Influence of materials on water intended for human
consumption - organic materials - determination of colour and turbidity of
water in piping systems, EN 14395 : Influence of organic materials on
water intended for human consumption – organoleptic assessment of water
in storage systems, EN 14718 : Influence of organic materials on water
intended for human consumption – determination of the chlorine demand,
EN 14944 : Influence of cementitious products on water intended for
508 human consumption – Test methods, EN 15664 : Influence of metallic
materials on water intended for human consumption and NF EN ISO 8795 :
Plastics piping systems for the transport of water intended for human
consumption – migration assessment.
3.7 Certification
Different levels of attestation of conformity exist. Certification, i.e. control by an
independent third body, is the most complete system to assess conformity of materials
for use in contact with drinking water. It exists for example in the USA (NSF
certification), in UK for products only and delivered by WRAS, in Germany on
materials and delivered by DVGW, in the Netherlands (KIWA certification called
ATA), and Belgium (Hydrocheck Procedures). According to the level of attestation,
these different certifications can include initial testing of materials and/or products,
quality system control of the manufacturer and continuous conformity control of
materials or products.
Where no certification exists, the assessment of materials in contact with drinking water
can be a type approval certificate given by health authorities or a notified laboratory.
4 Evolution of evaluation systems
4.1 Introduction
As presented above, big differences exist in evaluation systems. This shows evidence of
improving most of them due to the lack of some criteria. The following questions arise :
-
Is a type test of product and/or material valid if this test is carried out on
specimens which are not sampled by a third party ?
-
Is the assessment of a product and/or material valid if it is tested in cold
water whereas it is intended to be used in hot water ?
-
Is a five years validity of a certificate reasonable if no factory control
occurs during this period ?
-
What would be the behaviour of certain materials tested positively when
new after many years in function in a water distribution system ?
-
What is the validity of tests elaborated many years ago when the accuracy
of measurements was not so high and presence of certain chemicals in
water like pesticides or medicines unknown ?
The answers to these questions lead to propose an evolution of the evaluation systems.
4.2 More accuracy
The evolution of analytical technology leads to more and more accuracy in
measurements. This may induce the revision of certain parameters. But it can also show
509 the existence of substances that were up to now unexpected. These are for example
pesticides, antibiotics or human hormones now present in the water cycle.
The involvement of more actors in evaluation scene also raises the question of
reproducibility and repeatability of test results from one laboratory to another one.
4.3 New analytical methods
For identification of water leachable organic substances from materials in contact with
water intended for human consumption, GC/MS is now used.
For bacterial growth, PCR gives quicker results than culture, but different results. For
complete interpretation, both shall be used together.
For bacterial activity in a water distribution system or in evaluation of certain materials
in laboratory, ATP is a new promising technique but until now faced to reproducibility
and repeatability problems.
Electrochemical techniques for evaluation of passivation behaviour of certain metals
like stainless steel are widely developed in corrosion laboratories. They now appear in
standardised evaluation methods.
4.4 New requirements
In Europe now, the quality of water intended for human consumption is measured at the
point of use, i.e. the final user's tap. This involves building water distribution products
and materials.
The existing evaluation systems are based on tests made with materials and products
when new and ready to be put on the market. What they become after a few years in use
(effect of temperature or pressure, or ageing in contact with water or different chemicals
that it may contain) is not taken into consideration. This has to be evaluated.
The necessity of quick answer when new materials and/or products are put on the
market obliges to think to different and quicker test methods. One way can be the use of
modelling. A proposition of modelling method to estimate the expected migration of
organic substances from synthetic materials to drinking water has been made. Works in
order to find a short test on nickel release from chromium plated surfaces or a method to
determine the amount of lead on the surface of copper alloys are on progress.
510 5 Unification work
5.1 Introduction
An open market, as existing today has important consequences on the quality of
materials and products intended for the use of contact with drinking water. Existing
national regulation and evaluation methods protects consumers from health risks, but
create barriers and duplicated procedures for industry.
5.2 The reasons
Unification of evaluation of materials in contact with drinking water is a logic evolution
and an answer to the requirement that everywhere human beings could benefit of the
same health protection, evaluated with the same requirements and the same accuracy all
around the world. A unique worldwide system is also a progress towards simplification
and cost reduction for the industry.
5.3 The possible ways
The first intention within Europe was to create a unified evaluation method, common to
all countries and based on a CE marking of products intended to come in contact with
drinking water. Due to high differences between national systems in the evaluation
methods, in the standards used for tests and in the limit values adopted for assessment
parameters, this idea was, for the moment, abandoned. Tendency today is first to unify
test methods, although test methods are influenced by the requirements on expected
result values and accuracy. A second step is to recognise existing certifications managed
by existing testing institutes. However, this recognition needs to unify also certification
rules (quality system requirement, frequency of audits and samplings). Positive lists
shall also be merged together and a corresponding responsible body (European Food
and Safety Authority in Europe, for example) appointed. Recognition with other
international actors like NSF in the USA would also be a progress in the field of
materials in contact with drinking water.
5.4 CSTB as an actor for evaluation of materials in contact with drinking water
Notified body for the assessment of materials in contact with drinking water at
European level, CSTB is accelerating its scientific involvement and its strategic
contribution to water management. This is why Aquasim was founded, with the
objective of making realistic and accelerated simulations of physical and chemical
events occurring in the water chain within the building – plot environment system.
Located on CSTB's Nantes site, it includes the collection and production of water, its
transportation and use, treatment and release in the environment. It can represent
accelerated failures and ageing, it takes into account the impact of use and gives a
realistic framework for nuisance studies.
511 Figure 5 – Aquasim, a major facility for water management
6 Conclusions
When put in contact with drinking water, materials react and are subjected to ageing
phenomena. This leads to impairment of water quality and a potential risk for the water
consumer. Dissolution of different substances in water, bacterial growth, modification
of taste and odour are evaluated in different ways, depending on analytical traditions or
local regulations. However, tendency towards unification of these evaluation systems is
on process now, but the way is long and needs the achievement of common test
standards, certification rules and regulation in different countries.
7 References
1. Gaborit S., Lenes D. and Ponthieux A. (2008) : Taste and Odors in Potable
Water, Journées information Eaux, APTEN, Poitiers
2. Rabaud B., Rozental-Evesque M., Boulanger G., Glucina K. (2008) : Behaviour
of Plastic Materials Used in Potable Water Systems in Presence of Chlorinated
Disinfectants, Journées information Eaux, APTEN, Poitiers
3. Tomboulian P., Schweitzer L., Millin K., Wilson J.and Khiari D. (2004) :
Materials used in drinking water distribution systems : contribution to taste and
odor, Water Science and Technology, Vol 49 n. 9, pp 219-226
4. Barthelemy E.(2001) : Interactions entre l'eau et le poly(chlorure de vinyle)
chloré, Thèse, Université Aix-arseille 1
5. Rozental-Evesque M., Martin F., Bourgine F., Colin X., Audouin L., Verdu J.
(2006) : Etude du comportement de tuyaux en polyéthylène utilisés pour le
transport d'eau potable en présence de désinfectant chlore, Journées information
Eaux, APTEN, Poitiers
512 8 Presentation of Author
Francois Derrien is working at Centre Scientifique et Technique
du Bâtiment (France) project leader in "Hydraulics and Sanitary
Equipment" department. He is specialised in piping system
pathologies, materials in contact with water and water
conditioning.
513 List of Authors
Author
Page
Author
Page
Angus, Paul
Asano, Yoshiharu
Bosseler, Bert
Broome, Jeff
Cheng, C.L.
182
473
363
314
283,
441
39
55
39,
247
201
502
489
218
489
39,
55,
247
379
151
379
283
326
10
485
10
247
406
182
135
271
489
406
326
392
182
151
10
125
105
164
406
489
489
355
441
Lin, J.L.
Liao, W.J.
Massé, Anthony
Massolino, M.C.
Mui, K.W.
Murakawa, Saburo
283
283
489
39
26
105,
135,
151,
420
Nabeshima, Minako
Nakamura, Tsutomu
Nakayama, Satoshi
Nuijten, O.W.W.
Otsuka, Masayuki
151
10
10
164
326,
340,
456
69
201
259
105
406
441
55
238,
259
247
326
300
151
456
105,
151
55
182
85
164
441
26
420
456
Cheng, Liang Yee
Correia, G.
De Oliveira, L. H.
Demiriz, Mete
Derrien, F.
Faur, Catherine
Galowin, Lawrence S.
Gérente, Claire
Gonçalves, Orestes M.
Gormley, M.
Hamada, Yasuhiro
Hartley, C.
He, K.C.
Hongo, Norihiro
Hori, Shizuka
Hourlier, Fanny
Ichikawa, Noriyoshi
Ilha, Marina S.O.
Iizuka, Hiroshi
Ingle, Steven
Itagaki, Masaharu
Jack, L.B.
Jaouen, Pascal
Kamata, Motyasu
Kawasaki, Koichi
Kelly, D.A.
King, Derek
Kitayama, Hiroki
Kodera, Sadahiko
Kose, Hiroyuki
Koshikawa, Yasuo
Kuijpers, M.
Kutiyama, H.
Lakel, Abdel
Le Cloirec, Pierre
Lee, Eric Wai Ming
Lee, M.C.
Perc, Matjaz Nekrep
Petzolt, Ulrich
Pimentel-Rodrigues, C.
Sakamoto, Kazuhiko
Sakaue, Kyosuke
Shih, Z.Y.
Silva,G.
Silva-Afonso, Armando
Sousa Jr, W. C.
Suzuki, Kazutoshi
Swaffield, J.A.
Takaaze, Akiko
Takahashi, Yuta
Takata, Hiroshi
Tamaki, H.
Turner, John
Van der Schee, Walter
Van Wolferen, J.
Wang, R.Z.
Wong, L.T.
Yamate, Toshihiro
Yamasaki, Hironori
Yu, Y.C.
514 441
Download PDF
Similar pages