HPCycle Period vs Set-point Space Temperature - RiuNet

Document downloaded from:
http://hdl.handle.net/10251/47506
This paper must be cited as:
Corberán, JM.; Finn, D.; Montagud, C.; Murphy, F.; Edwards, K. (2011). A quasi-steady
state mathematical model of an integrated ground source heat pump for building space
control. Energy and Buildings. 43(1):82-92. doi:10.1016/j.enbuild.2010.08.017.
The final publication is available at
http://dx.doi.org/10.1016/j.enbuild.2010.08.017
Copyright
Elsevier
Elsevier Editorial System(tm) for Energy and Buildings
Manuscript Draft
Manuscript Number: ENB-D-10-00098R2
Title: A Quasi-Steady State Mathematical Model of an Integrated Ground Source Heat Pump for
Building Space Control
Article Type: Full Length Article
Keywords: Ground Source Heat Pump, Mathematical Model, Control Sensitivity Study, Optimisation
Analysis
Corresponding Author: Dr. Donal P. Finn, PhD, MEngSc, BE
Corresponding Author's Institution: University College Dublin
First Author: Jose M Corbéran, PhD
Order of Authors: Jose M Corbéran, PhD; Donal P Finn, PhD; Carla M Montagud, Dip.Ing; Fintan T
Murphy, BE, MEngSc; Kilian C Edwards, BEng
Abstract: This paper is concerned with the development of a mathematical model, capable of
describing the quasi steady state performance of an integrated ground source heat pump, which is
used for heating and cooling of an institutional building located in a Mediterranean climate. The model
is structured on functional basis according to the heat pump vapour compression or primary circuit, a
secondary ground loop circuit and a secondary building loop circuit. Heat pump heating and cooling
capacities, as well as COP, are considered to be dependent variables and are estimated in the model
using performance fitted maps. Independent variables include: compressor speed, circulation pump
speeds, ground loop return temperature and building circuit return temperature. The model is
validated using data from a full scale ground source heat pump installation. The validated model is
used to examine system capacity and performance sensitivity to different control optimisation
strategies, including set-point control of room air temperature, room air bandwidth temperature,
building loop return water temperature and building loop return bandwidth temperature.
Cover Letter
UCD School of Electrical, Electronic
and Mechanical Engineering
Scoil na hInnealtóireachta Leictrí,
Leictreonaí agus Meicniúla UCD
University College Dublin,
Belfield, Dublin 4, Ireland
An Coláiste Ollscoile Baile Átha Cliath,
Belfield, Baile Átha Cliath 4, Eire
T +353 1 716 1884/1787/1909
F +353 1 283 0534/283 0921
eem@ucd.ie
www.ucd.ie/eem
08 July 2010
Prof. Branislav Todorovic,
Editor
Energy and Buildings
Ms. Ref. No.: ENB-D-10-00098
A Quasi-Steady State Mathematical Model of an Integrated
Ground Source Heat Pump for Building Space Control
J.M. Corberan*, D.P. Finn+, C.M. Montagud*, F.T. Murphy+ and K.C. Edwards+
*
Escuela Técnica Superior de Ingenieros Industriales,
Universidad Politécnica de Valencia, Spain
+
School of Electronic, Electrical & Mechanical Engineering,
University College Dublin, Dublin, Ireland
Dear Prof. Todorovic,
Please find attached our further revised paper. We have included some
additional references as per your email of June 15th and we have also addressed the
question of the reviewer (see next page). We hope this addresses the issues raised.
Yours sincerely
Donal Finn, PhD
Senior Lecturer
University College Dublin
-----Original Message----From: ees.enb.0.9b723.23deadc7@eesmail.elsevier.com
[mailto:ees.enb.0.9b723.23deadc7@eesmail.elsevier.com] On Behalf Of Energy and
Buildings
Sent: 15 June 2010 17:22
To: donal.finn@ucd.ie
Subject: Your Submission
Editors comments:
Dear Sir,
I missed to write my remarks and suggestions. First you have 18 figures and 3
tables.
Please imagine how each figure would look after being published, published in
black and white and a maybe being reduced in the size. Hour in international
system is only h not hr. I also noticed that among references there is no one
concerning papers we published in our journal, and we had a lot. We recommend to
put in references these papers, not to be impression of the potential readers
that our journal never wrote anything about the problem you are treating. Please,
be so kind and go through our previous issues and be informed what we published.
And please add in the references. It is important to give some link and
contiunation with our previously published papers.
Regards, B.Todorovic, editor.
Reviewers' comments:
Reviewer #1: I recommend reviewing the reported overall heat transfer coefficient
(U) associated with the conditioned space. Please note that the value of U is
usually less than 1.0 W.m-2K-1.Could be 1.7 W.m-2K-1. instead of reported 17 W.m2K-1?
*Detailed Response to Reviewers
Ms. Ref. No.: ENB-D-10-00098
A Quasi-Steady State Mathematical Model of an Integrated
Ground Source Heat Pump for Building Space Control
J.M. Corberan*, D.P. Finn+, C.M. Montagud*, F.T. Murphy+ and K.C. Edwards+
*
Escuela Técnica Superior de Ingenieros Industriales, Universidad Politécnica de Valencia, Spain
+
School of Electronic, Electrical & Mechanical Engineering, University College Dublin, Dublin, Ireland
SECOND AMMENDMENTS July 2010
KEY
BLACK: Reviewer Comment
RED: Author response
BLUE: Paper modification
Additional clarification was sought on the building U value and load.
The following section has been added which hopefully clarifies the point raised by
the reviewer. We are sorry for any confusion caused.
The U values utilised in the model for the building external walls and windows were 0.6 W.m -2K-1 and
5.75 W.m-2K-1 respectively. By considering all thermal loads to include fabric gains, solar gains,
internal gains, infiltration and ventilation loads, a peak design load normalised with respect to floor
area, with an external design temperature of 35 oC and internal set-point temperature of 23.5oC, was
determined to be 102 W.m-2.
The Editor of Energy and Buildings asked that additional references be included.
Four additional references have been included as follows:
[8] Renedo, C.J, et al. (2006), ‘Optimum design for reversible water–water heat pumps’, Energy and
Buildings, 38, 1240–1247.
[9] Liu , X Hong, T. (2010) ‘Comparison of energy efficiency between variable refrigerant flow
systems and ground source heat pump systems’ Energy and Buildings, 42, 584–589.
*13+ Chen, Y. (2000), ‘Real-time predictive supervisory operation of building thermal systems with
thermal mass’, Energy and Buildings, 33, 141-150.
[14] Guo, W, Nutter, D. W. (2010), ‘Setback and setup temperature analysis for a classic doublecorridor classroom building’, Energy and Buildings, 42, 189-197.
The following additional text has been added to the body of the paper
The primary difference between ground and air-source heat pumps can be attributed to the external
source. GSHPs offer relatively constant operating parameter for short term measurements insuring a
smaller temperature difference between the condenser and evaporator when compared to ASHPs
[8]. In a study by Liu and Hong [9] an air source heat pump with variable compressor control was
compared to a GSHP with on/off control. The GSHP was shown to be more efficient than the ASHP
especially at high heating loads, but less so at part loads.Chen [13] and Guo and Nutter [14] also
studied optimal zone set-point temperatures for heating and cooling to improve system
performance.
The Editor of Energy and Buildings asked that that the colour of the diagrams be evaluated so that
the paper can be easily understood in the B&W print version.
The diagrams have amended so they are more easily understood in B&W.
FIRST AMMENDEMNTS APRIL 2010
Response to Reviewer #1
KEY
BLACK: Reviewer Comment
RED: Author response
BLUE: Paper modification
Reviewer #1: The general scientific level of the paper is good to very good. The
paper deals with an actual problem focusing on energy optimisation practical
issues on a basis of an experimental and numerical approach.
Recommendations
A comment should be given on how the energy efficiency of the building could have
impact on the obtained results.
A comment is included in the conclusions on this issue (given below). The authors
acknowledge that many factors will influence the heat pump performance including,
heat pump selection (and the choice of compressor therin), ground loop design,
building hydronic loop design and building design and efficiency. However within
the constransts of the current paper it has not been possible to consider these
issues.
Finally, it should be noted that overall system performance will be influenced by
many other factors including heat pump design, ground loop design, building
hydronic loop design and building efficiency. However the influence of these
issues on overall system performance were not explored in the current paper.
Building heating and cooling loads values are estimated in a very simple manner
given that they do not consider solar gains (office are East orientated!) nor
energy losses/gains associated to infiltration/ventilation rates. Also, the value
of the U is not specified (Eq. 20)!
A short section has been added at the end of Section 4.3 that addresses this
issue. It is as follows:
The U values utilised in the model for the building external walls and windows were 0.6 W.m-2K-1 and
5.75 W.m-2K-1 respectively. By considering all thermal loads to include fabric gains, solar gains,
internal gains, infiltration and ventilation loads, a peak design load normalised with respect to floor
area, with an external design temperature of 35oC and internal set-point temperature of 23.5oC, was
determined to be 102 W.m-2.
The results shown in Figure 9 need to be more explicit in terms of assumptions.
For instance, why not show the ambient temperature in free-floating regime in the
same graphic?
Figure 9 has been modified to indicate the ambient temperature.
To turn out the presentation more explicit, I recommend moving the first paragraph
of the introduction next to the last paragraph and reformulate. After the review,
the introduction would start with the state of the art and end with an outline of
the paper objectives and methods under these circumstances.
This has been done as suggested. The final paragraph now reads as follows:
This paper examines, for a ground source heat pump (GSHP), how control of building circuit variables
affects overall system performance. Building integrated ground source heat pumps usually
incorporate two secondary circuits; an indoor hydronic circuit for heating or cooling of a building and
an outdoor circuit, which is typically either a horizontal or vertical ground loop heat exchanger. This
work focuses on a sensitivity study, considered from a system control perspective and which
examines the role of various building circuit variables. The research was carried out by means of a
system mathematical model, developed as part of this research, which incorporates a heat pump
model, a building secondary loop model and a simple building space model. The various mechanical
components (heat pump, circulation pumps and fan coils) were individually modelled and
incorporated in an overall system model, which was validated by comparison to system
experimental data from an institutional building located in a Mediterranean climate. Variables
analysed include building hydronic circuit set-point temperature, building circuit hydronic
temperature bandwidth, building space set-point temperature and building space temperature
bandwidth.
I also recommend reviewing some other general aspects of the manuscript:
- the nomenclature appears not to include all variables. Although familiar to the
target scientific audience, SPF, COP, UA, etc. might need to appear as well...
COP, SPF and UA have been included in the nomenclature
- Vwater in addition of not being included in the nomenclature, appears with
capital letter in Eqs. 11 to 13, whereas in the remaining (13 and 14) is lowercase
letter...
V (velocity) has been included in the nomenclatuire and the anomaly corrected.
- the cooling thermal load appears in italic in nomenclature whereas in the body
is normal.
This anomaly has been corrected.
- Celsius degrees symbol is ºC and not simply C.
This anomaly has been corrected throughout.
- Mc is described as "water flow rate" in body text (page 4) while in nomenclature
M is used for describing "mass". Mc is in fact a mass flow rate such as defined
right above the M symbol in nomenclature and by Vce and Vci elsewhere in the text.
These anomalies have been corrected. In order to avoud confusion, Mc has been
removed from the Nomenclature and is explicitly defined for Fig 2 as follows:
Fig. 2. Validation of IMST-ART heat pump model for different evaporator and condenser
flow rates [15]. (Note: Mc = Condenser water mass flow rate kg.hr-1)
- where appears Yang and Pedersen (page 4) should be Yang et al. according to
references.
This anomaly has been corrected.
*Manuscript
Click here to view linked References
A Quasi-Steady State Mathematical Model of an Integrated
Ground Source Heat Pump for Building Space Control
J.M. Corberan*, D.P. Finn+, C.M. Montagud*, F.T. Murphy+ and K.C. Edwards+
*
Escuela Técnica Superior de Ingenieros Industriales,
Universidad Politécnica de Valencia, Spain
+
School of Electronic, Electrical & Mechanical Engineering,
University College Dublin, Dublin, Ireland
Abstract
This paper is concerned with the development of a mathematical model, capable of
describing the quasi steady state performance of an integrated ground source heat
pump, which is used for heating and cooling of an institutional building located in a
Mediterranean climate. The model is structured on functional basis according to the
heat pump vapour compression or primary circuit, a secondary ground loop circuit
and a secondary building loop circuit. Heat pump heating and cooling capacities, as
well as COP, are considered to be dependent variables and are estimated in the model
using performance fitted maps. Independent variables include: compressor speed,
circulation pump speeds, ground loop return temperature and building circuit return
temperature. The model is validated using data from a full scale ground source heat
pump installation. The validated model is used to examine system capacity and
performance sensitivity to different control optimisation strategies, including set-point
control of room air temperature, room air bandwidth temperature, building loop return
water temperature and building loop return bandwidth temperature.
1
Nomenclature
A
heat transfer surface area (m2)
Cp
specific heat capacity at constant pressure (J.kg-1K-1)
COP
Coefficient of performance
FREQ
circulation pump frequency (Hz)
mass flow rate (kg.s-1)
ṁ
M
mass (kg)
P
perimeter (m)
power (W)
Ṗ
Q
SPF
T
TRCI
TRCE
t
U
UA
v
V
VCI
VCE
V
t
x
heat transfer rate (W)
seasonal performance factor
temperature (ºC)
water return temperature for the building circuit (K)
water return temperature for the external circuit (K)
time (s)
overall heat transfer coefficient (W.m-2 K-1)
UA value (W.K-1)
velocity (m.s-1)
volume (m3)
water mass flow rate in the internal circuit (kg.s-1)
water mass flow rate in the external circuit (kg.s-1)
volumetric flow rate (m3.s-1)
time step
distance between two consecutive nodes
Greek symbols
heat exchanger effectiveness


density (kg.m-3)
Subscripts
c
FC
HP
ICP
i
o
SYSTEM
cooling
fan coil
heat pump
internal circulation pump
inlet
outlet
heat pump, fan coils, internal pump, external pump
2
1. Introduction
Research concerning control issues related to vapour compression systems has largely
focused on refrigeration and air-conditioning and to a lesser extent on heat pumps.
Considering refrigeration and air-conditioning systems, reported research has
examined various capacity regulation techniques including compressor modulation by
speed control, on/off cycling, as well as modulation using tandem compressors [1, 2,
3]. Control related research has focused on optimisation of evaporator and expansion
valve operation through superheat control and regulation [4, 5]. Other capacity control
techniques such as clearance volume control, suction pressure control, cylinder
unloading, back-pressure and discharge gas bypass regulation, although reported in
the literature, appear to have been less commonly deployed in commercial systems [6,
7]. The primary difference between ground and air-source heat pumps can be
attributed to the external source. GSHPs offer relatively constant operating parameter
for short term measurements insuring a smaller temperature difference between the
condenser and evaporator when compared to ASHPs [8]. In a study by Liu and Hong
[9] an air source heat pump with variable compressor control was compared to a
GSHP with on/off control. The GSHP was shown to be more efficient than the ASHP
especially at high heating loads, but less so at part loads.
Research to date on control of ground source heat pumps has focussed on capacity
control issues and to a lesser extent on control of secondary side working fluids.
The most common type of capacity control currently deployed in commercial ground
source heat pumps appears to be on/off compressor cycling [2]. Karlsson and Fahlen
[10] compared on/off compressor control with variable speed control for a brine-towater heat pump. They noted that the main benefit of using a variable speed
compressor was a reduction in the need for supplementary heating. In another study,
Karlsson and Fahlen reported that heat pumps are generally sized to match 60% of the
heating load [11]. Control was achieved using either on/off cycling, when operating at
part load conditions, or by supplemental heating when demand exceeds the heat pump
capacity. Karlsson and Fahlen examined intermittent control and variable-speed
capacity control by comparing the performance of two capacity-controlled heat pumps
and one standard heat pump with a single-speed compressor. Test data was then used
for seasonal performance factor (SPF) calculations. The SPF calculations indicated
that despite improved performance at part load, the variable-speed controlled heat
pump did not necessarily improve the annual efficiency compared with the
intermittently operated heat pump. Zhao et al. [12] examined the use of variable speed
compressors in small scale geothermal heat pumps. COP was observed to increase as
the heat sink temperature decreased, at a constant compressor frequency. For sink
temperatures of 30ºC and 35ºC, an increasing trend of COP with decreasing frequency
was observed. Lower frequencies reduced the refrigerant flow rate in the evaporator
and condenser, which allowed more time for the refrigerant to exchange heat with the
secondary fluid. This reduced the associated mean temperature differences between
the refrigerant and the secondary fluids in each heat exchanger. The difference
between the saturation temperatures (and pressures) of the evaporator and condenser
were therefore reduced. Chen [13] and Guo and Nutter [14] also studied optimal zone
set-point temperatures for heating and cooling to improve system performance.
Yang et.al. [15] created a mathematical model of an under-floor heat pump system.
The model was used to assess system performance with different controllers, namely;
3
P, PID and PID with pre-filtering of inputs, and relay (on/off control) for controlling
the heat pump capacity. The use of a PID pre-filtering controller with a variable speed
compressor was predicted to reduce power consumption, when compared to the relay
type control. This model does not appear to take into account the inefficiencies
inherent in variable speed compressors, such as the inverter and motor efficiency at
part load.
This paper examines, for a ground source heat pump (GSHP), how control of building
circuit variables affects overall system performance. Building integrated ground
source heat pumps usually incorporate two secondary circuits; an indoor hydronic
circuit for heating or cooling of a building and an outdoor circuit, which is typically
either a horizontal or vertical ground loop heat exchanger. This work focuses on a
sensitivity study, considered from a system control perspective and which examines
the role of various building circuit variables. The research was carried out by means
of a system mathematical model, developed as part of this research, which
incorporates a heat pump model, a building secondary loop model and a simple
building space model. The various mechanical components (heat pump, circulation
pumps and fan coils) were individually modelled and incorporated in an overall
system model, which was validated by comparison to system experimental data from
an institutional building located in a Mediterranean climate. Variables analysed
include building hydronic circuit set-point temperature, building circuit hydronic
temperature bandwidth, building space set-point temperature and building space
temperature bandwidth.
2. System Description
The GSHP system utilised in this research is located in an institutional building at the
Universidad Politécnica de Valencia, Valencia, Spain. The overall heat pump system
consists of a heat pump, an indoor circuit and an outdoor circuit as shown in Fig. 1(a).
2.1 Ground Source Heat Pump
The heat pump, a prototype unit, is a water-to-water reversible GSHP, which uses
propane (R290) as its primary refrigerant. The nominal heating and cooling capacities
are 18 kW (45ºC return /16ºC return) and 14 kW (30ºC return/12ºC return)
respectively. The outdoor loop consists of a ground source heat exchanger (GSHX),
which is coupled to the heat pump by an external hydronic loop. The GSHX itself
consists of six vertical boreholes connected in a balanced parallel configuration. Each
borehole has a depth of 50 m and contains a single polyethylene U tube of 25 mm
diameter bore, with a 70 mm separation between the upward and downward tubes.
The borehole overall diameter is 150 mm. The six boreholes are arranged in a 2x3
rectangular grid (18 m2), with a 3 m separation between boreholes.
2.2 Building Description
The building (see Fig. 1(b)) which is heated and cooled, comprises approximately
250 m2 floor area and includes a corridor, nine offices (located on the east façade of
the building), a computer room and a service room with office equipment and other
internal loads. The building loop consists of a series of 12 parallel connected fan coils,
an internal hydronic loop and a water storage tank (160 L). Each office, along with
the service room, is equipped with one fan coil, except for the computer room which
4
has two installed fan coils. The corridor does not have a fan coil unit present. Each fan
coil can be individually regulated by means of a thermostat and comfort temperature
and fan speed can be selected by the user. The control for each fan coil is governed by
a three way valve that allows the heating/cooling water to be modulated through the
fan coil. The valve is controlled by the thermostat of the room.
Fig. 1. GSHP system (a) GSHP system schematic (b) Building plan
2.3 System Control Components
The operation of the heat pump is governed by an electronic controller which,
depending on the building water return temperature, switches on/off the heat pump
compressor. The default values for the building circuit return temperatures are
between 37ºC and 43ºC for heating mode and 12ºC and 15ºC for cooling mode. The
ground circulation pump is controlled by the heat pump controller, which activates the
external pump 60 seconds before compressor activation. When the compressor
switches off, the external pump continues to operate for a further period of one
minute. A timer controls overall system operation, which was programmed to operate
between 0700 and 2100 hours, 5 days per week. Finally, in order to vary the fan coil
and GSHX water flow rates, two frequency inverters were installed, one for each
circulation pump.
2.4 Instrumentation and Data Acquisition System
Pt100 RTD devices are used to measure the inlet and return temperature for each
hydronic circuit. The mass flow rate for each circuit is measured by means of a
coriolis meter. The power consumption associated with the compressor and external
pump, the internal pump and fan coils are measured by two separate power meters.
The data acquisition system is programmed, such that the power consumption of each
individual component, i.e., the internal circulation pump, the external circulation
pump, the fan coils and the heat pump compressor unit, can be calculated from the
data collected by the two power consumption meters. Climatic data is collected using
a meteorological station located on the building roof such that air temperature, air
humidity, wind speed and solar irradiation are recorded every five minutes. The data
acquisition system has been in operation for a period in excess of 24 months prior to
the current work [16].
3. Methodology
A system level mathematical model was developed from the GSHP heat pump system
using Engineering Equation Solver (EES) [17]. The vapour compression software
package IMST-ART was used to model the behaviour of the GSHP as a standalone
system [18]. In order to do this, parametric data associated with the compressor,
condenser, expansion valve and evaporator was utilised within IMST-ART to model
the performance of the heat pump. The IMST-ART model of the heat pump was
validated using experimental data. Sensitivity studies using the validated IMST-ART
heat pump model facilitated the production of system performance maps of heat pump
capacity and compressor power consumption as a function of building and ground
water return temperatures for different mass flow rates. These heat pump performance
maps were correlated using polynomial equation fits, which were incorporated within
5
the EES system mathematical model, along with separate pump and fan coil
performance maps, which were also integrated into the EES model. The overall
system model was then validated against data from the full UPV GSHP system. The
validated model is used to examine system capacity and performance sensitivity with
different control strategies, including set-point control of room air temperature, room
air temperature bandwidth, building loop return water temperature and building loop
return temperature bandwidth.
4. System Mathematical Model
4.1 Heat Pump Model: IMST-ART Model and Experimental Validation
The IMST-ART heat pump model incorporates the key elements of the vapour
compression circuit including the evaporator, condenser, compressor, expansion valve
and connecting pipe work. Using IMST-ART, the heat pump model was constructed
on a component by component basis, thereby allowing validation between the IMSTART predictions and experimental test data, as shown in Fig. (2). By constraining
condenser and evaporator water inlet temperatures to be 10.8ºC and 22ºC respectively
and considering three condenser water mass flow rates (2000, 2300 and 2600 kg.h-1)
in conjunction with five evaporator water mass flow rates (2000, 2300, 2600, 3100
and 4200 kg.h-1), five data points were obtained for each condenser mass flow rate
condition. Comparison between the experimental results and IMST-ART predictions
for these mass flow rate permutations show that cooling capacity and COP of the heat
pump were within a ±2.5% error band [19]. Elsewhere, other more general validation
studies of IMST–ART demonstrated maximum error bands of less than ±4% for a
wider range of operating conditions where other heat pump systems were modelled
[20].
Fig. 2. Validation of IMST-ART heat pump model for different evaporator and
condenser flow rates [19]. (Note: Mc = Condenser water mass flow rate kg.h-1)
4.2 IMST ART Performance Maps and Correlations
Using the IMST-ART software package, a performance map describing the sensitivity
of heat pump capacity and power consumption subject to variation in external and
internal water mass flow rates (VCE,VCI) and external and internal water return
temperatures (TRCE,TRCI) was made. This was carried out for five different pump
speeds (20, 30, 40, 50, 60 Hz) for both heating and cooling modes. In cooling mode,
five different internal (building) water return temperatures (T RCI) (8, 10, 12, 14, 16ºC)
and five different external return temperatures (T RCE) (23, 24, 25, 26, 27ºC) were
examined. In heating mode, the internal water return temperatures examined were 45,
42.5, 40, 38, 36ºC and the external return temperatures consisted of 20, 17.5, 15, 12.5
and 10ºC. In total, this gave rise to 125 data points for each mode. Using this data,
correlations for heating and cooling mode, were established as follows:
Qevaporator
 f  VCI , VCE , TRCI , TRCE 
(1)
(2)
(3)
Qcondenser  f  VCI , VCE ,TRCI,TRCE 
PHP
 f  VCI , VCE , TRCI , TRCE 
6
where the water mass flow rates (VCI, VCE) are expressed in kg.s-1, the water return
temperatures are expressed in degrees Kelvin (K), the heat transfer capacities and the
power consumption are in Watts (W). Polynomial correlations, given by Eqs. (4) to
(9), were obtained by means of a quadratic regression curve fitting to the data with
regression values R2 equal to or better than 0.998.
Cooling Mode Correlations
T
Qevaporator  A0  (A1 * VCE )  (A 2 * VCE2 )  (B1 * VCI )  (B2 * VCI2 )  (C1 * TRCI )  (C2 * TRCI2 )  (D1 * RCE )  (D2 * VCI * TRCI )
TRCI
(4)
T
Qcondenser  A0  (A1 * VCE )  (A 2 * VCE2 )  (B1 * VCI )  (B2 * VCI2 )  (C1 * TRCE )  (C2 * TRCE2 )  (D1 * RCE )  (D2 * VCI * TRCI )
TRCI
(5)
PHP  A0  (A1 * VCI )  (A2 * VCI2 )  (B1 * VCE )  (B2 * VCE2 )  (C1 * TRCI )  (C2 * TRCI2 )  (D1 * TRCE )  (D2 * TRCE2 )  (E1 * TRCI * VCE )
(6)
Heating Mode Correlations
Qevaporator  A0  (A1 * VCE )  (A 2 * VCE2 )  (B1 * VCI )  (B2 * VCI2 )  (C1 * TRCE )  (C2 * TRCE2 )  (D1 *
TRCI
)
TRCE
(7)
Qcondenser  A0  (A1 * VCE )  (A 2 * VCE2 )  (B1 * VCI )  (B2 * VCI2 )  (C1 * TRCI )  (C2 * TRCI2 )  (D1 *
TRCI
)
TRCE
(8)
PHP  A0  (A1 * VCE )  (A2 * VCE2 )  (B1 * VCI )  (B2 * VCI2 )  (C1 * TRCE )  (C2 * TRCE2 )  (D1 * TRCI )  (D2 * TRCI2 )  (E1 * TRCE * VCI )
(9)
In order to optimise Eqs. (4) to (9), it was considered to be probable that a relationship
would exist between heat exchanger capacity, heat exchanger water mass flow rate
and water return temperatures. Several linear and quadratic terms for the water mass
flow rates of the internal and external circuits (VCI, VCE) were considered to account
for this relationship. In order to include the effect of water return temperature, it was
first necessary to distinguish between cooling and heating modes. In cooling mode the
evaporator is coupled to the building circuit and the condenser to the ground loop,
whereas in heating mode the evaporator is coupled to the ground loop and the
condenser to the building. To account for this, linear and quadratic terms for the water
return temperature from the building circuit were added to correlate the evaporator
capacity in Eqs. (4) and (8). Similar terms were included in Eqs. (5) and (7), in order
to determine the condenser capacity based on the water return temperature of the
external circuit. As the compressor power input is affected by both the external and
internal water return temperatures, both linear and quadratic terms were utilised in
Eqs. (6) and (9). Second, the ratio between the water return temperature for the
external and internal circuits was incorporated in order to correlate the evaporator and
condenser capacities (TRCE/TRCI). Taking into account, that in cooling mode, the
condensation temperature is influenced by the external circuit water temperature and
the evaporation temperature is influenced by the building return water temperature,
this term was also coupled to the pressure ratio. Similarly for heating mode, the ratio
would correspond to the water return temperature for the internal and external circuit
(TRCI/TRCE). Third, a further cross term (VCI*TRCI) was considered in Eqs. (4) and (5),
to account for the effect on evaporator and condenser capacities varying in the
internal circuit due to changes in water mass flow rate and the water return
temperature. Finally, another cross term was incorporated into the power input
correlations (Eqs. (6) and (9)), which accounts for the influence of the water mass
flow rate in the condenser (VCI for heating mode and VCE for cooling mode) and the
7
water return temperature in the evaporator (T RCE for heating mode and TRCI for
cooling mode).
A comparison between experimental results and predictions was carried out for a
typical day, both for cooling and heating. For cooling mode, Fig. 3 to 5 show the
comparison between experimental results and correlated predictions using Eqs. (4) to
(6) for a 10 hour period. As can be observed on the right hand axes of Fig. 3 to 5,
experimental measurements for cooling capacity and power input are predicted by the
polynomial correlations with a maximum mean deviation better than 5% for cooling
capacity and 2.4% for power input. However, a maximum mean deviation of
approximately 13.8% was observed for heating mode. The maximum deviations were
observed at start-up, most likely due to transient conditions. These deviations were
corrected in the model.
Fig. 3. Evaporator capacity (Cooling): Correlated data fit and experimental data
Fig. 4. Condenser capacity (Cooling): Correlated data fit and experimental data.
Fig. 5. Compressor power (Cooling): Correlated data fit and experimental data.
4.3 GSHP System Simulation Model
A system mathematical model of the overall GSHP unit, as outlined in Fig. 1(a), was
developed using Engineering Equation Solver [17]. This model incorporated each of
the system components including: fan coils, circulation pumps, hydronic pipe
network, storage tank, building conditioned space, as well as the heat pump
performance maps as described by Eqs. (4) to (9). Each of these sub-systems is
discussed further in the following sections.
Fan Coils: Experimental characterisation of the fan coil units was carried out to allow
performance maps to be established for the fan coil UA value (dependent variable) as
a function of the air volumetric flow rate and the water mass flow rate (independent
variables). The internal circulation pump frequency was varied from 20 to 60 Hz (in
10Hz increments), in conjunction with the three fan speed settings (V1, V2 and V3),
and the experimental UA value noted. For each fan speed, the value of the air flow
rate was based on data supplied from the manufacturer (290 m3/h, 440 m3/h,
590 m3/h). Fig. 6 shows the correlated UA values as a function of water mass flow
rate for one fan speed (V1), and the associated UA correlation. Similar correlations
were established for the other two fans speeds (V2 and V3).
Fig. 6. Fan coil UA correlation (Fan speed V1).
Circulation Pumps (Internal and External): The water flow rate for each circuit is
dependent on the pump and system characteristics. Calculation of the mass flow rates
for each pump was based on fitting empirical polynomial correlations to the
experimental data. Fig. 7 shows the relationship between the measured water mass
flow rate (dependent variable) and the internal pump frequency (independent
variable). A similar approach was used to map the performance of the external pump.
The pump inverter efficiency was experimentally characterised and the following
correlation was determined for inverter efficiency as a function of pump frequency:
Inverter efficiency = [-0.0261(FREQ)2+ 2.335(FREQ) + 43.765] / 100
8
(10)
Fig. 7. Circulation pump correlation (Internal Pump)
Hydronic Pipe Network: The internal hydronic circuit connects the heat pump
refrigeration circuit with the building fan coil units. The network is modelled using
the Lax-Wendroff model for 2D conduction transport [21]. The flow in the pipes is
governed by the transport equation which can be written as follows:
Twater
T
P·U
 v water · water 
· Twater  T 
t
x
water ·A·Cpwater
(11)
As the network is assumed to have negligible losses to the ambient, Eq. (11) can be
simplified as follows:
Twater
T
 v water · water
t
x
(12)
Eq. (12) is a hyperbolic partial differential equation and it is solved using the Lax &
Wendroff explicit discretisation approximation. This is governed by the CourantFriedrichs-Lewy (CFL) condition, which is subject to the following constraint:
t 
x
(13)
v water
The CFL condition implies that for a given velocity of the water circulating in the
pipes, the time step is controlled by the distance between the two consecutive nodes
considered. The discretised version of Eq. (12) is re-written as follows:
 Twater i1  2·Twater i  Twater i1
dTwater v water 
Twater i1  Twater i1  


dti
2·dx 
dx




·v water ·dt 



(14)
Solving (14) gives an explicit solution for each network node N, as follows:
t,final
Twater i  Twater initial,i 

0
 dTwater 

dt
 dti 
for i = 2 to N-1
(15)
Storage Tank: A storage tank is located at the outlet of the internal circulation pump
and provides thermal inertia within the system. It was modelled as follows:
dT
Mwater .Cpwater · water  mICP .Cpwater .(Tout,HP  Twater )  UA ·(Twater  T )
dt
(16)
Assuming negligible heat losses from the tank, Eq. (16) can be simplified to:
mICP .(Tout,HP  Twater )
dTwater

dt
Mwater
(17)
Building Conditioned Space: Each of the spaces within the building was modelled
assuming a constant volume closed system for the space as follows:
Mair cpair
dTair
 QFanCoil  Qload
dt
(18)
where QFanCoil is the heating/cooling capacity associated with the fan coil and is given
by:
9

QFanCoil  mair cpair FC · Tair  Tin,water

(19)
and Qload is the heating or cooling load associated with the space and is calculated as
follows:
(20)
Qload  UA· Tamb  Tair   Qint ernalgains
where Qint ernalgains is the thermal load associated with the presence of people,
computers and any other internal loads. For this study, an average load of 200W was
empirically determined for each space and an occupancy level of 50% was assumed.
The U values utilised in the model for the building external walls and windows were
0.6 W.m-2K-1 and 5.75 W.m-2 K-1 respectively. By considering all thermal loads to
include fabric gains, solar gains, internal gains, infiltration and ventilation loads, a
peak design load normalised with respect to floor area, with an external design
temperature of 35oC and internal set-point temperature of 23.5oC, was determined to
be 102 W.m-2.
5. Model Validation
Validation of the system model was undertaken by comparing model predictions and
experimental data taken from the UPV GSHP installation. Fig. 8 compares model
predictions for space and water temperatures against experimental data for a typical
heat pump cycle period. For this data, the following boundary conditions were
applicable for the system mathematical model:
 Space set-point temperature 23.5ºC
 Space temperature bandwidth ±0.5ºC
 Building return water set-point temperature 10.4ºC
 Building return water bandwidth ±1.6ºC
 Internal circulation pump frequency 60 Hz
 External circulation pump frequency 50 Hz
 External ambient temperature corresponding to July 29 th, 2009
Fig. 8 shows the evolution of the inlet and outlet temperatures for the internal circuit
and the space temperature for one office in the building. In order to analyse system
performance, critical points have been identified (Points 1 to 7) in Fig. 8. The heat
pump can be observed to switch off, when the return temperature reaches the lower
set-point of 8.8ºC (Point 1). A delay of approximately one minute can be observed
until equalisation of the supply and return temperatures occurs, which is attributed to
the thermal inertia of the evaporator. Once the heat pump has been switched off, the
supply temperature increases until point 2, and the return temperature continues to
decrease due continued circulation of the remaining 6ºC chilled supply water until it
reaches point 3. At point 3, as the heat pump is switched off, the building cooling load
results in a gradual increase in the building return temperature. A thermal time-lag,
attributed to the hydronic loop inertia, can be observed between points 2 and 3. At
point 4, as the return temperature reaches its upper set-point of 12ºC, the heat pump
switches on and the supply temperature drops until it reaches a quasi-steady
temperature of 8.8ºC at point 5. The return temperature continues to increase due to
space cooling load, until the chilled water once more reaches the fan coils, with an
associated time lag of approximately 3 minutes evident between points 4 and 6. This
delay is produced by the thermal inertia of the internal hydronic circuit. Once chilled
10
water reaches the fan coils, as the building load is less than heat pump capacity, the
return water temperature decreases until it reaches 8.8ºC at point 7, where the heat
pump switches off and cycle recommences.
Space temperature is also shown in Fig. 8 for one of the building offices. A space
temperature set-point of 23.5ºC was used, which is selectable by the user. Space air
temperature, which varies according to the fan coil operation, is governed by a 3-way
valve, that either circulates the chilled water through the fan coil or past it by means
of the valve bypass action. This valve is controlled by the room thermostat, such that
if the space temperature is 24ºC, the 3-way valve circulates the chilled water through
the fan coil, such that the space temperature decreases until it reaches 23ºC. At this
point, the chilled water will be diverted to the return circuit via the bypass. During the
phase, as fan cooling action ceases, the space temperature increases until it reaches to
24ºC and the cycle starts again.
Fig. 8. Comparison of system model predictions
with building experimental data (July 29, 2009)
Fig. 9 shows a set of simulation results for a 24 hour period, subject to a peak external
ambient boundary condition of 35ºC and an internal set-point temperature of 23ºC.
After an initial pull down period, quasi-steady behaviour is observed to occur, with an
increase in heat pump ON cycle time, observable between 12:00 and 18:00 hours,
reflecting increased cooling demand during the afternoon period.
Fig. 9. Daily temperature profile – system mathematical model
Fig. 10 shows the experimental measurements obtained for the 29th of July, 2009. It
can be observed that the simulation results for the return and supply water
temperatures are very similar to experimental data. The space temperature, however,
presents a different evolution. This is attributed to simplifications associated with
calculating the thermal load for each building space as discussed in Section 4.3.
Moreover, fan speed settings for each fan coil unit have been assumed at a single
fixed speed for all spaces.
Fig. 10. Daily temperature profile – system experimental measurements
In order to characterise overall performance, a number of additional parameters were
calculated as follows. The ON/OFF time operation of the heat pump was determined,
and is presented as a percentage for the heat pump in either ON and OFF mode, with
reference to a 24 hour period (1440 minutes). The seasonal performance factors of the
heat pump and the system were also calculated based on a daily analysis, as per Eqs.
(21) and (22).
t
 Q ·dt
c
SPFHP =
0
t
(21)
 P
HP
·dt
0
11
t
 Q ·dt
c
SPFSYSTEM =
0
(22)
t
 P
SYSTEM
·dt
0
Table 1 shows the results as determined for the experimental measurements and
simulation predictions. It can be observed that the model predictions are very close to
experimental data with a maximum absolute deviation of 3.26%.
Table 1 System performance parameters: experimental versus simulation results
6. Sensitivity Study
The sensitivity of GSHP performance subject to control of set-point temperature of
the fan coil water return temperature and the space air temperature are examined in
this section. In addition the role of bandwidth settings for the fan coil water return
temperature and space air temperature are also considered. The sensitivity studies
presented here are only for cooling mode, additional data is available for heating
mode, but is not considered in this paper.
6.1 Sensitivity Study Boundary Conditions
For cooling mode, the following reference boundary conditions were applicable:
 Space set-point temperature 23ºC
 Space temperature bandwidth ±1ºC
 Fan coil water return temperature 12ºC
 Fan coil water return temperature bandwidth ±1ºC
 Internal circulation pump frequency 60 Hz
 External circulation pump frequency 50 Hz
 External ambient temperature corresponding to the July 29th, 2009
6.2 Set-point Control of Water Return Temperature
The effect of set-point control of building water return temperature was examined
using the system model. Three water return temperatures, 10ºC, 12ºC and 14ºC were
considered. All other system variables were constrained at the boundary conditions
described in Section 6.1. Fig. 11 illustrates the effect of different water return setpoint temperatures on return temperature evolution for a typical 200 minute
operational period of the system. It can be observed that, as the building water return
set-point temperature is increased from 10ºC to 14ºC, the heat pump compressor ON
time decreases due to the increased cooling capacity associated with the underlying
refrigeration cycle. Analysis shows that capacity increases by approximately 3.6% per
degree increase of the building return water temperature. Fig. 12 summarises the heat
pump compressor cycle data. As the building water return temperature increases, the
compressor ON cycle time percentage is observed to decrease from 42.3% to 37.1%,
which is equivalent to 610 minutes and 535 minutes respectively. Therefore the main
effect of increasing building water return temperature, is that resulting from increased
refrigeration capacity, an increased fan coil capacity is available, resulting in
increased space cooling capabilities, leading to a shorter heat pump ON period.
Fig. 11. Building water return temperature evolution
12
Fig. 12. Compressor cycle period data versus building water return temperatures
Fig. 13 shows the daily system power consumption, associated with the data in Fig.
11. Compressor power consumption is observed to decrease from 30.19 kWh to 27.07
kWh, as the building return temperature increases, as does the total system
consumption which decreases from 59.38 kWh to 57.34 kWh. Moreover, the daily
SPF for the heat pump increases from 4.75 to 5.24. On the other hand, the daily SPF
of the system remains almost constant, as the decrease in total power consumption is
offset by the decrease in thermal energy supplied to the system. Nevertheless, it is
important to point out that variations in the return water temperature would have
greater effect on the SPF of the system for milder ambient conditions, which may
present a lower building cooling demand and a thus lower fan coil power
consumption.
Fig. 13. Energy consumption and daily SPF versus building water return temperatures
6.3 Set-point Control of Space Temperature
Space temperature is controlled by an individual thermostat located in each room.
Three separate space set-point temperatures (21ºC, 23ºC and 25ºC) were analysed and
the simulation predictions are shown in Fig. 14. All other boundary conditions were
maintained as outlined in Section 6.1. Space temperature set-point is seen to have a
significant influence on heat pump cyclic behaviour. An increase in space set-point
temperature will result in a decrease in the building heat gains. At higher space
temperatures, due to the higher mean temperature difference between the space and
the coil mean water temperature, a larger space to fan coil heat transfer load is
possible, and thus the water return temperature is observed to cool faster during the
heat pump ON periods, given that the heat pump capacity remains constrained and is
therefore unchanged. Thus for higher space set-point temperatures, the building water
return temperature reaches the lower set-point temperature more quickly, resulting in
a more frequent heat pump cycling associated with higher space temperature setpoints.
Fig. 14. Set-point control of space temperature
Fig. 15 summarises the heat pump ON/OFF cycle time for each of the space set-point
temperatures, whereas Fig. 16 gives daily power consumption and SPF of the heat
pump and the system. Total system power consumption decreases as Tspace is
increased, which is due to the decreased compressor ON time and its associated power
requirement. It can be observed in Fig. 16, that the daily SPF of the heat pump
remains practically constant, whereas SPF of the system decreases as the space setpoint temperature increases from 21ºC to 25ºC. This is because the building cooling
load demand decreases at the same rate as heat pump energy consumption but at a
greater rate than the total energy consumption.
Fig. 15. Compressor cycle data
Fig. 16. Energy consumption and daily SPF as a
function of space set-point temperature
13
6.4 Building Water Return Bandwidth
Bandwidth control for building water return temperature was examined for
bandwidths of ±0.5ºC, ±1ºC, ±1.5ºC, ±2ºC and the simulation predictions are shown
in Fig. 17. All other variables were maintained at the reference boundary conditions
as outlined in Section 6.1. Reference to Table 2 shows that changes in return water
bandwidth had a negligible effect on system performance characteristics. The ON
time operation of the heat pump can be observed to be not significantly affected by
the return temperature bandwidth, so the heat pump energy consumption remains
practically the same. In this case, total consumption would only be affected by the fan
coils, but it also appears to be not significantly influenced by changes in water return
bandwidth, because total consumption of the system is almost constant. Finally,
values for the SPF of the system and the heat pump are affected by less than 0.5% by
return water temperature bandwidth. This is because thermal energy supplied to the
building is constant.
Fig. 17. Building return temperature as a function of return water bandwidth
Table 2 System performance parameters as a function of return water bandwidth
6.5 Building Space Temperature Bandwidth
The role of space temperature bandwidth control was considered for bandwidths of
±0.5ºC, ±1ºC and ±1.5ºC. Simulation results are given in Fig. 18 and system
performance is summarised in Table 3. For these results, a building space set-point
temperature of 23ºC was used, along with the other boundary conditions given in
Section 6.1 Examining Fig. 18, it can be observed that adjusting of the space
temperature bandwidth has a modest effect on heat pump cyclic performance, as well
as overall system performance. The ON time operation of the heat pump stays
practically the same; this is why heat pump energy consumption remains constant. In
this case, as total consumption is only affected by the fan coils, it also appears to be
insignificantly affected by changes in space bandwidth temperature as total
consumption is practically constant. Finally, values for the SPF of the system and the
heat pump are slightly affected, less than 1%, by space temperature bandwidth.
Fig. 18. Building return temperature for different space temperatures bandwidths
Table 3 System performance parameters as a function of return water bandwidth
7. Conclusions
This paper examines for ground source heat pumps, how control of building circuit
variables affects overall system performance. Variables analysed included: building
hydronic circuit set-point temperature, building circuit hydronic temperature
bandwidth, building space set-point temperature and building space temperature
bandwidth. Assessment was carried out by means of a system mathematical model
which was developed using Engineering Equation Solver.
The dominant factor affecting system power consumption was found to be building
space set-point temperature. Space set-point temperature directly affects heat gains to
14
the building and by association the cooling load. Under quasi-steady state conditions,
as the set-point temperature increased from 21ºC to 25ºC, the daily system power
consumption decreased significantly from 67.54kWh to 48.23kWh. The daily SPF of
the heat pump remained practically constant, whereas the daily SPF of the system was
decreased from 2.58 to 2.21. Moreover, as Tspace increases, the rate of heat pump
cycling was also noted to decrease as compressor ON time was reduced from 48.85%
to 30.05%.
The effect of varying building return water temperature was found to have less
influence than space temperature set-point on system power consumption. As building
water set-point was increased from 10ºC to 14ºC, compressor power consumption
decreased from 30.19kWh to 27.07kWh, while compressor ON time was also
observed to decrease from 42.33% to 37.13%. Moreover, the daily SPF for the heat
pump improved from 4.75 to 5.24 and the daily SPF of the system increased slightly
from 2.41 to 2.47. Therefore, it would appear that higher water return temperatures
reduces the ON time operation of the heat pump, and results in an increase of daily
heat pump SPF of approximately 5% per each degree increase of the building water
return temperature.
Building water return bandwidth was noted to have almost a negligible effect on
system and compressor power consumption and compressor ON time was noted to
decrease only marginally with increased temperature bandwidth. Daily SPF values for
the system and the heat pump were not significantly affected by water return
bandwidth. Space temperature bandwidth was also observed to have a negligible
effect on compressor and system performance. Finally, it should be noted that overall
system performance will be influenced by other factors including: heat pump design,
ground loop design, building hydronic loop design and building efficiency. However
the influence of these factors on overall system performance were not explored in the
current paper.
Acknowledgements
This work was supported under the FP7 programme “Advanced ground source heat
pump systems for heating and cooling in Mediterranean climates” (GROUND-MED
FP7-ENERGY-2007-2-TREN-218895).
References
[1] L. Gasser, B. Wellig, K. Hilfiker, Exergy analysis for increasing the efficiency of
air/water heat pumps, Forschungsprogramm UAW Umgebungswärme, Lucerne
University of Applied Sciences and Arts, Engineering & Architecture, CC Thermal
Energy Systems & Process Engineering, Technikumstrasse 21, CH-6048, 2008.
[2] N. Flach-Malaspina, J. Lebreton, D. Clodic, Performance of a new air-to-water
heat pump system with controlled capacity, Int. Refrigeration and Air Conditioning
Conference, Purdue, 2004.
15
[3] P. Riviere, F. Malaspina, J. Lebreton, A new installation for part load testing of air
to water single stage chillers and heat pumps, International Refrigeration and Air
Conditioning Conference, Purdue, July 12-15, 2004, 1-8.
[4] C. Aprea, R. Mastrullo , Experimental Evaluation Of Electronic And Thermostatic
Expansion Valves Performances Using R22 And R407C, Applied Thermal
Engineering vol. 22(2) (2002) 205-218.
[5] B. P. Rasmussen, A. G. Alleyne, Control-Oriented Modeling of Transcritical
Vapor Compression Systems. J. Dyn. Sys., Meas., Control. 126(1) (2004) 54.
[6] R.J. Dossat, Principles of Refrigeration, fifth ed., Prentice Hall, 2001, ISBN-13:
978-0130272706.
[7] M. Yaqub, S. M. Zubair, Capacity Control for Refrigeration and Air-Conditioning
Systems: A Comparative Stud. J. Energy Resour. Technol. 123 (1) (2001) 92.
[8] Renedo, C.J, et al. (2006), „Optimum design for reversible water–water heat
pumps‟, Energy and Buildings, 38, 1240–1247.
[9] Liu , X Hong, T. (2010) „Comparison of energy efficiency between variable
refrigerant flow systems and ground source heat pump systems‟ Energy and
Buildings, 42, 584–589.
[10] P. Fahlén, F. Karlsson, Optimizing And Controlling Media Flows In Heat Pump
Systems, 8th IEA Heat Pump Conference, Las Vegas, USA, 2005.
[11] F. Karlsson, P . Fahlen, Capacity Controlled Ground Source Heat Pumps in
Hydronic Heating Systems, Int. Journal of Refrigeration 30(2) (2007) 221-229.
[12] L. Zhao, L. L. Zhaob, Q. Zhang, G. L. Ding, Theoretical and basic experimental
analysis on load adjustment of geothermal heat pump systems, Energy Conversion
and Management, 44 (1) (2003) 1-9.
[13] Chen, Y. (2000), „Real-time predictive supervisory operation of building thermal
systems with thermal mass‟, Energy and Buildings, 33, 141-150.
[14] Guo, W, Nutter, D. W. (2010), „Setback and setup temperature analysis for a
classic double-corridor classroom building‟, Energy and Buildings, 42, 189-197.
[15] Z. Yang, , G. K. M. Pedersen, L. F. S. Larsen, Modeling and Control of Indoor
Climate Using a Heat Pump Based Floor Heating System, 33rd Annual Conference of
the IEEE Industrial Electronics Society (IECON), Taipei, Taiwan Nov. 5-8, 2007.
[16] J. F. Urchueguía, M. Zacarés, J. M. Corberán , A. Montero, J. Martos , H. Witte,
Comparison between the energy performance of a ground coupled water to water heat
pump system and an air to water heat pump system for heating and cooling in typical
conditions of the European Mediterranean coast, Energy Conversion and Management
49 (2008) 2917-2923.
16
[17] Engineering Equation Solver, F-Chart Software, The University of Wisconsin,
Madison, 2009.
[18] J. M. Corberán, J. Gonzálvez-Macia, IMST-ART, a computer code to assist the
design of refrigeration and air conditioning equipment, www.imst-art.com, IMST,
Universidad Politécnica de Valencia, Spain, 2009.
[19] J. M. Corberan, C. Radulescu, J. G. Macia, Performance Characterisation Of A
Reversible Water To Water Heat Pump, 9th International IEA Heat Pump Conference,
Zürich, Switzerland, 20-22 May 2008.
[20] J. Gonzálvez-Maciá, F. Vera-García, J. R. García-Cascales, J. M. CorberánSalvador, C. Radulescu, (in Spanish) Estudio experimental de la influencia del flujo
másico de agua en las prestaciones de una bomba de calor de compresión de propano,
Análisis detallado mediante modelado con el programa de cálculo IMST-ART,
Congresso Iberico e II Congresso Iberoamericano, Ciencias E Tecnicas de Frio, Porto,
Portugal, 2007.
[21] P. D. Lax, B. Wendroff, System of Conservation Laws, Comm. Pure
& Applied Mathematics 13, 1960, pp. 217-237.
17
Table 1
System
Performance Parameters
HP ON (%)
HP OFF (%)
Energy Consumption - HP (kWh)
Energy Consumption - System (kWh)
Heat Pump Daily SPF (Eqn. 21)
System Daily SPF (Eqn. 22)
Simulation
Results
41.08
58.92
29.43
58.54
4.78
2.40
Experimental
Results
41.29
58.71
29.95
56.69
4.64
2.45
Abs
Deviation
(%)
0.5
0.35
1.74
3.26
3.04
1.90
Table 1 System performance parameters: experimental versus simulation results
Table 2
Building Water Return Bandwidth
HP ON (%)
HP OFF (%)
Energy Consumption - HP (kWh)
Energy Consumption - System (kWh)
Daily SPF of the HP
Daily SPF of the system
±0.5C
40.04
59.96
28.68
58.35
4.99
2.45
±1C
39.97
60.03
28.72
58.40
4.99
2.45
±1.5C
39.90
60.10
28.65
58.32
5.01
2.46
Table 2 System performance parameters as a function of return water bandwidth
±2C
40.39
59.61
28.65
58.32
4.95
2.44
Table 3
Tspace Bandwidth
Compressor On (%)
Compressor Off (%)
Energy Consumption - HP (kWh)
Energy Consumption - System (kWh)
Daily SPF of the HP
Daily SPF of the system
± 0.5
39.49
60.51
28.40
57.88
4.98
2.44
± 1.0
39.97
60.03
28.72
58.40
4.99
2.45
± 1.5
40.74
59.26
29.15
59.09
5.01
2.47
Table 3 System performance parameters as a function of return water bandwidth
Fig 1
Fig. 1. GSHP system (a) GSHP system schematic (b) Building plan
Fig 2
Fig. 2. Validation of IMST-ART heat pump model for different evaporator and condenser
flow rates [18]. (Note: Mc = Condenser water mass flow rate kg.hr -1)
Fig 3
Evaporator (Building) Capacity (Cooling mode)
Cooling Capacity (kW)
Cooling Capacity (kW)_exp
Cooling Capacity (kW)_correl
Mean Deviation (%)
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
400 420
440
460
480 500
520 540
560 580
600
620
640
660
680
700 720
740
760
780 800
820
840
860
880
900
920 940
960 980 1000
Time (mins)
Fig. 3. Evaporator capacity (Cooling): Correlated data fit and experimental data
Fig 4
Condenser (Ground loop) Capacity (Cooling mode)
Heating Capacity (kW)
Heating Capacity (kW)_exp
Heating Capacity (kW)_correl
Mean Deviation (%)
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
400 420 440 460 480 500 520 540 560 580 600 620 640 660 680 700 720 740 760 780 800 820 840 860 880 900 920 940 960 980 1000
Time (mins)
Fig. 4. Condenser capacity (Cooling): Correlated data fit and experimental data.
Fig 5
Compressor Power consumption (kW)
Power Consumption (kW)
PI(kW)_exp
PI(kW)_correl
Mean Deviation (%)
4
4
3
3
2
2
1
1
0
0
400
420
440
460
480
500
520
540
560
580
600
620
640
660
680
700
720
740
760
780
800
820
840
860
880
900
920
940
960
980 1000
Time (mins)
Fig. 5. Compressor power (Cooling): Correlated data fit and experimental data.
Fig 6
UA - Mass Flowrate Water
UA(W/ºK)
Volume Flowrate Air 290(m3/hr)
190
180
170
160
150
140
130
120
110
100
0.02
y = 1939.4x + 78.361
0.025
0.03
0.035
0.04
0.045
Mass Flowrate Water (kg/s)
Fig. 6. Fan coil UA correlation (Fan speed V1).
0.05
0.055
Fig 7
Frequency-Mass Flowrate Water Int. Pump
4500
3
Mass Flowrate Water (Kg/h)
4000
2
y = -0.0094x + 0.6414x + 65.323x - 380.86
R2 = 0.9999
3500
3000
2500
2000
1500
1000
500
0
0
5
10
15
20
25
30
35
40
45
50
55
Frequency (Hz)
Fig. 7. Circulation pump correlation (Internal Pump)
60
65
Fig 8
Daily Temperature Profile
Temperature (C)
Tspace_exp(ºC)
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
Tspace(ºC)
Tsupply(ºC)
Treturn(ºC)
4
Tsupply_exp(ºC)
Treturn_exp(ºC)
6
2 3
5
730 732 734 736 738 740 742 744 746 748 750 752 754 756 758 760 762 764 766 768 770 772 774 776 778 780 782 784 786 788 790 792 794 796 798 800 802 804 806 808 810 812 814
1
Time (mins)
7
Fig. 8. Comparison of system model predictions with building experimental data (July 29, 2009)
Fig 9
Daily Temperature Profiles
Tsupply(ºC)
Treturn(ºC)
Tspace(ºC)
Tamb (ºC)
35
Temperature (ºC)
30
25
20
15
10
5
400
460
520
580
640
700
760
820
880
940
1000
1060
Time (minutes)
Fig. 9. Daily temperature profile – system mathematical model
1120
1180
1240
1300
1360
Fig 10
Daily Temperature Profiles
Tsupply_exp(ºC)
Treturn_exp(ºC)
Tspace_exp(ºC)
Tamb (ºC)
35
Temperature (ºC)
30
25
20
15
10
5
400
460
520
580
640
700
760
820
880
940
1000
1060
1120
Time (minutes)
Fig. 10. Daily temperature profile – system experimental measurements
1180
1240
1300
1360
Fig 11
Treturn (ºC) vs Time (minutes)
Treturn=10ºC
Treturn=12ºC
Treturn=14ºC
16
Temperature (ºC)
15
14
13
12
11
10
9
8
700
720
740
760
780
800
820
Time (minutes)
Fig. 11. Building water return temperature evolution
840
860
880
900
Fig 12
HPCycle Period vs Return Set-point Temperature
On (%)
Off (%)
Time (%)
100
80
60
40
20
0
10
12
14
Return Temperature (ºC)
Fig. 12. Compressor cycle period data versus building water return temperatures
Fig 13
Total Consumption
HP consumption
DAILY SPF HP
DAILY SPF SYSTEM
70.00
7.00
60.00
6.00
50.00
5.00
40.00
4.00
30.00
3.00
20.00
2.00
10.00
1.00
0.00
0.00
10
12
14
Return Temperature (ºC)
Fig. 13. Energy consumption and daily SPF versus building water return temperatures
SPF
Power Consumed (kWh)
System Performance vs Return Set-point Temperature
Fig 14
Treturn (ºC) vs Time (minutes)
Tspace=21ºC
Tspace=23ºC
Tspace=25ºC
16
Temperature (ºC)
15
14
13
12
11
10
9
8
700
720
740
760
780
800
820
Time (minutes)
Fig. 14. Set-point control of space temperature
840
860
880
900
Fig 15
HPCycle Period vs Set-point Space Temperature
On (%)
Off (%)
100
Time (%)
80
60
40
20
0
21
23
Room Temperature (ºC)
Fig. 15. Compressor cycle data
25
Fig 16
Total Consumption
HP consumption
DAILY SPF HP
DAILY SPF SYSTEM
70.00
7.00
60.00
6.00
50.00
5.00
40.00
4.00
30.00
3.00
20.00
2.00
10.00
1.00
0.00
0.00
21
23
25
Room Temperature (ºC)
Fig. 16. Energy consumption and daily SPF as a function of space set-point temperature
SPF
Power Consumed (kWh)
System Performance vs Space Set-point Temperature
Fig 17
Treturn (ºC) vs Time (minutes)
+/-0.5ºC
+/-1ºC
+/-1.5ºC
+/-2ºC
16
Temperature (ºC)
15
14
13
12
11
10
9
8
700
720
740
760
780
800
820
840
860
Time (minutes)
Fig. 17. Building return temperature as a function of return water bandwidth
880
900
Fig 18
Treturn (ºC) vs Time (minutes)
+/-0.5ºC
+/-1ºC
+/-1.5ºC
16
Temperature (ºC)
15
14
13
12
11
10
9
8
700
720
740
760
780
800
820
840
860
880
Time (minutes)
Fig. 18. Building return temperature for different space temperatures bandwidths
900
Download PDF
Similar pages