ON-SITE WASTEWATER TREATMENT – POLONITE AND OTHER FILTER MATERIALS FOR

ON-SITE WASTEWATER TREATMENT – POLONITE AND OTHER FILTER MATERIALS FOR
ON-SITE WASTEWATER
TREATMENT – POLONITE AND
OTHER FILTER MATERIALS FOR
REMOVAL OF METALS, NITROGEN
AND PHOSPHORUS
Agnieszka Renman
June 2008
TRITA-LWR PhD Thesis 1043
ISSN 1650-8602
ISRN KTH/LWR/PHD 1043-SE
ISBN 978-91-7415-024-7
Agnieszka Renman
TRITA LWR PhD Thesis 1043
© Agnieszka Renman 2008
Doctoral Thesis
Department of Land and Water Resources Engineering
Royal Institute of Technology (KTH)
SE-100 44 STOCKHOLM, Sweden
Photo on cover: Interior view of the filter well Biop®, using Polonite® as reactive material, for
treatment of domestic wastewater.
ii
On-site wastewater treatment-Polonite and other filter materials for removal of metals, nitrogen and phosphorus
P REFACE AND ACKNOWLEDGEMENTS
This thesis was produced as part of the requirements for a doctoral degree at the Royal Institute
of Technology, Stockholm.
My Master’s degree (in Engineering) is from Warsaw Agricultural University (SGGW), where my
interest for ecotechnology began. Thanks to scholarships from the Swedish Institute I was able to
join the ecotechnological research programme, which was established by Assoc. Prof. Gunno
Renman in joint cooperation with my home university. During the first years I performed research together with Dr Lena Johansson Westholm until she graduated with a PhD in 1998. I
also had the opportunity to work with my friends from SGGW, Dr Joanna Kwapisz and Dr
Agnieszka Karczmarczyk. In 2001 I finally had the possibility to start my PhD studies at KTH.
This study was conducted in close cooperation with companies: Telge Återvinning AB, Hubbinettes Prefab AB and Bioptech AB. In the very beginning of this research in which I assisted, I
also had good cooperation with NCC AB. This latter contact has continued through Mr Magnus
Alfredsson, who organised the preparation of the first generation of Polonite® for my experiments, and later on through Mr Claes Thilander, Bioptech AB.
I would like to thank my advisor, Professor Per-Erik Jansson, for his support, especially in the
last phase of thesis preparation. Secondly, I would like to thank Assoc. Prof. Jon Petter
Gustafsson, Assoc. Prof. Lars Hylander and Assoc. Prof. Gunno Renman for their help, valuable
advice, discussions and comments on manuscripts. Prof. Dr hab. Zygmunt Brogowski and Prof.
Dr hab. Józef Mosiej from SGGW are acknowledged for challenging me and thus improving my
work. I am also especially grateful to Engineer Bertil Nilsson and Research Engineer Monica
Löwén for discussions and technical support in the laboratory. I wish to thank all my colleagues
at the Department of Land and Water Resources Engineering for the enjoyable atmosphere. I
thank Mr Börje Öhrvall and Mr Gustaf Persson, Viksta By, for their great help with the fieldscale experiments and nice “fika” events in their houses. Many thanks are due to Dr Mary
McAfee for revision of the English in this thesis.
The funding for this research was provided by the Swedish Environmental Protection Agency,
the Swedish Research Council Formas, Axel and Margaret Ax:son Johnsons Stiftelse, Telge
Återvinning AB and the Swedish Institute.
The deepest and warmest gratitude goes to my family, who have always supported and encouraged me along my way. I especially want to send my deepest love to my Mother, who is no longer
with us, for being so very important to me. Thank you for everything.
Stockholm, May 2008
Agnieszka Renman
iii
Agnieszka Renman
TRITA LWR PhD Thesis 1043
iv
On-site wastewater treatment-Polonite and other filter materials for removal of metals, nitrogen and phosphorus
S AMMANFATTNING
Småskalig, lokal rening av avloppsvatten – Polonite och andra filtermaterial för avskiljning av metaller, kväve och fosfor
Infiltration, markbäddar och konstruerade våtmarker har länge varit en förhärskande teknik att
rena avloppsvatten som genereras från enskilda hushåll. Dagvatten och lakvatten renas också ofta
på motsvarande sätt.
I denna avhandling har en ny typ av reningsteknik undersökts som innebär att avloppsvatten aktivt filtreras genom kemiskt reaktiva material. Dessa kan omvandla eller binda ämnen så
att de inte blir skadliga i miljön, t ex ammoniumkväve blir kvävgas eller binds genom jonbyte till
materialet.
Sammanlagt filtermaterial undersöktes varav de viktigaste är restprodukten masungsslagg,
Nordkalk Filtra P, Polonite® och wollastonite. Materialens förmåga att avskilja fosfor, kväve och
metaller har studerats genom kolonn- och fullskaleförsök i långa tidsserier. Mekanismer för fosforns och metallernas fastläggning har också klarlagts genom skakförsök och kemisk modellering.
Avloppsvatten härrörande från kommunalt nät och enskilda hushåll och lakvatten från avfallsdeponi har använts vid de experimentella försöken. Dessutom har ett konstgjort avloppsvatten med
enbart oorganisk fosfor och kväve utnyttjats vid ett kolonnförsök.
Resultaten från skak- och kolonnförsöken med konstgjort avloppsvatten visade att Nordkalk Filtra P och Polonite är bra filtermaterial för avskiljning av fosfatfosfor. Långtidsförsök med
Polonite i en patenterad typ av filterbrunn visade att detta material fungerar i praktisk tillämpning
och binder fosfor så att godkända utsläppsvärden erhålls i renat vatten (reningsgrad < 90%; utgående genomsnittlig fosfor koncentration < 1 mg/l). Polonite i filterbrunnar ska bytas efter intervall som bestäms av den belastning som filtret utsatts för. Materialåtgången uppskattas till minst
1-2 kg Polonite (fraktion 2-5 mm) för varje renad kubikmeter avloppsvatten i optimerade system.
De studerade reaktiva filtermaterialen är mindre lämpliga för avskiljning av kväve. I småskaliga
reningssystem krävs kompletterande enheter som svarar för detta.
Kunskap som erhölls från skak- och krukförsök visade att den bundna fosforn i filtermaterialen är tillgängliga för växter eftersom den till viss del förelåg som kalciumfosfater. Samtidigt
visade delstudier att metaller i avloppsvatten binds till filtermaterialen. Koncentrationerna var
dock mycket låga i förhållande till vad som är godkänt i slam för spridning på åkermark. Det
betyder att filtermaterial kan återföras till växtodling men fler undersökningar krävs för att en mer
allmängiltig slutsats ska kunna dras.
De reaktiva materialens metallavskiljande förmåga kan utnyttjas avsiktligt i andra reningssammanhang, nämligen vid filtrering av lakvatten och dagvatten. Masungsslaggen släppte vissa
metaller t ex mangan till vattenfasen vilket är en nackdel.
Fortsatt forskning krävs så att den reaktiva filterbäddstekniken kan förfinas och appliceras i större skala och omfattning.
v
Agnieszka Renman
TRITA LWR PhD Thesis 1043
vi
On-site wastewater treatment-Polonite and other filter materials for removal of metals, nitrogen and phosphorus
T ABLE OF C ONTENT
PREFACE AND ACKNOWLEDGEMENTS ............................................................................................... III
SAMMANFATTNING ..................................................................................................................................... V
TABLE OF CONTENT ................................................................................................................................ VII
LIST OF PAPERS ........................................................................................................................................... IX
ABSTRACT ......................................................................................................................................................... 1
INTRODUCTION............................................................................................................................................. 1
Background ........................................................................................................................................................ 1
The treatment technology ................................................................................................................................... 3
Research problem ............................................................................................................................................... 4
Objectives .......................................................................................................................................................... 4
STATE OF THE ART .......................................................................................................................................5
MATERIALS AND METHODS .......................................................................................................................7
Filter materials .................................................................................................................................................... 7
Industrial by-products ............................................................................................................................................................... 7
Natural materials and man-made products................................................................................................................................ 8
Column experiments (Papers I-VI) .......................................................................................................................................... 8
Pot experiment (Paper II) ................................................................................................................................. 10
Batch experiments (Paper III) ........................................................................................................................... 11
Modeling approach (Papers III, V) ................................................................................................................... 11
Full-scale treatment system experiment (Papers V, VI)...................................................................................... 11
Chemical and physical analyses ......................................................................................................................... 11
Evaluation of treatment performance ............................................................................................................... 12
RESULTS AND DISCUSSION ....................................................................................................................... 13
Phosphate removal and pH development in the columns.................................................................................. 13
Phosphate removal and pH development in full-scale bed filter ........................................................................ 16
Phosphorus sorption to different materials by layers ......................................................................................... 17
Transformation and removal of nitrogen .......................................................................................................... 19
Metal removal................................................................................................................................................... 21
Removal mechanisms- heavy metals and phosphorus ....................................................................................... 24
Recycling of filter materials back to agriculture ................................................................................................. 25
Bed filter lifetime for P removal........................................................................................................................ 28
CONCLUSIONS .............................................................................................................................................. 29
REFERENCES ................................................................................................................................................ 33
vii
Agnieszka Renman
TRITA LWR PhD Thesis 1043
viii
On-site wastewater treatment-Polonite and other filter materials for removal of metals, nitrogen and phosphorus
L IST OF PAPERS
This thesis is based on a summary of the following six articles, which are referred to throughout
the text, by their Roman numerals. The two first papers are published under my former surname
of Kietliľska.
I.
Kietliľska, A., Renman, G., 2005. An evaluation of reactive filter media for treating
landfill leachate. Chemosphere 61: 933-940.
The author was responsible for planning and conducting column experiment and wrote most of the manuscript in co-operation with the second author.
II.
Hylander, L.D., Kietliľska, A., Renman, G., Simán, G., 2006. Phosphorus retention in
filter materials for wastewater treatment and its subsequent suitability for plant production. Bioresource Technology 97: 914-921.
The author was responsible for conducting the study at laboratory (except pot experiment) and writing the
paper.
III.
Gustafsson, J.P, Renman, A., Renman, G., Poll, K., 2007. Phosphate removal by mineral-based sorbents used in filters for small-scale wastewater treatment. Water Research
42(1): 189-197.
The author’s contribution was to conduct the experiment, make laboratory analyses and, write part of the
paper.
IV.
Renman, A., Hylander, L.D., Renman, G., 2008. Transformation and removal of
nitrogen in reactive bed filter materials designed for on-site wastewater treatment.
Ecological Engineering, (submitted).
The author’s contribution was to conduct the analyses at laboratory and write the paper in co-operation
with the other authors.
V.
Renman, A., Renman, G., Gustafsson, J.P., Hylander, L. D., 2008. Metal removal by
filter materials used in domestic wastewater treatment. Journal of Hazardous Materials,
(submitted).
The author’s contribution was the same as in paper I and IV.
VI.
Renman, A., Renman, G., 2008. Phosphorus removal by Polonite® from wastewater
column experiments and a compact bed filter trial. Manuscript.
The author’s contribution was to make all laboratory work, work up the material and, write the first version of the paper.
Reprints are published with the kind permission of the journals concerned, and the papers are
appended at the end of the thesis.
ix
Agnieszka Renman
TRITA LWR PhD Thesis 1043
x
On-site wastewater treatment-Polonite and other filter materials for removal of metals, nitrogen and phosphorus
A BSTRACT
Bed filters using reactive materials are an emerging technology for on-site wastewater treatment.
Chemical reactions transfer contaminants from the aqueous to the solid phase. Phosphorus is
removed from domestic wastewater by sorption to filter materials, which can then be recycled to
agriculture as fertilisers and soil amendments. This thesis presents long-term column and fieldscale studies of nine filter materials, particularly the novel product Polonite®. Phosphorus, nitrogen and metals were removed by the mineral-based materials to varying degrees. Polonite and
Nordkalk Filtra P demonstrated the largest phosphorus removal capacity, maintaining a PO4-P
removal efficiency of >95%. Analysis of filter bed layers in columns with downward wastewater
flow, showed that phosphorus, carbon and nitrogen content was vertically distributed, with decreasing values from surface to base layer. Polonite and Filtra P accumulated 1.9-19 g P kg-1.
Nitrogen in wastewater was scarcely removed by the alkaline filter materials, but transformation
from NH4-N to NO3-N was >90%. Pot experiments with barley (Hordeum vulgare L.) revealed that
after wastewater treatment, slags and Polonite could increase plant production. Batch experiments and ATR-FTIR investigations indicated that amorphous tricalcium phosphate (ATCP) was
formed in the materials, so some of the accumulated PO4-P was readily available to plants. Low
heavy metal contents occurred in the materials, showing that they can be applied as soil amendments in agriculture without contamination risks. A full-scale treatment system using Polonite as
filter material showed an average PO4-P removal efficiency of 89% for a 92-week period, indicating the robustness of the filter bed technology.
Key words: alkaline materials; heavy metals; mechanisms; nutrient removal; sorption; speciation
modelling
ter treatment of the close to one million onsite wastewater treatment facilities in Sweden.
Such facilities in sparsely populated rural
areas are responsible for considerable discharges of phosphorus (P) to lakes, rivers and
streams. About 40% of these systems are
considered unacceptable by the Swedish
Environmental Protection Agency (SEPA).
The situation in other countries surrounding
the Baltic Sea is, for instance, the same or
even worse. On a global scale the sanitation
problem in rural areas of developing countries is large and the need for sustainable
technical solutions is urgent (WHO &
UNICEF, 2006). During many years, different technologies have been tested that may
improve the P removal efficiency in private
wastewater treatment systems (Crites &
Tchobanoglous, 1998) This development is
likely to accelerate because of the new requirements for small-scale wastewater treatment (less than 25 person equivalents connected) issued by SEPA, according to which
70-90% of total P should be removed.
I NTRODUCTION
Background
This thesis deals with the use of reactive bed
filters for on-site treatment of wastewater.
The focus is on ecological engineering systems that are more environmentally friendly
than many systems based on traditional techniques, as they usually require little or no
input of chemicals and electric energy for the
process (Mitsch & Jørgensen, 1989; Jenssen,
1996). They are also characterised as more
cost-effective in the long run. These properties are beneficial for developing ecotechnological treatment processes as an interesting alternative to on-site wastewater
treatment, urban stormwater and landfill
leachate. However, in some cases advanced
and expensive treatment systems (Renou et
al., 2008) have to be used together with
ecotechnological systems to meet treatment
goals.
There is a reason to raise the level of ambition when it comes to the domestic wastewa1
Agnieszka Renman
TRITA LWR PhD Thesis 1043
larly that from motorways, should not be
treated in sewage treatment plants. Solutions
for on-site treatment exist but efforts to
produce a cleaner and sustainable sludge for
recycling will be meaningless as long as
leachate and stormwater are mixed with
municipal wastewater. Leachate in an untreated form is unsuitable for direct discharge
into water courses, as the high metal and
ammonia concentrations would have a severe
impact on the ecology of the recipient water
body. There is a strong argument for the
introduction of filter systems using reactive
media prepared from natural minerals or
from by-products of steel production, such
as blast furnace slag. Such filters could be a
possible solution for the removal of metals
and could be used as a pre-treatment step
before leachate handling for N removal in a
constructed wetland (Kietliľska, 2004; Kietliľska et. al., 2005). Filter materials saturated
with heavy metals have to be replaced and
stored in a safe way. In addition, the knowledge obtained from research on filter materials is of interest for treatment of drinking
water and small industrial effluents. Implementation of the reactive bed filter technology can also be one option to reach the sanitation goals in developing countries. Bed
filters using reactive materials may provide an
effective, low cost and low maintenance
approach for on-site wastewater treatment.
Application of this technology can limit the
amount of P entering surface water systems
and groundwater from wastewater effluents,
and may provide significant environmental
benefits in rural, un-sewered areas (Heistad et
al., 2006).
Sand filter beds, soil infiltration and constructed wetlands (CW) are widely used systems for on-site treatment. All these systems
have two main disadvantages from the viewpoint of construction engineering, namely
that they are outdoor systems subjected to
impacts from precipitation and temperature
and that they have large area requirements.
The treatment capacity varies and is generally
low for P, while P recycling is not possible.
During the past 15 years, an increasing number of papers describe the possibility of using
materials in filter beds and CWs that are
In Sweden, the general recommendation has
been to construct a soil infiltration system or
sand-filter system after the septic tank. However, such systems have often shown poor P
removal efficiency. A set of new or upgraded
technologies for on-site treatment was recently tested in Sweden, including e.g. package treatment plants, chemical precipitation
in combination with large sand filters (Hellström & Jonsson, 2006). All systems installed
by the manufacturing companies showed
good performance during the project period,
partly as a result of regular monitoring and
maintenance. Many other different on-site
solutions have been launched, such as constructed wetlands and compact filter systems.
The latter group includes the Filtralite system, comprising a biofilter consisting of
coarse expanded clay aggregates (LWA) and a
bed filter with Filtralite P®. This technology,
developed in Norway, has proven very effective for most treatment parameters except
nitrogen and has a long P-retention lifetime
(Heistad et al., 2006). The newest generation
of Filtralite P has proven interesting for reuse
as a P-fertiliser in agriculture, although to
date only on the basis of results from pot
experiments (Nyholm et al., 2005).
Novel compact filter systems for on-site
wastewater treatment, similar to that introduced in Norway, have been developed and
launched in Sweden. Reactive filter materials
are increasingly used in these systems and
such materials are also of interest for producers of other treatment systems, see for example
the
Swedish
website
(www.avloppsguiden.se/ent_pro/ent_pro_fil
ter.htm). The filter materials can for instance
be used in filter wells as a post-treatment step
to reduce excessive effluent P concentrations.
On-site treatment is not only of concern for
domestic wastewater, but can also be applied
for treatment of leachate emanating from
landfills or for treatment of urban stormwater (Kietliľska, 2004; Hallberg, 2007). Treatment of these highly polluting wastewaters is
becoming mandatory world-wide. Discharge
to a municipal sewage treatment plant is
often difficult and expensive since the landfill
and the sewage plant are not located on the
same site. Furthermore stormwater, particu2
On-site wastewater treatment-Polonite and other filter materials for removal of metals, nitrogen and phosphorus
The treatment technology
A reactive bed filter (RBF) is designed according to the purpose of treatment, e.g. for
stormwater, domestic wastewater or leachate
water. The construction may be a filter well
(Fig. 1, 2) or a CW where the most important
component is the reactive medium or sorbent. Different types of artificial adsorbents
or ion exchange materials are available as
commercial products and most of them are
utilised when very high water treatment standards are required. However, these materials
are not useful for wastewater treatment because of their high cost. Many different filter
materials for wastewater purification have
been presented recently (Babel & Kurniawan,
2003; Johansson Westholm, 2006). Mineralbased materials such as Filtralite, Nordkalk
Filtra P and Polonite are those most studied
by researchers in Scandinavia.
The criteria for selection of a filter material
are also related to the purpose of treatment,
but usually include the following:
more efficient in terms of P removal. However, very few of these laboratory investigations of filter materials have been tested in
field-scale conditions. Two examples of filter
materials that have a decade of full-scale
experience are Filtralite P (including previous
types) and steel slag (Dobbie et al., 2005;
Ádám et al., 2007). The treatment systems
designed for these materials use large
amounts, implying large area requirements.
Recovering P from domestic wastewater and
returning it to arable or other productive land
is of major concern. The reason for this is
primarily to avoid environmental problems
such as eutrophication, which is caused by
nutrients from wastewater. However problems associated with P mining, fertiliser production and in the long term the finite
amount of minable phosphate minerals are
also important. Domestic wastewater contains other elements and substances that are
also of interest in recycling such as inorganic
N and potassium (K) and organic matter.
The prerequisites for P recovery should be
good as a result of suggested lower content
of contaminants of such plants and their
proximity to forest and cultivated land.
However, various structural issues must be
clarified, such as the division of responsibility, before recovery can take place. In accordance with the Swedish National Environmental Goals, levels of metal and alien
substances in the environment should be
close to zero within one generation (SEPA,
2002).
For this purpose it is considered important to
define long-term requirements on the quality
of wastewater fractions, to which reactive
filter materials belong.
x Availability of material
x Cost
x Physical characteristics; pH, porosity,
specific surface area
x Chemical composition
x Sorption capacity
This filter technology is a treatment system
where wastewater is allowed to percolate,
normally by gravity, through a reactive porous medium that removes the contaminant(s) from the water.
5
6
1
8
2
3
4
7
Fig. 1. A filter system layout (not to scale): (1) Inlet of household wastewater, (2) Septic tank,
(3) Dosing pump, (4) Filter well, (5) Biofilter, (6) Polonite bed filter, (7) Pump, (8) Sampling
well and outlet (Paper VI).
3
Agnieszka Renman
TRITA LWR PhD Thesis 1043
replaced regularly and returned to agriculture as P fertilisers and soil
amendments
x Minor or no use of electricity, since
pumps and other energy-consuming
devices are not needed at all or only
to a minor extent in the system
x A robust system in terms of function,
control and management
Research problem
The advantage of using sorbents or reactive
filter materials (RFM) for improving water
quality has been proven in several investigations. However, most studies have concentrated on the removal of P and have been
performed as batch experiments in the laboratory, while little is known about the success
of RFM in removing heavy metals and nitrogen. Furthermore, due to the lack of knowledge of operating systems with RFM over
longer time periods, further research is required. Generally, the longevity of filter materials for the removal of metals, N and P is
unknown due to lack of long-term studies in
of full-scale and column experiments. Furthermore, understanding of the removal and
transport processes in the reactive materials
is crucial for the control of full-scale operating systems, and for more widespread use in
on-site wastewater treatment. In addition,
more knowledge is needed about the
possibility to recycle the spent materials as P
sources and soil amendments in agriculture.
Fig. 2. Filter well for a single house using the
material Polonite.
If this type of treatment is successful, the
benefit is that a significant mass of the contaminant is accumulated in a finite and accessible volume of material, which allows for
future collection and disposal if necessary.
The latter is not the case if a natural and large
wetland is used for treatment purposes.
Reactive materials can be classified into two
distinct groups. One type promotes chemical
reactions that destroy the contaminant or
transform it to a more benign species (e.g.
reductive dehalogenation, denitrification,
biodegradation); the second attempts to
transfer the contaminant mass from the
aqueous phase to the solid phase (e.g. adsorption, ion exchange, precipitation). The
latter reactions provide environmental benefits by concentrating the contaminant mass in
a finite and known volume of material, which
then allows for easier collection and future
disposal, or possibly even acceptable rates of
release (Baker et al., 1998).
It is believed that the RBF can replace the
commonly used soil infiltration of wastewater and other types of on-site treatment technologies, which do not fulfil new environmental requirements and goals.
The RBF technology has several advantages,
the most important of which are:
Objectives
The aim of this thesis was to evaluate the
removal processes for metal, N and P that
take place in reactive bed filter materials, for
development of technologies used in on-site
wastewater treatment. The overall context is
how this technology can contribute to a
sustainable flow of P, N and materials in
society. This is in agreement with the longterm national aim to return all nutrients in
wastewater that can be recovered back to
arable soil or other land. For these reasons,
the thesis covers a broader spectrum of issues, such as P removal, recycling and engineering application.
x Purification is achieved by treatment
in materials of natural origin or byproducts from industry
x Used solid filter materials or sludge
from filtration and sedimentation are
4
On-site wastewater treatment-Polonite and other filter materials for removal of metals, nitrogen and phosphorus
several contaminants that have to be removed from the wastewater. The term sorbent refers not only to adsorption, but also to
processes such as precipitation, ion exchange,
complexation and mechanical filtration
(McCay, 1996). Sorption depends heavily on
conditions such as pH, concentration of
pollutants, ligand concentration, competing
ions and particle size. Sorbents may consist
of natural materials that are available in large
quantities and at a low cost, or of byproducts from industrial or agricultural operations. Since they are non-expensive, these
materials can be disposed of without expensive regeneration, although one must bear in
mind that they can contain hazardous substances after use and have to be treated accordingly. Examples of such sorbents are
bark and other tannin-rich materials, chitosan
and seafood processing wastes, zeolites, clays,
fly ash and peat moss (Babel & Kurniawan,
2003). Recent research has identified the use
of active filtration through alkaline media for
the removal of P from domestic wastewater
(Johansson & Gustafsson, 2000; Drizo et al.,
2006; Shilton et al., 2006; Ádám et al., 2007).
Active filtration belongs to a family of promising techniques for small-scale wastewater
treatment, which can be used in combination
with other treatments such as wetland treatment
systems
and
source
separation/collection (Valsami-Jones, 2001; Shilton
et al., 2006; Hedström, 2006).
In the literature, numerous filter materials are
described; those removing metals (Bailey et
al., 1999), those removing organic compounds (e.g. O’Hannesin & Gillham, 1998),
those removing N (Ahsan et al., 2001), and
those with removal capacities for P (Kløve &
Mæhlum, 2000; Johansson Westholm, 2006).
Recent studies have focused on filter materials for the removal of P from wastewater,
among which are Blast Furnace Slags (BFS),
electric arc furnace steel slag (EAF), Filtralite®, Polonite®, wollastonite, shell sand,
dolomite, red mud, peat and others (Sakadevan & Bavor, 1998; Bailey et al., 1999;
Hill et al., 2000; Kløve & Mæhlum, 2000;
Brooks et al., 2000; Brown et al., 2000; Heavey, 2003; Brogowski & Renman, 2004; Søvik
& Kløve, 2005; Drizo et al., 2006).
The research was performed in close cooperation with small enterprises that work with
patent-driven development of small treatment facilities. This approach made it possible to critically test the findings from laboratory experiments in full-scale and real
treatment conditions.
The specific objectives of this thesis were to:
x Investigate selected bed filter materials for their removal capacity of
phosphorus, nitrogen and metals
from wastewater under different
treatment conditions (Papers I-VI).
x Cast light on the phosphorus and
metal removal mechanisms by means
of solubility experiments and geochemical modelling (Papers III, IV).
x Study the used filter materials as soil
amendments and phosphorus sources
on the effectiveness of plant production (Paper II) and, quantify the content of metals in the filter materials,
particularly the Polonite used in
small-scale wastewater treatment, and
relate that content to statutory limits
on metal content in sludge for use in
agriculture (Papers I, V).
x Assess the long-term treatment performance and bed filter lifetime for
phosphorus removal by Polonite and
other filter materials (Paper VI).
S TATE OF THE ART
Many recent investigations have shown that
the removal efficiency of particular contaminants can be enhanced if a filter medium of
high sorption capacity is used in treatment
systems such as constructed wetlands for
leachate treatment (Mæhlum, 1998) and for
domestic wastewater treatment (Zhu, 1998;
Arias & Brix, 2005; Vohla et al., 2005; Hedström, 2006). Another approach involving in
situ CW upgrading with reactive filter media
has been developed, where separate filter
wells are constructed as a step preceding the
CW (Kietliľska, 2004). Besides the filter
construction, the most important part is the
selection of material or sorbent (Brix et al.,
2001). The sorbent is ‘reactive’ for one or
5
Agnieszka Renman
TRITA LWR PhD Thesis 1043
Phosphorus usually occurs in wastewater in
the form of organic phosphate (e.g. detergents) and inorganic phosphates (ortho- and
poly-phosphates). Various techniques have
been employed for P removal in wastewater
treatment as well as recovery technologies
(Morse et al., 1998). The broad categories of
P effluent treatment include physical, chemical, biological and crystallisation methods.
Chemical removal techniques are the most
effective and well-established methods to
date, including P precipitation with lime, or
aluminium and iron salts. However, the use
of metal salts may hinder widespread application of the P resource, as the recovery of P
from sludge is difficult (de-Bashan & Bashan,
2004). Consequently, the removal of P compounds through sorption processes onto
various filter materials has been tested increasingly during the past decade (e.g. Oguz,
2004; Kostura, 2005; Ganrot et al., 2007).
The possibility of recycling the spent reactive
materials or sorbents to agriculture as a fertiliser and soil conditioner has been investigated in some studies (Hylander & Simán,
2001; Kvarnström et al., 2004; Hylander et al.,
2006; Cucarella et al., 2008). Their content of
P might be low compared with industrial
fertilisers but their alkaline properties have
positive effects on acid soils (Cucarella et al.,
2007; Cucarella et al., 2008). However the
benefits of using these materials in agriculture
can be outweighed by their content of heavy
metals, which can be transferred from soil to
crops by plant uptake and become hazardous
for man and the environment. Domestic
wastewater contains different amounts of
metals depending on its source (Moriyama et
al., 1989). These metals can accumulate in the
filter material during filtration of the wastewater and add to any indigenous content that
might be present (Gustafsson et al., 2008).
Investigations carried out on treatment of
landfill leachate and urban stormwater clearly
show the ability of reactive materials to remove metals from the liquid to the solid
phase, but leaching from the filter matrix has
also been observed (Kietliľska & Renman,
2005; Hallberg & Renman, 2007).
Several mechanisms, such as sorption, complexation and precipitation, can control metal
concentrations in effluent from bed filters
employed for wastewater treatment. Divalent
metal cations tend to favour sorption on
colloidal particles with negatively charged
sites, calcite, clay minerals, organics and oxides of Fe, Mn, Al and Si (Trivedi & Axe,
2001). Complexation is the combination of
metal ions with non-metallic ligands by covalent bounds. The humic-like substances
formed from wastewater decomposition can
serve as ligands for metal complexes (Metcalf
& Eddy, 2003). Precipitation occurs when a
metal species falls out of solution as a solid.
Sulphides and carbonates are capable of
forming precipitates with Cd, Ni, Zn, Cu and
Pb (Kamara et al., 1989; Papadopoulus &
Rowell, 1989).
When the aim is to remove metal, N and P
by filtration through a reactive filter material,
the presence of dissolved and particulate
organic matter is of concern. Filtration is the
process of passing a liquid through a porous
medium, for example sand, either in natural
formation or filter constructions with the
expectation that the effluent will have a better quality than the influent. High concentrations of dissolved organic matter can result in
increased sorption but aqueous complexation
with metal ions can also result in a decreased
sorption (Düker et al., 1995; Jönsson et al.,
2006). Humic substances dominate the total
organic content in municipal wastewater
(Omoike & Vanloon, 1999). It is therefore
expected that organic species can inhibit P
sorption mechanisms (Van der Houwen &
Valsami-Jones, 2001) and have importance in
the removal of wastewater contaminants
(Katsoyiannis & Samara, 2007). Organic
matter is responsible for different kinds of
clogging that can occur in filter constructions. Deposited solids at the particle surface
in the pores leads to inner blockage. This
volume filtration is the main process of solid
removal in most filter beds. Surface filtration
may occur if a filter cake develops (Kiely,
1997). The mechanisms by which granular
filter materials remove colloidal matter from
water are complex and not fully understood.
This particularly applies to the bed materials
investigated in this thesis (cf. Meyer, 2004).
6
On-site wastewater treatment-Polonite and other filter materials for removal of metals, nitrogen and phosphorus
M A T E R I A LS
The P removal potential of a particular filter
material can be obtained from batch and/or
column studies. The results achieved from
column studies are considered more realistic
than the maximum P adsorption capacities
determined from batch experiments using
high P concentrations (Drizo et al., 2002).
The main disadvantage of the batch experiment is that particles can break, increasing
the specific surface area, which leads to overestimation of the amount of P removed by
the material (Hylander et al., 1999). However,
because of the time constraints for long-term
column studies, most investigations of P
sorption capacities of filter materials have
been performed by laboratory batch experiments (Cucarella & Renman, 2008). Phosphorus sorption experiments tend to yield
diverging results even with a material of very
similar chemical composition (Kostura et al.,
2005). The large scatter of results in P sorption on different materials investigated relates
to the fact that these materials are not fully
and unambiguously defined by their chemical, mineralogical and phase compositions
(Kostura et al., 2005). The design capacity for
P removal can not either be easily calculated
on the sorption results achieved in a batch
experiment, as demonstrated on filter beds
with sand and oil-shale ash (Vohla et al.,
2007).
AND METHODS
This chapter contains a description of materials and methods applied in batch, field and
column experiments and a pot experiment,
according to the objectives of the study.
Filter materials
Filter materials can be divided into the following classes: Industrial by-products, manmade products and natural materials (Johansson Westholm, 2006). The following materials were used in this thesis and are accordingly described in the following.
Industrial by-products
Three kinds of slag were used in the column
experiments: pretreated water-cooled blast
furnace slag (WCBFS), crystalline and amorphous blast furnace slag. All slag materials
were supplied by Merox AB (Oxelösund,
Sweden).
WCBFS is a product from the steel industry
and contains amorphous glasses and a small
amount of crystalline silicates. It had a particle size of 0-4 mm and was pre-treated with
1% CaO to increase its alkaline reaction. The
original composition of WCBFS was as follows (in g/kg dry matter): Ca (216); Si (155);
Mg (98); Al (70); and some amounts of K,
Mn, Na and Fe. Virgin material had a pH of
9.4. This filter material was tested for phosphorus removal from artificial solution in
Paper III.
Blast furnace slag is an industrial byproduct resulting from the process of extracting iron from iron ore at steel-works. It had a
particle size of 0.25-4 mm; 2-5.6 mm and 2-7
mm and was used in the experiments presented in Paper I, II, IV and V. Blast furnace
slag is characterised by high amounts of SiO2
(36.2%) and CaO (35%). Amount of other
components such as MgO and AL2O3 were
over 13% and 10%, respectively. Pure material had a pH ranging from 9.13 (Paper I) to
between 9.4-9.9 (Papers II, IV, V).
Blast furnace slag is produced in large
amounts by the steel industry and most of it
is reused in a variety of applications such as
road construction, liming materials in agriculture and to some extent for wastewater
treatment.
7
Agnieszka Renman
TRITA LWR PhD Thesis 1043
Natural materials and man-made products
lic conductivity since the mineral fractions
were coarse. Peat itself can act as an adsorbent of heavy metals but has not been found
to be capable of removing metals from landfill leachate to acceptable levels (Ringqvist et
al., 2002). Ammonia, on the other hand, can
be removed by peat to some extent (Heavey,
2003). Peat was used as an admixture with
mineral sorbents in experiments presented in
Papers I, III and VI.
Sand was used as inert reference material.
The experiment in Paper I used a washed
quartz sand produced by Ahlsell AB. Sand
had a particle size of 0.2-1.5 mm (Paper II)
and 2-5.6 mm (Paper I). The latter was taken
from a gravel-pit 20 km SW of Södertälje,
Sweden. The main components of sand are
SiO2 (69.3%), Al2O3 (13.4%), K2O (3.4%)
and Fe2O3 (3.1%). Contents of Cr, Cu, Ni
and Zn in the virgin material were 35.8, 20.6,
32.6, 49.4 ppm, respectively (analysed in this
study). The pure material before mixing with
peat had pH of 6.25 (Paper I).
Wollastonite (CaSiO3) is a natural calcium
meta-silicate. The material used in column
experiments was supplied by Minpro AB
(Stråssa, Sweden). It had a particle size of 1-3
mm and contained approximately 30% of
wollastonite, 11% of diopside, 11% of quartz
and 40% of feldspars. This filter material was
used to study removal of P from artificial
solution in Paper III.
Filtra P (Nordkalk Filtra P) used in column
experiments is produced after heating a mixture of limestone, gypsum and Fe oxides.
This filter material consists mainly of
Ca(OH)2 (20%), which gives strongly alkaline
properties. Filtra P is granulated in a particle
size of 2-13 mm. In this study samples of
Filtra P were obtained from Nordkalk (Pargas, Finland). This material was used for P
removal studies in a column experiment
(Paper III).
Opoka belongs to the group of silica-calcite
sedimentary rocks, a marine deposit composed of the remains of minute marine organisms from the late Cretaceous period
called Mastrych. This formation is found in
Poland, Lithuania, Ukraine and Russia.
Opoka consist mainly of SiO2 and CaCO3.
Depending on the ratio between those compounds, opoka can be classified as lightweight (> SiO2) or heavy-weight (> CaCO3).
In Polish literature this type of rock can be
classified either as geza (synonymous with
gaize) when the silica content is high or as
opoka when calcite dominates (Brogowski &
Renman, 2004). The pH of the opoka used in
the study was 8.3 and the particle size was 25.6 mm (Papers II, IV, V).
Polonite is the most novel filter material of
those used in experiments in this thesis.
Polonite® is a product manufactured from
the cretaceous rock opoka and is intended for
use in wastewater treatment (Brogowski &
Renman, 2004). This material is known for its
high sorption capacity of soluble phosphorus
and usefulness for recycling of nutrients in
agriculture.
Limestone was from the Ignaberga quarry,
South Sweden. Limestone, a sedimentary
rock, is used within several branches of industry such as the steel and concrete industries and the pulp industry. It is also used as a
liming agent in agriculture. Due to its high
content of calcium, mainly CaCO3, limestone
has attracted attention as a candidate substrate for P removal. The pH of the limestone used in the study was 8.9 and the particle size was 1-2 mm (Papers II, IV, V).
Peat used in experiments was a manufactured product from Hasselfors Garden AB: a
natural Sphagnum peat without any additives,
moderately decomposed (humification degree
H3-H4) and with a density of 70 kg m-3. It
had a pH ranging between 3.0-4.0.
Calcium-rich materials can react with the
sulphate commonly found in leachate and
create gypsum, but by including an organic
component this effect can be reduced. Another reason for including peat in the filter
matrix was to decrease the saturated hydrau-
Column experiments (Papers I-VI)
Three column experiments were performed
in this thesis, using three kinds of feed water;
landfill leachate, municipal wastewater and
synthetic solution, respectively.
Five columns (K0 – K4) made of PVC, each
having an overall height of 60 cm and an
8
On-site wastewater treatment-Polonite and other filter materials for removal of metals, nitrogen and phosphorus
and sprinkled equally over the surface area of
the column materials every second hour for
67 weeks. The six filter materials and their
grain size used were: coarse amorphous
furnace slag, 0.25-4 mm (ASC); coarse
crystalline furnace slag, 0.25-4 mm (CSC);
very coarse crystalline furnace slag, 2-7 mm
(CSVC); limestone, 1-2 mm; opoka, 2-5.6
mm; Polonite® , 2-5.6 mm; and quartz sand,
0.2-1.5 mm. They were filled to a height of
50 cm in 30-cm wide columns (two materials
in duplicate) and received wastewater under
unsaturated flow conditions. The sand was
used as a reference material (Ritter &
Eastburn, 1988).
Sampling was performed regularly during 67
weeks with a total number of 40 duplicate
samples (100 mL polypropylene bottles) of
influent and effluent of each column. The
pH and electric conductivity (EC) of the
liquids were determined within 1 h,
whereafter they were frozen (-18 oC) until
time of analysis. Samples were filtered
through a 0.45 Ƭm filter (Satorius) before
each analysis. The influent and effluent
concentrations of P (Papers II, VI) and inorganic forms of N (Paper IV), i.e. ammonium
(NH4-N), nitrate (NO3-N) and nitrite (NO2N), were analysed.
It was not possible to maintain the desired
wastewater flow of 6 L day-1 equally to all
columns, so the incoming volume was monitored by regular manual measurements and
verified by tipping-bucket flow meters connected to a data logger (Campbell CR10X).
In Paper III a column leaching experiment
was carried out to provide information on
long-term P removal performance of the
filter materials. Columns were constructed of
PVC tubes of 60 cm length and 10 cm inner
diameter and filled with a 50 cm layer of the
appropriate filter material. However, the
Polonite was sieved for removal of particles
less than 2 mm and its composition was
changed by addition of 10% (w/w) Sphagnum
peat. The bottom of the columns was filled
with a 2 cm layer of gravel, and with a coarse
plastic (HDPE) filter to prevent loss of material from the columns. The experiment was
carried out at room temperature (20 oC)
under saturated conditions.
internal diameter of 9.8 cm, were used for
testing landfill leachate (Paper I). All columns
were filled with substrate to a height of 50
cm. The top of each substrate was covered
with polyester filter to prevent media scouring and clogging during leachate addition.
Peat with a moisture of 77.5% was mixed
with the mineral substrates in a ratio of 1:4
by volume. In the column experiment the
following substrates were used: sand, Polonite, blast furnace slag (BFS), and peat. Column K0 consisted of sand/peat, columns K1
and K3 were filled with Polonite/peat and
columns K2 and K4 with BFS/peat. Peat was
intended to prevent clogging because of
chemical reactions between sulphur and
calcium. Landfill leachate was transported
from the Tveta Landfill in Södertälje to the
laboratory in eight separate batches during
the experimental period. The leachate was
stored at a temperature of 4 oC and brought
in portions of 25 L to the column test, performed at room temperature. Leachate water
was distributed by a peristaltic pump through
Teflon tubes to each column with a flow rate
of 7 mL min-1, corresponding to a hydraulic
loading rate of 1.34 m d-1. The system was
operated intermittently for 8 h per day, and
each column received approximately 300 L of
leachate during the whole experiment. The
experiment was run under saturated conditions.
A long-term experiment aimed at overall
studies of elements (P, metals, and inorganic
forms of N) and their removal by six substrates was carried out indoors (temperature
20 oC) at the Loudden wastewater treatment
plant in Stockholm (Papers II, IV, V, VI).
The plant receives domestic wastewater and
is designed for a flow of 16 000 m3 d-1. The
raw wastewater was pumped to a
experimental
set-up
consisting
of
pretreatment units and columns imitating the
conditions of on-site treatment with filter
beds (Fig. 3). A two-chamber septic tank was
the first step, aimed at removing suspended
solids (SS). It was followed by a 2.5 cm thick
layer of mineral wool at the top of each
column. This wool was exchanged three
times during the experimental period. On
average, 0.5 L h-1 of wastewater was pumped
9
Agnieszka Renman
TRITA LWR PhD Thesis 1043
A
B
0.5 m
1
2
3
4
5
6
7
8
9
0.3 m
Fig. 3. The experimental set-up with pre-treatment system (A), distribution device (B) and
columns (1-9) filled with sand, opoka, Polonite®, limestone and blast furnace slags (Paper IV).
A synthetic solution with PO4-P and NH4-N
concentrations of 5 mg L-1 and 30 mg L-1
respectively was prepared by adding KH2PO4
and NH4Cl to 200 L of tap water. The solution was stored in a container and pumped
automatically to the top of each column three
times per day throughout the experimental
period of 68 weeks. The loading rate was
differentiated so that the pumped volumes
were roughly proportional to the pore volume of the particular filter material. Samples
were taken from influent and effluent weekly
during the first 40 weeks, thereafter biweekly.
The pH was analysed directly after collection.
The samples were then stored in a freezer at
-18 oC until analysis of PO4-P. The synthetic
solution and the loading schedule were selected to mimic conditions frequently encountered in on-site wastewater treatment
using reactive filter media. The concentrations of P and N represent concentrations
typically found in wastewater.
wastewater treatment plant as fertilisers in a
pot experiment.
Mitscherlich pots were filled (3.6 kg dry soil
per pot) with the A-horizon (5-20 cm) of Pdepleted soil from agricultural land under
permanent cultivation from Bjärröd, Scania,
Sweden. This sandy moraine soil is poor in P,
rather poor in K, and has an average level of
Mg.
Each pot received 1.5 g K2SO4, 1.0 g MgSO4
and 1.0 g N as NH4NO3 as basic fertilisation.
The soil was mixed with P in the form of
K2HPO4 or P sorbed from wastewater to the
different filter materials in quantities corresponding to 0.03 g P per pot. Each treatment
was performed in triplicate, with filter material from the column layer indicated (generally the surface, 0-5 cm).
Barley was sown and thinned after eight days
of growth to 29 blades per pot. The pots
were kept outdoors during the period JuneSeptember. They were watered once a day or
every second day to keep soil moisture between 60% and 80% of maximum waterholding capacity.
At harvest, the blades were cut 10 mm above
the soil surface, dried at 55 oC and weighed.
Pot experiment (Paper II)
Barley (Hordeum vulgare) was grown using the
P-enriched filter materials (opoka, sand,
Polonite, CSC, ASC, Limestone, CSVS) from
the column experiment at the Loudden
10
On-site wastewater treatment-Polonite and other filter materials for removal of metals, nitrogen and phosphorus
by the authors (Christoffersen et al., 1990).
These thermodynamic data should only be
considered as rough estimates, not least because the observed Ca:P ratios in the precipitates were between 1.28 and 1.38, which is
slightly less than the ideal ratio of 1.5. The
solubility data for OCP were also calculated
from experiments conducted by Christoffersen et al. (1990).
To investigate whether equilibrium with solid
phases might have controlled metal solubility
in the materials, Visual MINTEQ ver. 2.53
(Gustafsson, 2007) was applied for speciation
of the column effluents (Paper V). In these
calculations, complexation with dissolved
organic matter (DOM) in the effluents was
estimated with the Stockholm Humic Model
(Gustafsson, 2001), assuming that 70% of the
DOM consisted of fulvic acid. Because
measurements of alkalinity and dissolved
organic carbon (DOC) were made only for a
few samplings, the results should be considered approximate.
Soil pH was determined at sowing in a 1:2.5
(w:V) soil:water suspension.
Batch experiments (Paper III)
Batch experiments were carried out on both
unused filter material samples and on samples from the 0-5 cm layer of the used filter
materials in a column experiment (Paper III).
In brief, 2.00 g material (for wollastonite
samples 3.00 g) were suspended in 35 mL
solutions of variable composition in polypropylene centrifuge tubes and then equilibrated
for 5 d in an end-over-end shaker at 21 oC
and at a background ionic strength of 0.1 mol
L-1 (NaNO3 was used as supporting electrolyte). For the used filter materials wet samples
were used, which means that the actual liquid
to solid (L:S) ratios were larger; 56.6 for Filtra
P, 27.7 for Polonite, 27.3 for WCBFS and
12.0 for wollastonite. To produce a range of
pH values and PO4-P concentrations in the
suspensions, various amounts of nitric acid
(0-10 mmol L-1) were added. To some suspensions with 10 mmol L-1 nitric acid, an
additional 5-20 mmol L-1 oxalic acid were
added, to further increase the variation in pH
values and PO4-P concentrations. After
equilibration, the suspensions were centrifuged, pH was measured (using a Radiometer
combination electrode) and they were filtered
with 0.2 Pm single-use filters (Acrodisc PF).
Directly after filtration, the alkalinity of the
samples was measured through addition of
0.02 M HCl to pH 5.4.
Full-scale treatment system experiment
(Papers V, VI)
A reactive bed filter system for a one-family
house, situated 20 km NE of Uppsala city in
Sweden, was used for studies of treatment
efficiency. The reactive material used was
Polonite mixed with 8% peat (w/w). The
system consists of a pre-treatment step with
septic tank and biofilter followed by filter
well filled with Polonite (volume 800 L, dry
weight 560 kg), which receives wastewater
intermittently in relation to its production in
the household. After 2 years and 3 months of
operation, the filter material was removed
and exchanged with a new Polonite filter
material.
Modeling approach (Papers III, V)
The speciation of leachates from the batch
experiments was processed with Visual
MINTEQ (Gustafsson, 2006) using equilibrium constants for aqueous complexes (i.e.
CaHCO3+, CaPO4-, CaHPO40, etc.) from the
default Visual MINTEQ database, which
mostly relies on the NIST Critical Stability
constants, version 7.0 (Smith et al., 2003).
Calculated ion activity products were compared with solubility constants given in the
literature (Table 3 in Paper III). For ATCP,
for which few solubility data exist, constants
were calculated from raw data given for two
amorphous calcium phosphates of varying
crystallinity, referred to as ACP1 and ACP2
Chemical and physical analyses
The analysis of three forms of nitrogen
(NH4-N, NO3-N, NO2-N) and phosphorus
as PO4-P was performed using Flow Injection Analysis (FIA, Aquatec-Tecator
autoanalyser) (Paper I-VI). Before analysis,
the samples were filtered through a 0.45 Ƭm
Micropore filter. The pH and electric conductivity of liquids were measured by the
following instruments: Hanna HI 8424 mi11
Agnieszka Renman
TRITA LWR PhD Thesis 1043
concentrations in influent and effluent water.
Mass removal (g kg-1 dry substrate) of each
constituent entering and leaving the columns
was estimated from water quality and flow
data (Papers I-VI).
Fertiliser effectiveness can be summarised as
the relative effectiveness (RE%) of P bound
to the filter materials (PFM) in increasing
barley dry matter production in relation to
yield with the standard fertiliser (K2HPO4)
(Paper II). This was calculated according to
the following formula (1):
crocomputer pH meter or Radiometer PHM
82 and Hanna HI 8733 conductivity meter.
BOD7 was determined at the laboratory of
Stockholm Water AB, according to Swedish
Standards (SS028143-2 mod. and SS-EN
25814-1) (Papers II, IV, V).
The liquids intended for metal analyses were
preserved with a few drops of concentrated
HNO3 and kept in a cold-storage room at 4
o
C prior to analysis (Papers I, V). The analyses were performed using ICP-AES (Inductively Coupled Plasma Atomic Emission
Spectrometry) for all elements. Metal concentrations in the virgin materials, in the
liquids and in the solids from 0-5, 5-10, 1020, 20-30, 30-40, and 40-50 cm layers of each
column were determined using inductively
coupled plasma atomic emission spectrometry (ICP-AES) (Paper V).
Carbon and N were determined by dry
combustion on a LECO combustion system
coupled to an IR-detector (LECO, 1995)
(Paper II). Total element contents of carbon,
and N in individual layers of the different
materials were determined by extraction with
conc. HNO3 in an autoclave (120 oC for 30
minutes, material:solution 1:20; modified
from SIS (1986) (PaperII).
Total concentrations of P were determined
in three layers (0-5, 5-10, 40-50 cm, Paper
III) or in random samples (Paper VI) of the
used filter materials using ICP-OES. Samples
from different layers (0-5, 5-10, 40-50 cm) in
columns studied with synthetic solution were
air-dried, gently crushed in a mortar, and
then examined by a Perkin-Elmer S2000
Fourier Transform infrared (FTIR) spectrometer. The instrument was run in the
attenuated total reflectance (ATR) mode
using a Golden-Gate diamond cell. The
spectra were compared with those recorded
for filter material samples that were not used
in the column experiment. Finally, total
dissolved Ca, Mg, and Si were determined by
plasma emission spectroscopy using a JobinYvon JY24 ICP instrument (Paper III).
RE %
>Y with PFM Y control @ 100
>Y with K 2 HPO4 Y control @
where Y represents the yield.
The percentage removal of inorganic PO4-P
was calculated as the difference between
influent and effluent concentrations for the
samples collected at the same time. The sorption of PO4-P to the filter matrix (Sp, mg kg1
) was calculated using the following mass
balance equation (Papers II, III, IV):
Sp
Ci Ce V
(2)
m
where Ce is the effluent concentration and Ci
is the influent concentration, V is the volume
of wastewater treated during the experiment
and m the mass (kg) of the filter material.
The removal capacity of the filter materials
(Rm, %), i.e. the retention or leaching of
dissolved elements during the infiltration of
wastewater, was calculated according to the
following equation (Paper V):
Rm
§ Ce
¨¨1 © Ci
·
¸¸ 100
¹
(3)
where Ce is the effluent concentration and Ci
is the influent concentration.
Evaluation of treatment performance
The percentage removal efficiencies of P,
heavy metals and N by the column substrates
were calculated as the difference between
The sorption to, or release of, dissolved
elements (mainly metals) from the filter ma-
12
On-site wastewater treatment-Polonite and other filter materials for removal of metals, nitrogen and phosphorus
trix (Sm, mg kg-1) was calculated using the
following mass balance equation (Paper V):
Sm
Ci C e V
R ESULTS AND DISCUSSION
Phosphate removal and pH development
in the columns
Filter materials in the column experiment
using synthetic solution were exposed to a
high loading rate (> 400 L m-2d-1) (Papers III,
VI) under saturated conditions. The columns
received between 1.1 to 1.8 m3 of solution
during 68 weeks. Technical problems with
pumping caused less discharge of solution
than was intended. The column filled with
Filtralite P clogged after 971 pore volumes
(pv) and was closed for that reason. Its pore
volume also decreased and was found to be
0.45 dm3 less than at the start.
Clear material-specific differences in PO4-P
removal were observed during the course of
the experiment. Wollastonite was the least
efficient material in terms of PO4-P removal;
on average 51.1% of P was removed. This is
partly explained by the low pH in these columns; already after 100 pore volumes, the
pH of the effluent was more or less equal to
that of the inflow. Water-cooled blast furnace
slag (WCBFS) removed more than 95% P
during the first 300 pore volumes, but thereafter the P retention efficiency decreased
steadily, so for the whole time period the
mean P removal was 85.6%.
The two most strongly P-retaining materials
were Polonite and Filtra P. Filtra P was
slightly more efficient; on average it retained
98.2% P, whereas Polonite retained 96.7%.
However, Polonite treated much more solution (1410 pv) than Filtra P. Disintegration of
Filtra P was observed, particularly in the
beginning of the experiment, as a yellowbrown colour in the effluent. In both cases
the pH remained alkaline, particularly so for
Filtra P. Before termination of the experiment, there was a tendency for the pH in the
Polonite column to decrease, which may
suggest that the most reactive CaO phase had
dissolved completely.
A column filled with Polonite and run in
parallel under intermittent saturated flow
conditions had a slightly lower removal
(96%). Obviously, the two modes of flow
conditions did not have an influence on the
removal capacity.
(4)
m
where Ce is the effluent concentration and Ci
is the influent concentration, V is the volume
of wastewater treated during the experiment
and m the mass (kg) of the filter material.
The total metal content (mg kg-1) of filter
materials in the columns was calculated as the
average of the content in each layer. Analyses
were performed in duplicate for the 0-5 cm
layers (Paper V).
The removal capacity of the filter materials,
i.e. loss of organic matter (BOD7) and of N
during the infiltration of wastewater, was
calculated according to the following equation (Paper IV):
ª C º
Loss = «1 e » 100
¬ Ci ¼
(5)
where Ce is the effluent concentration and Ci
is the influent concentration.
13
Agnieszka Renman
TRITA LWR PhD Thesis 1043
At the end of the experiments the concentration of PO4-P was still very low in the effluent from columns filled with Polonite, both
under saturated and intermittent saturated
conditions (0.13 and 0.17 mg L-1), respectively). However, the effluent concentration
started to increase by the end of the experiment, as did the pH, indicating a beginning
of breakthrough (Fig. 4).
Only Polonite showed good PO4-P removal
efficiency for the whole experimental period
using municipal wastewater as feed solution
in columns (Paper II). Effluent from column
3 and 7, filled with Polonite (Fig. 5), had a
low mean concentration of P of 0.1 mg L-1
throughout the experiment and a simultaneous decrease in pH from 12.8 to 8.9. The P
removal capacity decreased successively in
the other columns studied and, at the end of
the experiment, reached P concentrations in
the effluent higher than 2 mg L-1 (Fig. 6). An
exception was the column filled with sand,
which did not remove soluble P at all from
the incoming wastewater. The pH in the
effluent from the sand column closely
matched the influent wastewater pH, while
effluent from the other columns had a pH of
8.9-9.2.
Effluent
Influent pH
Effluent pH
6
14
5
12
10
4
pH
8
3
6
2
4
68
60
50
40
0
30
0
20
2
10
1
1
Concentration PO 4-P (mg/l)
Influent
The difference in retention capacity between
Polonite and opoka was large, more than tenfold, due to oxidation of CaCO3 to the more
reactive form CaO in Polonite. High temperature thermal pretreatment of opoka also
results in sintering and a subsequent large
decrease in the surface area of the product
Polonite®.
The comparably low sorption capacity of
limestone is also explained by its Ca being in
the carbonate form. Burning limestone
(CaCO3) to form quick lime (CaO) increases
the removal capacity (Hylander & Simán,
2001), but results in a powder with a hydraulic conductivity too low to permit an use in
the designed filtration system in its pure
form.
In addition to element content and chemical
form, the differences in the particle size of
the materials can also explain the variations
in P sorption that are clearly apparent when
comparing the coarse and very coarse crystalline slags. The process of carbonation should
also be considered here. In this process,
carbonation may occur in a gas-solid-liquid
or gas-solid condition (Huijgen et al., 2006).
Weeks
Fig. 4. Changes in pH and influent and effluent concentrations of PO4-P for column
run under saturated flow and with artificial solution over a 68-week period.
14
Effluent 1
Effluent 2
Influent pH
Effluent 1 pH
Effluent 2 pH
3
6
2
4
1
2
0
0
67
8
47
4
40
10
35
5
29
12
23
6
8
14
pH
Influent
7
1
Concentration PO 4-P (mg/l)
On-site wastewater treatment-Polonite and other filter materials for removal of metals, nitrogen and phosphorus
Weeks
Fig. 5. Changes in pH and influent and effluent concentrations of PO4-P for two columns
filled with Polonite and fed with municipal wastewater over a 67-week period.
Influent
CSVC
ASC
Opoka
Limestone
Sand
Polonite 2
Polonite 1
CSC 2
CSC 1
Concentration PO4-P (mg/l)
7
6
5
4
3
2
1
67
47
40
35
29
23
8
1
0
Weeks
Fig. 6. Long-term development of PO4-P concentrations in effluents from nine columns
with different filter materials fed by municipal wastewater.
15
Agnieszka Renman
TRITA LWR PhD Thesis 1043
Phosphate removal and pH development
in full-scale bed filter
The full-scale on-site system investigated
showed an overall removal of 89% for a
period of 92 weeks. The phosphate breakthrough curve for this trial is presented in
Fig. 7. After 128 pore volumes had been
treated in the bed filter in downflow mode,
the Ce/Ci ratio started to rapidly increase,
coinciding with the pH drop seen in Fig. 8.
After 75% breakthrough, the rate of increase
of the breakthrough curve seemed to even
out to the end of the experimental period.
The effluent PO4-P concentration was 2.49
mg L-1 by then. The filter well was operated
for an additional 4 months with the same
Polonite material after termination of this
experiment and analyses of a few samples of
the effluent showed that complete breakthrough (Ce=Ci) was not achieved in this
period. This suggests that PO4-P removal
continued, possibly by precipitation and/or
biological mechanisms.
Both processes occur in the surface layer of
the filter materials while the aqueous carbonation route appears in the deeper layers,
i.e. leaching of Ca, dissolution of CO2 and
subsequent conversion of (bi)carbonate
species, followed by nucleation and growth
of CaCO3.
There were no operational problems such as
clogging causing hydraulic failure and overflow in either experiment. Chemical clogging
was expected from previous experience (unpublished data) with Polonite and problems
with clogging have been observed in many
experiments due to formation of precipitates
in the outflow system (e.g. Drizo et al., 2006),
so a small amount of peat was incorporated
to prevent this in the experiment with synthetic solution.
However, the Polonite filters fed with municipal wastewater performed excellently
under the particular hydraulic load despite
being peat-free.
0,8
Ce/Ci
0,6
0,4
0,2
17
3
15
0
12
5
10
0
75
50
25
1
0
Pore volume
Fig. 7. Phosphate breakthrough curve for Experiment III (part with down-flow feeding of filter).
16
On-site wastewater treatment-Polonite and other filter materials for removal of metals, nitrogen and phosphorus
Effluent
Influent pH
Effluent pH
12
14
10
12
10
8
pH
8
6
6
4
4
74
57
17
31
0
23
0
12
2
7
2
1
Concentration PO4-P (mg/l)
Influent
Weeks
Fig. 8. Changes in pH and influent and effluent concentrations of PO4-P in a filter well used for
domestic wastewater treatment. The points represent the last 74 weeks of the sampling period
(total 92) when the filter was fed in downward flow.
et al., 2004). Total P content at the surface
and in the 5-10 cm layers of the columns
filled with Polonite was nearly the same and,
due to sorption of phosphate, was more than
six times higher than in the bottom layer,
which contained only P from the initial material without any P sorbed (Fig. 9). The similar
P content in the two upper layers and the
continuous decrease in lower layers indicate
that only the surface layers were saturated or
close to saturated with P. The absence of P
retention in the bottom layer and only limited
P sorption in the two overlying layers indicate that the column would have been able to
efficiently retain P from the wastewater for
twice as long a period of time.
In the case of columns operated with synthetic solution, the concentrations also decreased significantly with depth of the column, although not clearly in the column filled
with wollastonite (Paper III).
The discharge of liquid was much higher to
these columns compared with those using
municipal wastewater (approx. 6 times).
Phosphorus sorption to different materials by layers
The top layer (0-5 cm, 0-10 cm) of each
material showed the highest concentration of
accumulated P in column and full-scale experiments. The concentration decreased
significantly with depth of the columns; this
was especially the case in the Filtra P and
Polonite materials. The columns fed by municipal wastewater and filled with Polonite
increased from an original 0.2 mg g-1 to 1.3
mg P g-1 dry matter in the top layer. These
were followed by sand, which had a higher P
content in its surface layer than revealed by
its capacity to remove phosphate from percolating wastewater. However, P content decreased significantly in underlying layers of
sand, which should not be the case with an
effluent rich in phosphate, where the total P
content originated mainly from sorbed phosphate. Phosphorus from the wastewater was
retained in the surface layer by the sand acting as a filter due to its fine texture and by
bacteria living in the surface layer of the
column rather than in lower layers (Renman
17
Agnieszka Renman
TRITA LWR PhD Thesis 1043
Depth (cm)
Polonite
0-5
5-10
10 20
20 30
30 40
40 50
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7
-1
P (mg g )
Depth (cm)
Sand
0-5
5-10
10 20
20 30
30 40
40 50
0.0
0.1
0.2
0.3
P (mg g-1)
0.4
0.5
Fig. 9. Initial P content of the Polonite and sand (dark shaded area) and P enriched to different layers (light shaded area) after the column experiment. Bars indicate ±SE for duplicate
columns (Paper II).
The total concentration of P was also much
higher, as indicated by Polonite showing a
value of 7.39 g kg-1 dry matter. The highest
concentration of accumulated P was found
for Filtra P, approaching 19.4 g kg-1 dry matter. However, 12% of the dry matter was lost
from the column due to the release of gypsum from the material.
The Polonite used in the filter well showed a
P concentration of 1.27 g kg-1 dry matter.
Significantly lower concentrations were
found in other layers of the filter bed.
This can indicate that the wastewater was not
evenly distributed in the material, despite the
filter bed being operated by a batch-wise
filling-up and drainage mode.
Ádám et al. (2006) studied the spatial distribution of P in horizontal flow, Filtralite P systems. They concluded with the help of a Brtracer experiment that preferential flow could
appear in the upper parts of the containers
used. Drizo et al. (2002) showed that when
electric arc furnace slag was drained and left
to rest for four weeks, it was able to increase
its P removal capacity by 74%.
18
On-site wastewater treatment-Polonite and other filter materials for removal of metals, nitrogen and phosphorus
By this regeneration method the P saturation
value was increased from 1.35 to 2.35 g of P
kg-1. None of the experiments carried out in
this thesis used such long resting periods for
the materials tested. Completely drained
conditions occurred only for hours up to
days in the experiments performed in Papers
II and VI. More studies are needed on the
effects of resting periods and how the contact between solute and solid can be enhanced, as well as how to detect preferential
flows. The latter can be performed by using
stable isotopes and tracers (Ronkanen &
Kløve, 2007).
An NH4-N effluent concentration of below 1
mg L-1 was recorded rather soon after the
start of the experiment (Paper IV). The
temporal variation in NH4-N concentration
in the effluent of CSVC slag represents the
overall pattern found, with the exception of
the two Polonite columns (3 and 7). In these
latter columns Polonite exhibited a much
slower
development
of
NH4-N
transformation and stable, low effluent
values as in the other columns were not fully
reached.
However NH4-N was reduced to less than 7
mg L-1 after the wastewater had percolated
through the Polonite filter material for a few
weeks of operation (Fig. 10). The
concentration of NO2-N in influent
wastewater was 55.5 μg L-1 (range 1.8-290.3
μg L-1) and it remained at this low level in
effluents from all columns except for those
filled with Polonite, where it was 2.93 mg L-1.
Nitrate was also quite low in the influent
wastewater (0.25 mg L-1, range 0.013-0.49 mg
L-1, but on the other hand increased
markedly after passing through the materials,
resulting generally in concentrations of
between 0.03 and 87.6 mg L-1 in the
effluents.
The observed decrease in effluent NH4-N
concentration in columns 1, 2, 4, 5, 6 and 8
was obviously attributable to microbial
immobilisation
and
nitrification,
i.e.
transformation processes rather than
removal. At pH values above 9.3, NH4+ can
be converted into NH3. This NH3
volatilisation could have occurred in the
columns with Polonite and amorphous and
crystalline slag, i.e. columns 3, 4, 5, 7 and 9.
Transformation and removal of nitrogen
The N dynamics were monitored in the
experiment using municipal wastewater for a
period of 67 weeks (Paper IV). The removal
performance of NH4-N was over 90% in all
columns regardless of filter material. Only
two materials (Polonite and CSVC) were able
to reduce the total inorganic nitrogen (TIN)
content over the whole experimental period,
by 17.7% and 9.8% respectively (TIN = Ɠ
NH4-N, NO3-N, NO2-N).
The other filter materials leached NO3-N and
NO2-N, particularly the amorphous slag.
Three filter media (BFS, Polonite, sand) were
tested for their ability to remove nitrogen
from landfill leachate. As shown in Table 1,
TIN was reduced in various amounts by the
media used.Polonite achieved the highest
removal efficiency (average 18%).
The concentrations of NH4-N and total
inorganic nitrogen in the influent wastewater
were on average 26.6 ±5.5 mg L-1 and 26.9
±12.6 mg L-1, respectively. The influent
landfill leachate had the corresponding values
of 104.5 ±17 mg L-1 and 105.9 ±29 mg L-1.
Table 1. Total inorganic nitrogen (TIN = ƓNH4-N, NO3-N, NO2-N) in influent, and sorption
and removal efficiencies by three media (Paper I).
TIN
Influent
(mg L-1)
105.9±29
Sorption (g kg-1)
Sand Polonite BFS
0.25
1.9
0.55
19
Sand
4±22
Removal (%)
Polonite
BFS
18±15
8±22
Agnieszka Renman
TRITA LWR PhD Thesis 1043
pH C 3
pH C 7
Effluent C 3
Effluent C 7
8
14
7
12
10
5
8
4
6
3
4
2
67
60
0
45
0
30
2
15
1
1
pH
NH4-N mg/l
6
Weeks
Fig. 10. Temporal variation in the ammonium concentration at the exit of the Polonite columns
(C3, C7) and pH of the effluent (influent concentration 26.6 ±5.5 mg L-1; influent pH range 8.189.28). The samples represent the whole experimental period of 67 weeks (Paper IV).
The total N content was highest in the 0-5cm layer of all materials, as was found for P
too (exemplified in Fig. 11). This was accompanied by a high total carbon (C) content
particularly in the surface layers of sand,
Polonite and crystalline slag.
The highest C content in the whole filter
mass at termination of the experiment was
found for opoka. The vertical N distribution
differed between materials. The differences
were obviously related to particle size distribution and thereby to hydraulic conductivity
of each material.
Pell et al. (1990) clearly showed a general
pattern of decreasing numbers of bacteria
with depth in a sand filter, with bacteria
constituting a large part of the biomass in the
surface layer. This in fact was probably true
for all the filter materials studied here but was
more accentuated in the most alkaline types.
In the case of Polonite, bacterial activity and
mineralisation could have been expected only
in the surface layer, where pH was lower. The
loss of about 18% TIN from Polonite columns during the experiment with municipal
wastewater was caused by volatilisation rather
than denitrification.
This was also probably true for the loss observed in the columns feed by landfill
leachate. However adsorption and cation
exchange are also possible explanations for
that removal. A comparison of TIN in the
effluent when pH was above and below 9
revealed that the loss was associated with the
high pH range. The reactive filter materials
obviously create an alkaline environment that
inhibits denitrification even in the superficial
layer, where the biofilm could create a biologically active part of the filter. Unfavourable conditions for denitrification in sand and
limestone were caused by the mode of intermittent discharge of wastewater to the columns and the unsaturated flow. This was
intended to create oxygenated conditions in
the matrix of each material throughout the
whole experimental period.
It is known that the presence of competing
ions and organic matter can significantly
reduce the removal of NH4-N (Chen et al.,
2002) but the role of organic carbon for
nitrogen removal in reactive bed filter materials has to be further investigated.
20
On-site wastewater treatment-Polonite and other filter materials for removal of metals, nitrogen and phosphorus
Depth (cm)
Sand
0-5
5-10
10 20
20 30
30 40
40 50
0.0
0.2
0.4
0.6
0.8
-1
N (mg g )
Depth (cm)
Polonite
0-5
5-10
10 20
20 30
30 40
40 50
0.0
0.2
0.4
0.6
-1
N (mg g )
Fig. 11. Initial N content in the sand and Polonite® (dark spotted area) and N enriched to
different layers (light spotted area) after 67 weeks of wastewater treatment by infiltration. Bars
indicate ±SE where columns were duplicated (Paper IV).
Metal removal
The metal removal capacity of the filter materials was assessed in two experiments (Papers I, V). In Paper I, the purpose was to test
materials able to remove metals from landfill
leachate, while Paper II analysed the metal
removal by filter materials aimed for P removal and recycling from domestic wastewater.
Polonite, sand and BFS, all of them mixed
with peat, were used for treating landfill
leachate. The best performance was found
for Polonite, where Mn, Fe, Zn and Cu were
removed to 99%, 93%, 86% and 67%, respectively. Breakthrough curves are presented in Figs. 12 and 13.
Hydroxide precipitation was suggested as the
process for the high removal efficiency of
metals by Polonite, forming insoluble precipitates in the bed filter. The precipitation is
primarily dependent upon two factors,
namely the concentration of the metal and
the pH of the water. Chemical treatment
plants normally operate at a pH of approximately 9 when multiple metals are present.
The superior removal capacity of the Polonite filter is probably a combination of several factors, among which precipitation is the
most important. The BFS showed good
removal efficiency for Cu (66%), Ni (19%)
and Mo (16%). Sand did not demonstrate a
promising removal capacity for any of the
elements studied with the exception of Cu
(25%).
21
Agnieszka Renman
TRITA LWR PhD Thesis 1043
pH K1
pH K3
pH in
Zn
Mn
Cu
Ni
14
1.5
1.25
10.5
0.75
pH
Ce/Ci
1
7
0.5
3.5
0.25
0
0
27
40
60
97
130
160
202
238
Pore volume
Fig. 12. Breakthrough curves (Zn, Mn, Cu, Ni) and pH data for effluent from columns filled with
Polonite. Ce/Ci = ratio for effluent and influent mean concentration (Paper I).
pH K2
pH K4
pH in
Zn
Mn
Cu
Ni
9
1.4
1.2
1
0.8
pH
Ce/Ci
6
0.6
3
0.4
0.2
0
0
27
40
60
97
130
160
202
238
Pore volume
Fig. 13. Breakthrough curves (Zn, Mn, Cu, Ni) and pH data for effluent from columns filled with
BFS. Ce/Ci = ratio for effluent and influent mean concentration (Paper I).
22
On-site wastewater treatment-Polonite and other filter materials for removal of metals, nitrogen and phosphorus
Leaching of several elements occurred from
the filter media (e.g. Mn from BFS). Most
pronounced was the release of silica from
Polonite and BFS. Calcium was also leached
from Polonite but after 60 pore volumes of
treated leachate, the release of Ca decreased,
while Si release rapidly increased.
The changes in release of Si and Ca versus
normalised flow for the Polonite can be
related to the weathering of the solid material
and the presence of wollastonite. According
to other studies, the latter material is efficient
for heavy metal removal from aqueous solution (Ni, Pb, Cd) (Sharma et al., 1990; Yadava
et al., 1991; Sharma, 1995). The removal of
different elements was suggested to be a
combination of several factors, e.g. precipitation, ion exchange and adsorption.
The apparent propensity for accumulating
metals from domestic wastewater followed
the order Polonite > ASC > opoka > CSC >
limestone > CSVC > sand. All materials
removed Fe with an efficiency of around
60%. This element was actually the only one
removed by the sand. Zinc was removed by
all materials, except the sand, in the range 5383%. Only the three blast furnace slags (ASC,
CSC, CSVC) were able to remove Ni. Polonite removed over 98% of dissolved Mg and
Mn in the wastewater, while fairly efficient
removal of Cu was demonstrated by the
ASC. Analyses of metal concentrations in the
leachate (effluent) from columns showed
significant (p<0.05) release, i.e. an increase in
concentration, of some elements from sand,
CSC, ASC and CSVC.
Sand released Cu and Zn, while CSC and
ASC showed increased concentrations of Mg.
The very coarse crystalline slag (CSVC) exhibited a large loss of Mn, although only in
the beginning of the experiment.
The wastewater trickles through the filter
matrix and colloidal matter can attach to the
surface of each mineral particle (Yao et al.,
1971). Bacteria create a biofilm where microbial adsorption of heavy metals can occur
(Goyal et al., 2003) and where adsorption and
precipitation processes can also retain the
metals. The increasing concentrations of
dissolved metals in the effluent can be the
result of leaching from the materials.
The blast furnace slag material, from which
Mg and Mn were obviously released, demonstrates this in both experiments.
The presence of peat in the trial with landfill
leachate had an effect on the removal of
several metals. The sand did not remove
metals at all from domestic wastewater.
It is obvious from Table 2 that wide variations occurred in metal concentrations,
probably caused by the sources of pollution
(Moriyama et al., 1989;Vinnerås et al., 2006).
In fact, metal concentrations in influent were
lower than typical effluent discharge limits
with the exception of Cu (Metcalf & Eddy,
2003). Thus the wastewater used in the column experiment had very low dissolved
metal concentrations, similar to those expected to be found in wastewater from single
houses.
Table 2. Influent and column effluent concentrations of elements. The units are Ƭg
erwise stated (average, SD).
Element
Influent
Ba
Cr
Cu
Mg*
Mn
Ni
Zn
62.2±21.3
1.71±0.52
16.2±10.8
6.26±0.66
10.9±3.1
4.64±1.6
37.5±11.8
** Average
L-1, unless oth-
Column effluent
Sand
Opoka
Polonite**
68.4±12.6
2.71±3.52
26.4±9.6
6.88±1.86
31.2±32.2
10.1±5.7
221.3±164.5
39.7±12.3
1.51±0.34
18.1±6.4
7.68±1.25
4.21±4.0
5.52±1.58
11.8±5.5
90.4±24.7
26.8±32.6
13.6±7.4
0.09±0.06
0.24±0.19
6.37±4.15
14.2±10.2
and SD for duplicate materials
23
CSC**
34.7±8.9
1.68±0.53
15.01±14.1
32.1±5.99
29.3±65.3
3.26±0.77
6.57±1.96
ASC
CSVC
Limestone
45.3±9.9
1.55±0.5
11.6±4.8
28.4±10.3
1.77±2.15
3.11±0.49
7.47±1.63
32.6±7.6
1.75±0.52
15.7±5.5
7.26±0.86
82.6±203.1
3.48±1.34
9.54±2.86
53.7±26.9
1.72±0.59
15.0±8.6
26.3±7.03
6.04±6.7
5.79±5.01
17.36±23.5
Agnieszka Renman
TRITA LWR PhD Thesis 1043
tion/precipitation to sulphuric compounds in
blast furnace slag may also be important.
However, sulphur was leached in great
amounts from blast furnace slag columns in
the beginning of the experiment with municipal wastewater (Paper V), as previously
observed in the column experiment with
landfill leachate (Paper I).
The dissolution of alkaline minerals, particularly in the case of Polonite, can increase the
pH of the percolating wastewater above
solubility point, which causes metals to precipitate, probably as metal oxides and metal
carbonates (Petrovic et al., 1999).
The organic ligands present in wastewater
can bring about either enhanced, suppressed
or unaffected adsorption of metal ions on
clay minerals (Abollino et al., 2008); at low
pH an enhanced effect is often seen, whereas
at neutral to high pH (as in the present investigation) DOM is likely to suppress metal
sorption, especially for Cu. Jönsson et al.
(2006) studied the sorption processes in the
goethite system and found that the presence
of Cu(II) resulted in a increased adsorption
of DOM at high pH.
Phosphorus
Unused and spent filter materials (Filtra P,
Polonite, WCBFS, wollastonite) were characterised using ATR-FTIR (Paper III). All four
unused samples contained peaks at a1420
and 870 cm-1, characteristic of the carbonate
anion in calcite (Fig. 14).
The treatment preceding the filtration
through the column materials probably removed the majority of metal species. These
can be trapped by gravity settling of suspended solids in a primary clarifier, e.g. septic
tank (Artola et al., 1997; Wang et al., 2006).
Removal mechanisms- heavy metals and
phosphorus
The studies on removal mechanisms for
metals and P are presented and discussed in
the following.
Heavy metals
The filter materials used in the experiment
with municipal wastewater removed heavy
metals (Paper V). According to speciation
calculations with Visual MINTEQ, a very
large fraction of dissolved copper in the
column leachates was bound to DOM (the
median value was 99.99%). The corresponding figures for Cr(III) (assuming that all dissolved Cr existed as trivalent chromium), Zn
and Ni were 93, 59 and 24% respectively.
Other ions such as Mn, Ca and Ba were less
strongly bound to DOM. Dissolved chromium was in most cases slightly lower than
predicted by equilibrium with Cr(III) hydroxide. For copper, the results suggest that CuO
or a similar phase in the Polonite columns,
but not in the other columns, might have
controlled the dissolved Cu. Dissolved Zn
was very much undersaturated with respect
to any Zn mineral phase, suggesting that the
processes controlling Zn are likely to involve
adsorption/desorption and weathering processes. Lead was far from equilibrium with any
mineral phase. Magnesium was close to equilibrium with brucite in the Polonite columns
early in the sampling period, suggesting that
Mg removal by Polonite involved the precipitation of this mineral. However in the latter
part of the experiment there was undersaturation also in the Polonite leachates, despite a
continued very strong Mg removal in these
columns (> 95% throughout the experiment).
The mechanisms involved in metal retention
by blast furnace slag are thought to be ion
exchange with calcium on particle surfaces
and precipitation on Al(OH)3 and SiO2
(Dimitrova & Mehandgiev, 2000). Sorp-
The presence of wollastonite (E-CaSiO3) in
both Polonite and natural wollastonite was
indicated by peaks near 1056, 960, 902 and
680 cm-1 (Atalay et al., 2001). The observation
that wollastonite is an important constituent
of Polonite is in agreement with earlier studies employing chemical dissolution techniques and X-ray diffraction (Eveborn, 2003;
Brogowski & Renman, 2004).
The unused Filtra P sample was dominated
by a number of peaks related to sulphate
minerals such as gypsum and ettringite.
These minerals were present in large amounts
in the unused sample but had disappeared
completely in the sample that was in prolonged contact with the synthetic solution.
24
On-site wastewater treatment-Polonite and other filter materials for removal of metals, nitrogen and phosphorus
Filtra P 0-5 cm
Polonite 0-5 cm
A
A
Filtra P 15-20 cm
Polonite unused
Filtra P unused
4000
3200
2400
1800
1400
1000
600
4000
3200
2400
cm-1
1800
1400
1000
600
cm-1
Fig. 14. ATR-FTIR spectra of used and unused Filtra P and Polonite.
Other calcium phosphate compounds were
not indicated with the exception of amorphous tricalcium phosphate, which appeared
as a single peak near 1025-1030 cm-1.
In an earlier study with other filter materials
in which equilibrium was obtained only from
supersaturation, it was hypothesised that
hydroxyapatite (HAp) was formed (Johansson & Gustafsson, 2000). However, the data
presented here are not in agreement with this
hypothesis, because PO4-P in the used filter
materials was more soluble than would be
expected from this equilibrium. Most of the
samples were supersaturated with respect to
HAp. Two other candidates in controlling
PO4 solubility are DCPD (dibasic calcium
phosphate dihydrate; CaHPO4u2H2O) and
DCP (dibasic calcium phosphate, CaHPO4).
However, most samples were undersaturated
with respect to these phases, although the
most acid Filtra P samples seemed to be in
equilibrium with DCP or DCPD. At high
pH, however, the samples were far from
equilibrium, indicating that the results could
not be resolved easily in terms of equilibrium
with respect to DCP or DCPD. Many of the
samples were relatively close to the solubility
line of OCP (octacalcium phosphate,
Ca4H(PO4)3), indicating that OCP might be
involved in determining PO4-P solubility.
Finally ATCP, with stoichiometric composition Ca3(PO4)2, may also control PO4-P solubility.
As shown in Fig. 15, many samples were
close to equilibrium with the less noncrystalline ATCP termed ACP2 (Christoffersen et al., 1990). Moreover, the pH dependence of the ion activity products was in
better agreement with ACP2 than with OCP,
which suggested that solubility control by
ACP2 (or ATCP) was slightly more likely
although a prominent role for OCP could
not be ruled out from these observations
only.
The observation that a soluble Ca-P phase,
probably ATCP, had accumulated in the filter
materials suggests that at least part of the
PO4-P is readily available to plants. However,
only between 3.5 and 18% of the accumulated PO4-P was readily dissolved in the
batch experiments. There may be several
reasons for this, but one possibility is that
part of the PO4-P had crystallised to a slightly
less soluble phase (such as HAp). Thus more
detailed spectroscopic studies need to be
performed to fully elucidate the solid-phase
speciation and reactivity of the accumulated
phosphates.
Recycling of filter materials back to agriculture
The possibility of recycling used filter materials together with their accumulated contents
of P and N back to agriculture as fertilisers
and soil amendments was of interest in several of the studies (Papers II, III, IV, VI).
25
Agnieszka Renman
TRITA LWR PhD Thesis 1043
-38
-35
-39
-41
OCP
-39
Filtra P
WCBFS
Wollastonite
Polonite
4log{Ca 2+}+3log{PO 43-}
5log{Ca 2+}+3log{PO 43-}
-37
-43
-45
-47
-49
-51
HAp
-40
-41
-42
-43
-44
Filtra P
WCBFS
Wollastonite
Polonite
-45
-53
-46
-55
6
7
8
9
6
10
7
9
10
-24
-10
DCP
-25
3log{Ca 2+}+2log{PO 43-}
DCPD
-11
log{Ca 2+}+log{PO43- }
8
pH
pH
-12
-13
-14
Filtra P
WCBFS
Wollastonite
Polonite
-15
ACP1
-26
Filtra P
WCBFS
Wollastonite
Polonite
-27
ACP2
-28
-29
-16
-30
6
7
8
9
10
pH
6
7
8
9
10
pH
Fig. 15. Solubility diagrams for the 0-5 cm layer of the used filter materials. The points represent
the calculated solution activities in the batch experiment extracts. Data points above the lines
indicate supersaturation.
A particular column experiment was designed
for testing the capacity of materials for use
as fertilisers (Paper II). After termination of
the experiment, the P-enriched filter materials were tested for their fertiliser effectiveness
in a pot experiment where barley (Hordeum
vulgare L.) was cultivated. The application of
0.03 g P per pot as K2HPO4 produced the
highest yield, closely followed by P sorbed to
slags from the wastewater filters (Fig. 16).
This can be attributed to a fast release of
loosely bound P from the amorphous and
crystalline slags used in our experiment.
Yields from treatments with P sorbed to
Polonite were significantly lower. Yields with
P sorbed to opoka were even lower, despite
the application of K2HPO4, but higher from
soils that had not had fertiliser added (Fig.
16).
This indicates that P sorbed to opoka can be
utilised by barley, although not as readily as P
sorbed to slag or P in chemical fertiliser. The
low diffusion of P, as previously reported for
P transport in soils (e.g. Hylander & Ae,
1999), may explain the reduced availability of
P in opoka compared with slag. In this regard, Polonite represents a condition somewhere between opoka and slag.
This is because it originates from opoka and
has been thermally treated to change its physico-chemical properties, thereby activating its
sorption capacity (Brogowski &Renman,
2004).
In the sand filter, the fertiliser effectiveness
of accumulated P was also lower when compared with P applied as K2HPO4. The sand
was a washed quartz sand, which is inert.
26
On-site wastewater treatment-Polonite and other filter materials for removal of metals, nitrogen and phosphorus
14
100 %
12
76 %
-1
Yield (g DM pot )
71 %
10
8
34 %
35 % 44 %
20%
24 %
-19%
9%
6
4
2
Ca
CO
3
p
Ca a >
CO >B
3 S
O pa B
po
S
k
Po a vi
r
lo
ni gin
te
v
A irg
SC in
v
CS irg
C in
vi
rg
i
Co n
nt
Sa rol
nd
Fe CaC 0-5
rtP+ O3
Ca 0-5
CO
O 3 pa
p
Po oka
lon 0-5
i
Po te 0
lo
n 5A
Po ite 0
lon - 5
ite B
20
A - 30
SC
0
CS -5
C
05
Fe
rt P
0
Fig. 16. Yields (average and 95% confidence interval) of barley grown in soil treated with different filter materials containing P (white stacks) or with materials without P (gray stacks). P
fertiliser effectiveness (%) in relation to standard P fertiliser is indicated by figures above the
stacks (Paper II).
Therefore the P in the sand filter, retained
because of the physical filtering capacity of
sand, was expected to be mainly in organic
form. This filtering was confirmed by increased contents of C and N in the surface
layer (Fig. 17). Compared with inorganic
orthophosphate, organic P had a significantly
lower plant availability in the short term (i.e.
days to weeks) when applied to the studied
soil.
This can also explain the difference in fertiliser effectiveness between the surface layer
and deep layer Polonite (Fig. 16).
Despite the fact that P content in the surface
layer was twice that of other horizons, yield
and fertiliser effectiveness were higher for
barley grown in the deep layer material. All
filter media probably retained organic P at
the surface. The pH of the material from the
surface layer was lower than that from
deeper layers. Plant production can be affected by the pH of filter materials as shown
by Cucarella et al. (2007).
As demonstrated in Figure 16, relative effectiveness was negative or low for P sorbed by
limestone and opoka. This was caused by the
alkaline characteristics of these materials and
the large amounts needed to adsorb sufficient
P, as previously explained. Therefore limestone and opoka should only be considered
as filter materials if they are to be used in
soils with extremely low soil pH after being
saturated with P. Slag materials, however, are
suitable to use in moderately acidic soils,
where their relative effectiveness can be
assumed to be at least 70% of commercial P
fertiliser during the first year, and possibly
more in subsequent years.
The observation that a soluble Ca-P phase,
probably ATCP, had accumulated in the filter
materials suggests that at least part of the P is
readily available to plants (Paper III).
With this evidence from pot and batch experiments, it seems likely that used filter
materials from wastewater treatment may
contain readily soluble Ca phosphates such as
ATCP and DCP.
27
Agnieszka Renman
TRITA LWR PhD Thesis 1043
Depth (cm)
Sand
0-5
5-10
10 20
20 30
30 40
40 50
0.0
0.2
0.4
0.6
0.8
-1
N (mg g )
Depth (cm)
Polonite
0-5
5-10
10 20
20 30
30 40
40 50
0.0
0.2
0.4
0.6
-1
N (mg g )
Fig. 17. Initial N content in the sand and Polonite (dark spotted area) and N enriched to
different layers (light spotted area) after 67 weeks of wastewater treatment by infiltration. Bars
indicate ±SE where columns were duplicated (Paper IV).
Bed filter lifetime for P removal
The Polonite used in the full-scale filter well
in a volume of 800 L (560 kg) was a sufficient
amount for treating approximately 70 m3 of
domestic wastewater, i.e. 11 L (8 kg) were
required per cubic metre (Paper IV). These
results can be compared with the expected
lifetime of the Filtralite P material as calculated by Heistad et al. (2006), where a total
filter volume of 6 m3 was expected to meet
the required effluent limit of 1.0 mg P L-1 for
5 years. Based on the data given by Heistad et
al. (2006), it can be calculated that 1 m3 of
wastewater could be treated by filtration
through approximately 5 L of Filtralite P.
This is half the amount required for treatment to approximately the same effluent limit
by Polonite. A comparison of expected
maximum P sorption capacity has revealed
that Polonite has a much higher capacity than
Filtralite P (Brogowski & Renman, 2004;
Ádám et al., 2006; Cucarella & Renman,
2008).
This is supported by results from plant uptake experiments showing that the retained P
in similar filter materials is easily accessible
(Hylander & Simán, 2001; Kvarnström et al.,
2004). However, filter materials cannot be
expected to replace industrial P fertilisers, at
least not in the short-term, because of their
low contents of P.
The alkaline properties and the content of
micro-elements, such as in the Polonite, are
still of interest for the recycling of materials
as soil amendments (Cucarella et al., 2008).
The Polonite material used in both on-site
wastewater treatment and the column experiment showed heavy metal concentrations
that are much lower than the European limits
for sludge disposal. Hence the low content of
heavy metals in the filter materials will not
interfere with their use as a fertiliser or soil
amendment. However it must be pointed out
that the wastewater used in the experiments
had very low dissolved metal concentrations.
28
On-site wastewater treatment-Polonite and other filter materials for removal of metals, nitrogen and phosphorus
been expected for the Nordkalk Filtra P as
well, but it clogged and disintegrated during
the experiment performed (Paper III). This
will limit its practical lifetime in field applications.
For Polonite it is suggested that 1-2 kg would
be required as a design volume for treatment
of one cubic metre of wastewater. However,
other parameters such as nitrogen and BOD
have to be considered too and the Polonite
material and filter well design, including pretreatment steps, have to be further improved.
Drizo et al. (2002) argued that allowing the
system to rest for a certain time could enhance the P removal of slag-based systems.
Short resting periods were applied in the
filter beds studied in the present thesis, although clear advantages for the P removal
were not observed. In the case of strong
alkaline materials, resting periods can also
increase the interference with atmospheric
CO2, leading to the formation of CaCO3 (cf.
Drizo et al., 2002). The RBF technology
studied in this thesis has already been applied
for treatment of wastewater in summer cottages and part-time residences. This means
that longer resting periods will occur and the
question is how these periods will affect the
longevity of a filter material such as Polonite.
Further studies should be performed on the
effect of operating and resting periods, and
the integration of drying and wetting periods
for optimal removal efficiency.
Hence Polonite could be expected to have
shown better removal capacity in the fullscale treatment system tested. There are
several possible explanations for the observed lowered capacity.
The properties of the material can vary depending on the manufacturing process. The
Polonite used in the experiment had a starting pH of 11.9, compared with the normal
pH of 12.5 for this product.
Polonite is normally produced in the particle
size range 2-5.6 mm but it was observed that
the material used in the filter well contained
coarser fractions. Finer fractions of Polonite
give higher P sorption capacity (unpubl.
data). The loading of organic matter to the
filter was high during the first 18 weeks of
operation because of inadequate pretreatment in the septic tank. This improved
when the wastewater was discharged to the
filter through a biofilter but the effect of
dissolved organic matter could have persisted, which is suggested to reduce the P
sorption capacity of many alkaline filter materials.
Plotting the removal capacity against pH for
the filter well experiment revealed a good
correlation and a strong relationship between
PO4-P removal and pH for Polonite. Other
studies of alkaline filter materials have shown
the same relationship, although in batch
experiments (Ádám et al., 2007).
The pH could be an easy parameter to use
for decisions on replacing the filter material
in the filter well. Hence a certain pH could
correspond to a minimum effluent concentration or removal capacity according to local
discharge limits. In Norway (Heistad et al.,
2006) and Sweden (www.avloppsguiden.se),
the effluent criterion is normally a total P
concentration of 1 mg L-1. This value corresponds to a total P reduction of 90%. The
data presented here for the full-scale treatment system show that effluent pH cannot
indicate when a reactive bed filter has to be
exchanged. The filter material Polonite had
an extraordinarily high P removal capacity
during two-thirds of the operation time,
while a gradual breakthrough developed
during the last one-third. This could have
C ONCLUSIONS
The research in this thesis, with a combination of batch, long-term column and experiments with a full-scale treatment system,
demonstrated the potential of using reactive
filter materials for the treatment of wastewater. The removal of metals and nutrients by
different materials was shown to be influenced by many factors, such as pH, chemical
composition of material, element and organic
matter competition in wastewater. Experiments with wastewater are more difficult to
interpret than those performed with aqueous
solutions. The differences in the results obtained by investigators of reactive filter materials arise from the fact that wastewater,
particularly landfill leachate, has a complex
29
Agnieszka Renman
TRITA LWR PhD Thesis 1043
zeolites for NH4-N removal could be
included.
composition, which can interfere with the
adsorption of metals. Hence, this suggests
that it is not possible to draw conclusions of
a general or fundamental nature on the removal mechanisms involved.
The conclusions of this thesis are as follows:
x Batch experiments and ATR-FTIR
investigations indicate that amorphous tricalcium phosphate (ATCP)
forms in the materials. This means
that at least part of the accumulated
PO4-P is readily available to plants.
However, since only up to 18% of
the accumulated PO4-P was readily
dissolved in the experiments, the possibility cannot be excluded that part
of the phosphorus had crystallised to
slightly less soluble phases.
x Polonite and Nordkalk Filtra P are
promising filter materials for removal
of PO4-P from household wastewater, maintaining a PO4-P removal efficiency of >95%. Water-cooled blast
furnace slag and natural wollastonite
were also able to remove PO4-P, although less efficiently.
x Polonite used in wastewater filters
can be recycled as a combined fertiliser for plant production and liming
agent on acidic soils. While P retention capacity of blast furnace slag is
lower than that of Polonite, it nevertheless shows greater promise in P
fertiliser effectiveness. The low content of heavy metals in the filter materials will not interfere with their use
as a fertiliser or soil amendment.
x Investigations of the role of the feed
solution applied to the filter material
showed that synthetic wastewater
loading in small columns led to better
long-term P removal capacity than
real wastewater in a full-scale, bed filter trial. The Polonite in the column
systems examined was still ‘active’ at
the end of the experimental period,
showing no clear breakthrough tendency, while this clearly occurred in
the full-scale bed filter system.
x Two long-term column experiments
showed that dissolved metals in
wastewater (domestic and landfill
leachate) were removed in various
amounts by the filter materials studied. The kind of blast furnace slag
(BFS) used in our investigation did
not show good removal efficiency
over the whole contaminant spectrum. The reactive medium Polonite
did not reach the saturation point of
many of the major contaminants. A
drawback of the studied reactive media was the release of certain elements, such as Cr, from the BFS and
Polonite filter matrix. According to
speciation calculations with Visual
MINTEQ, a large fraction of dissolved Cu, Cr(III), Zn and Ni in the
column leachates were bound to
DOM.
x Cyclic loading of wastewater to the
filters interspersed with periods of
saturation/drying did not improve P
removal capacity. However, more
studies are needed in both column
and full-scale experiments.
x None of the six filter materials tested
here showed good potential for removing total inorganic nitrogen
(TIN). Ammonium-N, a dominant N
component in the wastewater, was
transformed in the filter beds to
NO3-N. Polonite removed 17.7% of
the influent nitrogen, which was
mainly associated with losses through
volatilisation. A similar removal capacity was found in the experiment
using landfill leachate. Although efficient for P removal, alkaline filter materials cannot be considered suitable
bed components for N removal in
engineered wetlands and compact filter wells. Instead, dual treatment systems should be developed where e.g.
x The pH proved not to be a simple
parameter for determining when the
filter material should be replaced. Instead, the P removal longevity of re30
On-site wastewater treatment-Polonite and other filter materials for removal of metals, nitrogen and phosphorus
active filter materials can be estimated
from the volume of wastewater
treated if a water meter is installed in
the house.
x A filter facility designed as a multilayer unit of several reactive media
could be a possible solution for more
effective, broad-spectrum removal of
pollutants. Prior to full-scale application, matrix selection, filter design
and operational procedures must be
developed to ensure appropriate hydraulic conditions and to facilitate periodic media replacement and possible recovery.
31
Agnieszka Renman
TRITA LWR PhD Thesis 1043
32
On-site wastewater treatment-Polonite and other filter materials for removal of metals, nitrogen and phosphorus
R EFERENCES
Abollino, O., Giacomino, A., Malandrino, M., Mentasi, E., 2008. Interaction of metal ions with
montmorillonite and vermiculite. Appl. Clay Sci. 38: 227-236.
Ádám, K., Søvik, A.K., Krogstad, T., 2006. Sorption of phosphorous to Filtralite-PTM – The
effect of different scales. Water Res. 40(6): 1143-1154.
Ádám, K., Krogstad, T., Vråle, L., Søvik, A.K., Jenssen, P.D., 2007. Phosphorus retention in the
filter materials shellsand and Filtralite P® - Batch and column experiment with synthetic P
solution and secondary wastewater. Ecol. Eng. 29(2): 200-208.
Ahsan, S., Kaneco, S., Ohta, K., Mizuno, T., Kani, K., 2001. Use of some natural and waste materials for waste water treatment. Water Res. 35(15): 3738-3742.
Arias, C.A., Brix, H., 2005. Phosphorus removal in constructed wetlands: can suitable alternative
media be identified? Wat. Sci. Technol. 51(9): 267-273.
Artola, A., Balaguer, M.D., Rigola, M., 1997. Heavy metal binding to anaerobic sludge. Water
Res. 31: 997-1004.
Atalay, S., Adiguzel, H.I., Atalay, F., 2001. Infrared absorption study of Fe2O3-CaO-SiO2 glass
ceramics. Mater. Sci. Eng. A304-306: 796-799.
Babel, S., Kurniawan, T.A., 2003. Low-cost adsorbents for heavy metals uptake from contaminated water: a review. J. Hazard. Mater. B97: 219-243.
Baker, M.J., Blowes, D.W., Ptacek, C.J., 1998. Laboratory development of permeable reactive
mixtures for the removal of phosphorus from onsite wastewater disposal systems. Environ. Sci. Technol. 32, 2308-2316.
Bailey, S.E., Olin, T.J., Bricka, R.M., Adrian, D.D., 1999. A review of potentially low-cost sorbents for heavy metals. Water. Res. 33(11): 2469-2479.
Brix, H., Arias, C.A., Bubba, M. del, 2001. Media selection for sustainable phosphorus removal in
subsurface flow constructed wetlands. Wat. Sci. Technol., 44(11-12): 47-54.
Brogowski, Z., Renman, G., 2004. Characterization of opoka as a basis for wastewater treatment.
Polish J. Environ. Stud. 13(1): 15-20.
Brooks, A.S., Rozenwald, M.N., Geohring, L.D., Lion, L.W., Steenhuis, T.S., 2000. Phosphorus
removal by wollastonite: A constructed wetland substrate. Ecol. Eng. 15: 121-132.
Brown, P.A., Gill, S.A., Allen, S.J., 2000. Metal removal from wastewater using peat. Water Res.
34 (16): 3907-3916.
Chen, J.P., Chua, M.L., Zhang, B., 2002. Effects of competitive ions, humic acid, and pH on
removal of ammonium and phosphorous from the synthetic industrial effluent by ion
exchange resins. Waste Manage. 22: 711-719.
Christoffersen, M.R., Christoffersen, J., Kibalczyc, W., 1990. Apparent solubilities of two amorphous calcium phosphates and of octacalcium phosphate in the temperature range 3042oC. J. Crystal Growth 106: 349-354.
Crites, R., Tchobanoglous, G., 1998. Small and decentralized wastewater management systems.
WCB/McGraw-Hill, 1084 pp.
Cucarella, V., Renman, G., 2008. Phosphorus sorption capacity of filter materials used for on-site
wastewater treatment determined in batch experiments – a comparative study. Submitted
to J. Environ. Quality.
Cucarella, V., Zaleski, T., Mazurek, R., Renman, G., 2008. Effect of reactive substrates used for
the removal of phosphorus from wastewater on the fertility of acid soils. Bioresour.
Technol. 99(10): 4308-4314.
33
Agnieszka Renman
TRITA LWR PhD Thesis 1043
Cucarella, V., Zaleski, T., Mazurek, R., Renman, G., 2007. Fertilizer potential of calcium-rich
substrates used for the removal of phosphorus from wastewater. Polish J. Environ. Stud.
16(6): 817-822.
de-Bashan, L.E., Bashan, Y., 2004. Recent advances in removing phosphorus from wastewater
and its future use as fertilizer (1997-2003). Water Res., 4222-4246.
Dimitrova, S.V., Mehandgiev, D.R., 2000. Interactions of blast-furnace slag with heavy metal ions
in water solutions. Water Res. 34: 1957-1961.
Dobbie, K.E., Heal, K.V., Smith, K.A., 2005. Assessing the performance of phosphorussaturated ochre as a fertilizer and its environmental acceptability. Soil Use and Manag. 21:
231-239.
Drizo, A., Comeau, Y., Forget, C., Chapuis, R.P., 2002. Phosphorus saturation potential: A parameter for estimating the longevity of constructed wetland systems. Environ. Sci. Technol. 36: 4642-4648.
Drizo, A., Forget, C., Chapuis, R.P., Comeau, Y., 2006. Phosphorus removal by electric arc furnace steel slag and serpentinite. Water Res. 40(8): 1547-1554.
Düker, A., Ledin, A., Karlsson, S., Allard, B., 1995. Adsorption of zinc on colloidal (hydr)oxides
of Si, Al and Fe in the presence of a fulvic acid. Appl. Geochem. 10(2): 197-205.
Eveborn, D., 2003. Småskalig rening av avloppsvatten med Polonite-filter. TRITA-LWR Master
Thesis 03-26. KTH, Department of Land and Water Resources Engineering, Stockholm.
Ganrot, Z., Dave, G., Nilsson, E., 2007. Recovery of N and P from human urine by freezing,
struvite precipitation and adsorption to zeolite and active carbon. Bioresour. Technol. 98:
3112-3121.
Goyal, N., Jain, S.C., Banerjee, U.C., 2003. Comparative studies on the microbial adsorption of
heavy metals. Adv. Environ. Res. 7: 311-319.
Gustafsson, J.P., 2001. Modeling the acid/base properties and metal complexation of humic
substances with the Stockholm Humic Model. J. Colloid Interface Sci. 244: 102-112.
Gustafsson,
J.P.,
2006.
Visual
MINTEQ
version
2.51.
http://www.lwr.kth.se/English/OurSoftWare/Vminteq/index.htm. Last accessed 2007-02-27.
Gustafsson,
J.P.,
2007.
Visual
MINTEQ,
version
2.53.
Web:
http://www.lwr.kth.se/English/OurSoftware/vminteq/index.htm.
Gustafsson, J.P., Renman, A., Renman, G., Poll, K., 2008. Phosphate removal by mineral-based
sorbents used in filters for small-scale wastewater treatment. Water Res. 42: 189-197.
Hallberg, M., Renman, G., 2007. Reactive filters for removal of dissolved metals in highway runoff. In: Highway and Urban Environment, Proceed. of the 8th Highway and Urban Environment Symposium, Springer (2007), 465-474.
Hallberg, M., 2007. Treatment conditions for the removal of contaminants from road runoff.
TRITA-LWR PHD 1032. Doctoral Thesis in Land and Water Resources Sciences. KTH
Architecture and the Built Environment, Stockholm.
Heavey, M., 2003. Low-cost treatment of landfill leachate using peat. Waste Manage. 23(5): 447454.
Hedström, A., 2006. Reactive filter materials for ammonium and phosphorus sorption in small
scale wastewater treatment. Doctoral Thesis, 2006:17, Luleå University of Technology.
Heistad, A., Paruch, A.M., Vråle, L., Ádám, K., Jenssen, P.D., 2006. A high-performance
compact filter system treating domestic wastewater. Ecol. Eng. 28: 374-379.
Hellström, D., Jonsson, L., 2006. Evaluation of small on-site wastewater treatment systems. Manage. Environ. Qual. 17(6): 728-739.
34
On-site wastewater treatment-Polonite and other filter materials for removal of metals, nitrogen and phosphorus
Hill, C.M., Duxbury, J., Geohring, L., Peck, T., 2000. Designing constructed wetlands to remove
phosphorus from barnyard runoff: A comparision of four alternative substrates. J. Environ. Sci. Health A35(8): 1357-1375.
Huijgen, W.J.J., Witkamp, G.J., Comans, R.N.J., 2006. Mechanisms of aqueous wollastonite carbonation as a possible CO2 sequestration process. Chem. Eng. Sci. 61: 4242-4251.
Hylander, L.D., Kietliľska, A., Renman, G., Simán, G., 2006. Phosphorus retention in filter materials for wastewater treatment and its subsequent suitability for plant production. Bioresour. Technol. 97: 914-921.
Hylander, L. D., Simán, G., 2001. Plant availability of phosphorus sorbed to potential wastewater
treatment materials. Biol. Fertil. Soils 34: 42-48.
Hylander, L. D., Makino, T. & Ae, N. 1999. Bray-2 phosphorus as influenced by soil fineness and
filtration time. Commun. Soil Sci. Plant Anal. 30:947-955.
Hylander, L.D., Ae, N., 1999. Nutrient dynamic around roots of brachiaria, maize, sorghum and
upland rice in an Andisol. Soil Sci. Plant Nutr. 45(3): 617-626.
Jenssen, P.D., 1996. Ecological engineering for wastewater treatment: fundamentals and examples. In: Recycling the Resource – Ecological Engineering for Wastewater Treatment
(Eds. J. Staudenmann, A. Schönborn, C. Etnier), Environmental Research Forum Vol. 56,. Transtec Publications, pp15-23.
Johansson, L., 1998. Phosphorus Sorption to Filter Substrates – Potential Benefits for On-site
Wastewater Treatment. Dissertation, TRITA-AMI PHD 1024, Royal Institute of Technology, Stockholm.
Johansson, L., Gustafsson, J.P., 2000. Phosphate removal using blast furnace slag and opoka –
mechanisms. Wat. Res. 34(1): 259-265.
Johansson Westholm, L., 2006. Substrates for phosphorus removal – Potential benefits for onsite wastewater treatment? Water Res. 40: 23-36.
Jönsson, J., Sjöberg, S., Lövgren, L., 2006. Adsorption of Cu(II) to schwertmannite and goethite
in presence of dissolved organic matter. Water Res., 40: 969-974.
Kamara, P., Jacob, S., Srinivasan, D., 1989. Removal of heavy metals from waste water by sulphide precipitation technique. Bull. Electrochem. 5: 572-574.
Katsoyiannis, A., Samara, C., 2007. The fate of dissolved organic carbon (DOC) in the wastewater treatment process and its importance in the removal of wastewater contaminants.
Env. Sci. Pollut. Res. 14(5): 284-292.
Kiely, G., 1997. Environmental Engineering Technologies. McGraw-Hill, 979 pp.
Kietliľska, A., Renman, G., Jannes, S. and Tham, G., 2005. Nitrogen removal from landfill
leachate using a compact constructed wetland and the effect of chemical pre-treatment. J.
of Environ. Sci. and Health, 40: 1493-1506.
Kieliľska, A., Renman, G., 2005. An evaluation of reactive filter media for treating landfill
leachate. Chemosphere 61: 933-940.
Kietliľska, A., 2004. Engineered wetlands and reactive bed filters for treatment of landfill
leachate. Licentiate Thesis,TRITA-LWR.LIC 2017, Royal Institute of Technology, Stockholm.
Kløve, B., Mæhlum, T., (Eds), 2000. On-site wastewater treatment and re-use in constructed
wetlands and filter media (special issue on ecological engineering). J. Environ. Sci. Health,
Part A, A35(8): 1037-1502 .
Kostura, B., Kulveitová, H., Lešco, J., 2005. Blast furnace slags as sorbents of phosphate from
water solutions. Water Res., 39: 1795-1802.
35
Agnieszka Renman
TRITA LWR PhD Thesis 1043
Kvarnström, M.E., Morel, C.A.L., Krogstad, T., 2004. Plant availability of phosphorus in filter
substrates derived from small-scale water treatment systems. Ecol. Eng. 22(1): 1-15.
LECO,1995. Instruction Manual CHN-900 carbon, hydrogen and nitrogen determinator/CHNS932 carbon, hydrogen, nitrogen, and sulfur determinator. LECOR Corporation, St.
Joseph, USA.
Mæhlum, T., 1998. Cold-climate constructed wetlands: Aerobic pre-treatment and
horizontal
subsurface flow systems for domestic sewage and landfill leachate purification. Doctoral
Scientiarum Theses 1998:9, Agricultural University of Norway, Ås.
Malandrino, M., Abollino, O., Giacomino, A., Aceto, M., Mentasti, E., 2006. Adsorption of heavy
metals on vermiculite: Influence of pH and organic ligands. J. Coll. Inter. Sci. 299: 537546.
McCay, G., (Ed), 1996. Use of adsorbents for the removal of pollutants from wastewater. CRC
Press, Inc, Boca Raton, Florida,186 pp.
Meyer, S., 2004. Hydraulic performance of reactive bed filters for wastewater treatment under
different flow conditions. Master´s Degree Project, LWR-EX-04-27, Royal Institute of
Technology, Stockholm.
Metcalf & Eddy, 2003. Wastewater Engineering – Treatment and reuse. 4th ed., McGraw-Hill.
1848 pp.
Moriyama, K., Mori, T., Arayashiki, H., Saito, H., Chino, M., 1989. Amount of heavy metals
derived from domestic wastewater. Wat. Sci. Tech. 21: 1913-1916.
Mitsch, W.J., Jørgensen, S.E., 1989. Introduction to ecological engineering. In: Ecological Engineering: An introduction to ecotechnology (Eds. W.J. Mitsch and S.E. Jørgensen),. Wiley,
New York, pp 3-19.
Morse, G.K., Brett, S.W., Guy, J.A., Lester, J.N., 1998. Review: Phosphorus removal and recovery technologies. Sci. Total Environ., 212: 69-81.
Nyholm, A.M., Yli-Halla, M., Kivistö, P., 2005. Wastewater treatment in filter beds: reuse of filter
material. MTT Agrifood Research Finland, Jokioinen, Finland, 40 pp.
Oguz, E., 2004. Removal of phosphate from aqueous solution with blast furnace slag. J. Hazard.
Mater., B114: 131-137.
O’Hannesin, S.E., Gilham, R.W., 1998. Long-term performance of an in-situ ”iron-wall” for
remediation of VOCs. Ground Water 36 (1): 164-170.
Omoike, A.L., Vanloon, G.W., 1999. Removal of phosphorus and organic matter removal by
alum during wastewater treatment. Water Res., 33(17): 3617-3627.
Papadopoulus, P., Rowell, D.L., 1989. The reactions of copper and zinc with calcium carbonate
surfaces. European J. Soil Sci. 40: 39-48.
Pell, M., Nyberg, F., Ljunggren, H., 1990. Microbial numbers and activity during infiltration of
septic-tank effluent in a subsurface sand filter. Water Res. 24(11): 1347-1354.
Petrovic, M., Kastelan-Macan, M., Horvat, A.J.M., 1999. Interactive sorption of metal ions and
humic acids onto mineral particles. Water Air Soil Pollut. 111: 41–56.
Renman, G., Kietliľska, A., & Cucarella Cabañas, V., 2004. Treatment of phosphorus and bacteria by filter media in onsite wastewater disposal systems. In: Ecosan closing the loop.
Proceedings of the 2nd international symposium, 7th-11th April 2003, Lübeck, Germany,
573-576.
Renou, S., Givaudan, J.G., Poulain, S., Dirassouyan, F., Moulin, P., 2008. Landfill leachate treatment: Review and opportunity. J. Hazard. Mater. 150: 468-493.
Ringqvist, L., Holmgren, A., Öborn, I., 2002. Poorly humified peat as an adsorbent for metals in
wastewater. Water Res. 36: 2394-2404.
36
On-site wastewater treatment-Polonite and other filter materials for removal of metals, nitrogen and phosphorus
Ritter, W.F., Eastburn, R.P., 1988. A review of denitrification in on-site wastewater treatment
systems. Environ. Pollut. 51: 49-61.
Ronkanen, A.K., Kløve, B., 2007. Use of stabile isotopes and tracers to detect preferential flow
patterns in a peatland treating municipal wastewater. J. Hydrol. 347: 418-429.
Sakadevan, K., Bavor, H. J., 1998. Phosphate Adsorption Characteristics of Soils, Slags and Zeolite to Be Used As Substrates in Constructed Wetland Systems. Water Res. 32(2): 393399.
SEPA, 2002. Action plan for recovery of phosphorus from wastewater. Swedish Environmental
Protection Agency, Report 5214 (In Swedish).
Sharma, Y.C., Gupta, G.S., Prasad, G., Rupainwar, D.C., 1990. Use of wollastonite th thr removal
of Ni (II) from aqueous solutions. Water Air Soil Pollut. 49: 69-79.
Sharma, Y.C., 1995. Economic treatment of cadmium(II)-rich hazardous waste by indigenous
material. J. Colloid Interface Sci. 173: 66-70.
Shilton, A.N., Elmetri,, I., Drizo, A., Pratt, S., Haverkamp, R.G., Bilby, S.C., 2006. Phosphorus
removal by an ’active’ slag filter - a decade of full scale experience. Wat Res. 40(1): 113118.
Smith, R.M., Martell, A.E., Motekaitis, R.J., 2003. NIST Critically Selected Stability Constants of
Metal Complexes Database. Version 7.0. NIST Standard Reference Database 46. National Institute of Standards and Technology, US Department of Commerce, Gaithersburg.
Søvik, A.K., Kløve, B., 2005. Phosphorus retention processes in shell sand filter system treating
municipal wastewater. Ecol. Eng. 25: 168-182
Trivedi, P. Axe, L., 2001. Predicting divalent metal sorption to hydrous Al, Fe, and Mn oxides.
Environ. Sci. Technol. 35: 1779-1784.
Valsami-Jones, E., 2001. Mineralogical controls on phosphate recovery from wastewaters. Miner.
Mag. 65(5): 611-620.
Van der Houwen, J.A.M., Valsami-Jones, E., 2001. The application of calcium phosphate precipitation chemistry to phosphorus recovery: the influence of organic ligands. Environ.
Technol. 22: 1325-1335.
Vinnerås, B., Palmquist, H., Balmér, P., Jönsson, H., 2006. The characteristics of household
wastewater and biodegradable solid waste – A proposal for new Swedish design values.
Urban Wat. J., 3(1): 3-11.
Vohla, C., Põldvere, E., Noorvee, A., Kuusemets, V., Mander, Ü., 2005. Alternative filter media
for phosphorous removal in a horizontal subsurface flow constructed wetland. J. Environ. Sci. and Health 40: 1251-1264.
Vohla, C., Alas, R., Nurk, K., Baatz, S., Mander, Ü., 2007. Dynamics of phosphorus, nitrogen and
carbon removal in a horizontal subsurface flow constructed wetland. Sci. Total Environ.,
380: 66-74.
Wang, J., Huang, C.P., Allen, H.E., 2006. Predicting metals partitioning in wastewater treatment
plant influents. Water Res. 40: 1333-1340.
WHO & UNICEF. 2006. Meeting the MDG drinking water and sanitation target: The Urban
and Rural challenge of the decade. Geneva, Switzerland: WHO Press.
Yadava, K.P., Tyagi, B.S., Singh, V.N., 1991. Effect of temperature on the removal of lead (II) by
adsorption on China clay and wollastonite. J. Chem. Tech. Biotechnol. 51:47-60.
Yao, K.M., Habbibian, M.T., O´Melia, C.R., 1971. Water and wastewater filtration: concepts and
applications. Env. Sci. Tech. 5: 2031-2038.
37
Agnieszka Renman
TRITA LWR PhD Thesis 1043
Zhu, T., 1998. Phosphorus and nitrogen removal in light-weight aggregate (LWA) constructed
wetlands and intermittent filter systems. Doctor Scientiarum Theses 1997:16, Agricultural
University of Norway.
38
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement