L IC E N T IAT E T H E S I S

On-Site Sanitation Systems

An Integrated Assessment of Treatment
Efficiency and Sustainability

Brenda Vidal

Urban Water Engineering

 On-Site Sanitation Systems

An Integrated Assessment of Treatment Efficiency

and Sustainability

Brenda Vidal
Luleå, 2018

Division of Architecture and Water

Department of Civil, Environmental and Natural Resources Engineering
Luleå University of Technology
Printed by Luleå University of Technology, Graphic Production 2018

ISSN 1402-1757
ISBN 978-91-7790-256-0 (print)
ISBN 978-91-7790-257-7 (pdf)
Luleå 2018
www.ltu.se
Preface
This thesis work was carried out in the Urban Water Research Group at the department of Civil,
Environmental and Natural Resources Engineering of Luleå University of Technology. The studies were
financially supported by the Swedish Research Council Formas (project number 942-2015-758), the J
Gustaf Richert Foundation and the Swedish Agency for Marine and Water Management (grant 1:12,
Measures for Sea and Water Environment).
I would like to thank the staff of the participating municipalities of Robertsfors, Skellefteå, Umeå and
Vännäs for their support with the identification and inspection of facilities during the field study. I am
grateful to the staff at the Development Centre for Water (Utvecklingscentrum för vatten) who helped
with logistics and provided the facilities to process our field samples. Special thanks to Jonathan Alm for
the fruitful discussions and for sharing his practical knowledge about on-site sanitation systems. Kerstin
Nordqvist, Snezana Gavric and Katharina Lange are gratefully acknowledged for the help with the
laboratory and field work. I am also very thankful to the members of the reference group for accepting
to participate in the multi-criteria study and for their valuable contributions.
I would like to express my deepest gratitude to my dear supervisors Annelie Hedström and Inga
Herrmann, for their never-ending support, encouragement and fruitful discussions. Your optimistic
attitude and professionalism guided me all the way through my Licentiate. I have learned a lot about the
scientific and human side of Academia from you, and for that I am very grateful. Thank you, Annelie,
for empowering me during these years. Thank you, Inga, for your easiness and for always receiving me
in your office with a big smile. To Maria Viklander, thank you for your kindness, guidance and
metaphors, and from creating such a welcoming and fun working environment in our group.
Since science is a social endeavour, I am very thankful to my colleagues at the Urban Water Group for
the interesting discussions and feedback, and for bringing our research to another level. Thank you, Jiri
Marsalek, for your valuable comments and constructive discussions and, thank you, Stefan Marklund, for
sharing your knowledge about wastewater systems with me (small systems, big problems!). Special thanks
to all my fellow PhD colleagues for making this journey so fun, for sharing your skills and for always
being there (in the stand-by list at least) when needed. Remember that we are the future of the water
sector! Special thanks to Hendrik, Youen and Joel, for all the weekend adventures, including fishing,
hiking, deep snow and cabins in the woods.
Finally, I would like to thank my dear EnvEuro community for the friendship, including fun scientific
discussions and gatherings by lakes and high mountains. It is very comforting to know that you are always
there despite the distance. To Franciska and Jan, for your lovely support and encouragement with my
career, and for all the cosy weekends in Stockholm. To my dear family in the Canary Islands, for the love
and the warmth you send all the way up to Luleå and for teaching me that home is wherever my heart is
at peace. To Joel, for your calm presence and unconditional support, for your easiness and willingness to
follow along when I got the PhD position in Luleå (the fishing is good up there, you said), and there we
went without hesitation.
Thank you to all of you who work with sustainability, and to those of you who try to make this planet a
better place to live.

Brenda Vidal
Luleå, November 2018

i
ii
Abstract
Small on-site sanitation systems for wastewater collection and treatment are prevalent in
suburban and rural areas in many countries. However, these systems often underperform, causing
potential impact to the receiving waters and increasing the risks to public health, thus hindering
the overall sustainability of the systems. Understanding the different sustainability dimensions
and trade-offs between assessment indicators can support the planning of sustainable on-site
sanitation systems for a specific context. The overall aim of this thesis was to evaluate the
sustainability and function of on-site sanitation systems by defining a set of indicators to assess
on-site sanitation options and estimating them for different scenarios, and by investigating the
treatment efficiency of on-site facilities for domestic wastewater treatment in a field study.
Particular attention was given to the removal of phosphorus (P) and indicator bacteria due to
their relevance in terms of eutrophication risk and public-health concern.
In a multi-criteria approach, twelve indicators were defined to assess nine on-site sanitation
systems. A reference group representing stakeholders’ views assigned weights to express the
relative importance of each indicator. The reference group assigned the highest weights to the
indicators robustness, risk of pathogen discharge and nutrient removal. Assessing the robustness proved
to be challenging, as there is a gap between how the sanitation systems are expected to perform
based on their design, and how they actually perform in practice, mainly due to incorrect
construction, operation and maintenance. The discriminating power of the indicators was
calculated using the entropy method, which showed that the indicators energy recovery and capital
cost had little impact in the final ranking of the alternatives. A sustainability ranking was obtained
by using the method ELECTRE III. A scenario analysis based on different settings of interest
based on socio-economic and geographical factors was done to evaluate the changes in the
ranking of alternatives. Overall, the greywater–blackwater separation system ranked the highest
in the baseline scenario and when nutrient-related indicators were important (Scenario 2),
together with the urine diversion system. The sand filter and drain field were the most sustainable
options when nutrient removal and recycling was not important (Scenario 1), and (in
combination with chemical P-removal) when the indicators related to energy and climate change
had the highest weights (Scenario 3). In terms of P-removal, chemical removal outranked the
alkaline P-filter.
In a field study, the effluent wastewater from twelve on-site wastewater treatment facilities with
sand and alkaline P-filters was evaluated in terms of removal and discharge of organic content,
total and dissolved phosphorus, and indicator bacteria (E. coli, total coliforms, intestinal
enterococci and C. perfringens). The results showed that the investigated sand filters generally had
low P-removal capacity and sometimes exceeded the criteria for excellent water quality set by
the EU bathing water directive with regard to intestinal enterococci and E. coli. Only one sand
filter of eight was confirmed to remove P according to the Swedish guidelines for areas with
non-sensitive receiving waters with a tot-P effluent concentration below 3 mg L−1. This indicates
that a downstream treatment step is needed to meet the guidelines regarding P discharge
concentrations. Alkaline P-filters generally removed P efficiently. Despite high effluent pH, the
collected data did not generally confirm a further reduction of the bacterial content of the
wastewater in the P-filters, as had been previously hypothesized. However, effluent
concentrations of indicator bacteria showed moderate positive correlations with effluent
concentrations of P and organic matter, indicating the potential of the P-filters to serve as a
polishing step also for bacteria.

iii
iv
Sammanfattning
Små lokala avloppssystem för insamling och rening av avloppsvatten är vanligt förekommande i
omvandlings- och landsbygdsområden i många länder. Dessa system fungerar ofta sämre än
förväntat vilket kan leda till effekter på recipienter, ökad risk för folkhälsan och begränsa
systemens hållbarhet. En djupare förståelse för hållbarhetsdimensioner och avvägningar mellan
olika indikatorer kan stödja planering och implementering av hållbara lokala avloppsystem för
specifika sammanhang. Det övergripande syftet med denna uppsats var att utvärdera hållbarhet
och funktion av enskilda avloppssystem. Detta utfördes genom att först definiera en uppsättning
indikatorer för bedömning av ett antal enskilda avloppsalternativ och sedan utvärdera dem för
olika scenarier. Vidare var syftet att undersöka några systems prestanda för rening av
hushållsavloppsvatten i en fältstudie. Särskilt fokus lades på reduktion av fosfor (P) och
indikatorbakterier på grund av deras relevans i relation till övergödningsrisk och
folkhälsoperspektiv.
I en multikriteriestudie definierades tolv indikatorer för att bedöma nio typer av enskilda
avloppssystem. En referensgrupp som representerade olika intressenter viktade indikatorerna för
att uttrycka den relativa betydelsen av varje indikator. Referensgruppen gav systemens robusthet,
risk för utsläpp av patogener, och näringsreduktion störst vikt. Att bedöma robusthet var en utmaning
i studien, eftersom det finns ett gap mellan hur systemen förväntas fungera, och hur de faktiskt
fungerar i praktiken, mestadels på grund av felaktig konstruktion, drift och underhåll.
Indikatorernas diskriminerande effekt räknades ut med entropimetoden, som visade att
indikatorerna energiåtervinning och kapitalkostnad hade liten inverkan på alternativens slutgiltiga
rangordning. En scenarioanalys genomfördes baserad på socioekonomiska och geografiska
faktorer. En hållbarhetsrankning erhölls genom att använda ELECTRE III-metoden. Överlag
rankades system som separerar gråvatten – svartvatten samt urinseparerande system högst i
basfallsscenariot och när näringsrelaterade indikatorer var viktiga (scenario 2). Markbäddar och
infiltrationsanläggningar var de mest hållbara alternativen när rening och återvinning av
näringsämnen inte var viktigt (scenario 1) och (i kombination med kemisk P-rening) när
indikatorer relaterade till energi och klimatförändringar viktades högst (scenario 3). När det gäller
P-rening, gav system med kemisk rening ett bättre utfall än de med alkaliska P-filter.
I en fältstudie utvärderades tolv enskilda avloppsanläggningar med markbaserade system och
alkaliska P-filter med avseende på rening och utsläpp av organiskt innehåll, totalt och löst P och
indikatorbakterier (E.coli, totala koliformer, intestinala enterokocker och C. perfringens).
Resultaten visade att de markbaserade systemen generellt hade låg P-reningskapacitet och ofta
överskreds kriterierna för utmärkt vattenkvalitet enligt EUs badvattendirektiv avseende
intestinala enterokocker och C. perfringens. Endast ett markbaserat system av åtta uppvisade en
P-rening som låg under de svenska riktlinjerna för områden med normal skyddsnivå, med en
utsläppskoncentration på under 3 mg L-1 tot-P. Detta indikerade att markbäddar behöver
nedströms placerade reningssteg för att uppfylla nuvarande riktlinjer för utsläpp av P. Alkaliska
P-filter avskilde generellt P effektivt. Trots högt pH-värde kunde ingen ytterligare minskning av
bakterieinnehåll i avloppsvattnet från P-filtren påvisas statistiskt. Utsläppen av indikatorbakterier
uppvisade emellertid måttliga positiva korrelationer med utsläpp av P och organiskt material,
vilket visar på en viss potential att P-filtren kan fungera som ett ytterligare poleringssteg även för
bakterier.

v
vi
Content
Preface ................................................................................................................................................ i
Abstract ............................................................................................................................................. iii
Sammanfattning ..................................................................................................................................v
List of papers ......................................................................................................................................ix
1. Introduction..............................................................................................................................1
1.1. Aim and research objectives .................................................................................................2
1.2. Thesis structure ....................................................................................................................2
2. Background ..............................................................................................................................5
2.1. Small-scale on-site sanitation systems ....................................................................................5
2.2. Removal of phosphorus and microorganisms in on-site sanitation systems .............................8
2.3. Sustainability assessment in on-site sanitation ...................................................................... 11
3. Methods ................................................................................................................................... 13
3.1. Multi-criteria analysis ......................................................................................................... 13
3.1.1. Sustainability indicators and alternatives for on-site sanitation ...................................... 13
3.1.2. The ELECTRE III method ........................................................................................ 15
3.1.3. Reference group and assignation of weights ................................................................ 17
3.1.4. Entropy study – discriminating power of the indicators ............................................... 18
3.1.5. Scenarios analysis ........................................................................................................ 19
3.1.6. On-site sanitation alternatives ..................................................................................... 19
3.2. Field investigation of existing sand filters and alkaline P-filters............................................. 22
3.2.1. Selection of facilities ................................................................................................... 22
3.2.2. Flow-proportional sampling........................................................................................ 24
3.2.3. Analyses of water samples ........................................................................................... 25
3.2.4. Data analyses .............................................................................................................. 25
4. Results ..................................................................................................................................... 27
4.1. Evaluating the sustainability of on-site sanitation systems..................................................... 27
4.1.1. Evaluation of on-site sanitation systems in terms of robustness, risk of pathogen discharge
and P removal ............................................................................................................ 27
4.1.2. Weighting of the indicators by the reference group and the entropy method ............... 30
4.1.3. Scenario analysis ......................................................................................................... 31
4.2. Removal of organic content, P and bacteria from sand filters and P-filters ........................... 32
4.2.1 Performance of sand filters .......................................................................................... 34
4.2.2 Performance of alkaline phosphorus filters................................................................... 35
5. Discussion ............................................................................................................................... 37
5.1. Robustness and function of on-site sanitation systems ......................................................... 37
5.2. Multi-criteria analysis methodology .................................................................................... 40

vii
 5.3. Removal of phosphorus and indicator bacteria in filters for on-site wastewater treatment
systems ................................................................................................................................... 42
5.4. Implications of the results in the Swedish context ............................................................... 43
6. Conclusions ............................................................................................................................. 45
References ...................................................................................................................................... 47
Appendices: Papers I–III

viii
List of papers

I Vidal, B., Hedström, A., Barraud, S., Kärrman, E., Herrmann, I. Assessing the
sustainability of on-site sanitation systems using multi-criteria analysis. (Manuscript)
II Vidal, B., Hedström, A., and Herrmann, I. (2018) Phosphorus reduction in filters for
on-site wastewater treatment. Journal of Water Process Engineering, 22, 210–217.
III Herrmann, I., Vidal, B., and Hedström, A. (2017) Discharge of indicator bacteria
from on-site wastewater treatment systems. Desalination and Water Treatment, 91,
365–373.

The author’s contribution to the papers is summarized in the table below:

Publication process
Paper Develop- Research Data Data processing Data Manuscript
Nr. ment of idea study design collection and analysis interpretation preparation Responding
for to reviewers
submission
I Contributed
Shared
Responsible Responsible Responsible Responsible NA
responsibility
II No
Contributed
Shared Shared Shared Shared
Responsible
contribution responsibility responsibility responsibility responsibility
III No
Contributed
Shared
Contributed Contributed Contributed Contributed
contribution responsibility

Responsible – developed, consulted (where needed) and implemented a plan for

completion of the task.
Shared responsibility – made essential contributions towards the task completion in
collaboration with other members of the research team
Contributed – worked on some aspects of the task completion
No contribution – with valid reason, has not contributed to completing the task (e.g.
joining the research project after the task completion)
NA – not applicable

ix
1. Introduction
Decentralized on-site sanitation systems are a steadily growing part of the wastewater
sector (Bradley et al., 2002; Gunady et al., 2015). In the US, Australia and Finland, about
20% of the population depends on on-site individual or community systems to treat their
wastewater (USEPA, 2014; Gunady et al., 2015; Martikainen et al., 2018). In Sweden,
approximately 700,000 on-site facilities (Ek et al., 2011) serve about 10% of the
population (Swedish EPA, 2014), including year-round and summer housing. The high
failure rate of on-site sanitation systems, e.g. up to 40% in Australia (Gunady et al., 2015)
and lack of compliance with the existing guidelines, e.g. 50% of the systems installed in
Sweden are estimated to provide inadequate treatment (Palm et al., 2002; Olshammar et
al., 2015), pose environmental, economic and social impacts. Poorly performing on-site
sanitation systems potentially release organic matter, nutrients and pathogens,
contributing to reducing water quality of natural waters and increasing risks to public
health. Inappropriate design, construction and maintenance of on-site systems for
domestic wastewater treatment are often the reason for the overall inadequate function
and performance (Gunady et al., 2015). Besides, the microbial quality of the effluent of
on-site sanitation systems is often neglected, unless sampling and analysis of indicator
bacteria are carried out. These systems could still represent a health risk in terms of
microbial contamination of ground water or receiving waters (Martikainen et al., 2018)
even if systems work hydraulically well.
Despite the fact that the application of fertilizers accounts for the major anthropogenic
perturbation of nitrogen (N) and phosphorus (P) cycles on a global scale (Steffen et al.,
2015), on-site sanitation systems in Sweden discharge as much P as municipal wastewater
treatment plants, accounting for 15% of the total anthropogenic release of P (Swedish
Environmental Emissions Data 2011). Research on the planetary boundaries and human
perturbations of the Earth system has shown that the biochemical flows of the nutrients
P and N are beyond the zone of uncertainty, calling for an urgent shift in the current
paradigm of development of human societies in order to maintain the resilience of the
Earth system (Steffen et al., 2015).
Often, systems that are simple in their design and have low cost and low maintenance
requirements are the ones implemented, sometimes to the detriment of the local
environment and human health (Bradley et al., 2002). The increasing environmental
awareness in the wastewater sector has set the focus on the sustainability of sanitation
systems, and the need for tools capable of assessing sustainability (De Feo and Ferrara,
2017). Evaluating the sustainability of on-site sanitation systems increases the knowledge
and understanding of the relations and trade-offs between indicators used to assess the
systems. Consequently, the evaluation of the potential impacts of on-site systems for
domestic wastewater treatment can also be improved (Diaz-Elsayed et al., 2017). Besides,
understanding the function and sustainability of on-site sanitation systems can facilitate
more informed decision-making when planning or selecting appropriate solutions in a
given context.

1
1.1. Aim and research objectives

The overall aim of the thesis was to improve the knowledge and understanding of the
sustainability of on-site sanitation systems, by assessing the sustainability of a number of
on-site sanitation options in a multi-criteria analysis and by investigating the performance
of existing on-site wastewater treatment facilities. Particular attention was given to the
removal of phosphorus (paper I and II) and bacteria (papers I and III), P being relevant
because of risk of eutrophication of natural waters and bacteria because of public health
concerns. Thus, the research objectives were
1. to define and evaluate a set of indicators to assess the sustainability of on-site sanitation
alternatives in a multi-criteria approach, considering different scenarios based on specific
contexts and preferences expressed in the form of weights and
2. to evaluate the treatment efficiency of existing on-site sand filters and P-filters by
conducting flow-proportional sampling and analyzing the treatment efficiency in terms
of organic matter, P and bacteria.

1.2. Thesis structure

This thesis is based on three papers, referred to as Papers I, II and III (Figure 1), that
discuss the sustainability and function of on-site sanitation systems. Paper I is a manuscript
and presents a multi-criteria analysis of nine on-site sanitation alternatives based on
sustainability indicators. In Figure 1, the indicators used in the analysis presented in Paper
I are grouped to emphasize nutrient-related indicators (yellow) used in Scenarios 1 and
2, energy and climate-related indicators (green) used in Scenario 3, and socio-economic
indicators (pink). Papers II and III are published and involve the evaluation and
assessment of twelve on-site wastewater treatment facilities featuring sand filters and P-
filters, with regard to their function and removal capacity in terms of P and indicator
bacteria. This thesis is presented in six chapters: Chapter 1 presents a brief introduction
to the thesis topic and the research objectives. Chapter 2 contains a literature review
providing an overview of the current status of on-site sanitation systems and the research
front with focus on sanitation technology for nutrients and microorganisms removal, as
well as the basis to assess the sustainability of sanitation systems using a multi-criteria
approach. Chapter 3 describes the methodology followed, namely the performed multi-
criteria analysis and the sampling study in the field. Chapter 4 presents selected results
from the three papers combined and Chapter 5 discusses the results in comparison to the
existing scientific literature. Chapter 6 presents the main conclusions from the work
presented. The three scientific papers are appended in their original form at the end of
this thesis.

2
 Paper II Function of on-site
sanitation systems
P removal
Paper I Robustness
N removal Paper III

Potential for Pathogen removal

nutrient recycling
Sustainability of on-site
sanitation systems Cost
Energy recovery

Energy use Social acceptance

Global warming
potential

Figure 1 Synthesis of the papers included in the thesis.

3
4
2. Background

2.1. Small-scale on-site sanitation systems

On-site sanitation systems (OSSs) are used to collect, store and/or treat domestic
wastewater in areas not connected to sewers and centralized wastewater treatment plants.
These facilities usually reduce the content of suspended solids, floatable grease, organic
matter, nutrients and pathogens from influent wastewater prior to discharge into the
receiving environment (USEPA, 2002). Conventional on-site treatment systems consist
of a septic tank, where the effluent is pre-treated, and a subsequent soil absorption field
or subsurface infiltration system. These are the most common on-site wastewater
treatment systems used, for example, in North America (USEPA, 2002), Australia (Green
and Ho, 2005) and Scandinavian countries (Heistad et al., 2006; Olshammar et al., 2015).
Constructed wetland technology has also been commonly used to treat wastewater when
space is available and is considered to be a reliable form of treatment system, with
subsurface systems being more common in Europe and free water surface systems in
North America and Australia (Vymazal, 2011).
Despite their long history of usage, OSSs often underperform due to inappropriate
installation and operation affecting their treatment capacity, and leading to the discharge
of poorly treated wastewater (Palm et al., 2002; Heinonen-Tanski and Matikka, 2017).
Furthermore, failures in their performance often lead to unnecessary extra costs and risks
to public health and water resources (USEPA, 2002). The scientific literature concerning
the performance of conventional systems, e.g. septic systems with soil infiltration, is
extensive and has mostly focused on the parameters affecting the treatment efficiency
(Rolland et al., 2009), the capacity to remove nutrients (Tanık and Çomakoğlu, 1996;
Vilpas and Santala, 2007) and the P removal capabilities of the soil and sand/gravel used
as infiltration material (Eveborn et al., 2012; Seguí et al., 2017).
The treatment efficiency of soil infiltration systems depends on the wastewater properties,
filter media used, temperature, biofilm development and hydraulic loads (Rolland et al.,
2009; Wilson et al., 2011). Despite their effective removal of organic matter and
suspended solids (Pell and Nyberg, 1989), the P-removal efficiency of the soil and
sand/gravel infiltration materials has been reported to vary considerably depending on
the occurrence of adsorption and precipitation processes, especially in a long-term
perspective (Eveborn et al., 2012; Seguí et al., 2017). Within this context, technologies
with physico-chemical methods have been developed to adapt or upgrade existing
treatment systems (Bunce et al., 2018), with reactive media filters for P removal showing
favourable results (Gustafsson et al., 2008; Cucarella and Renman, 2009). Studies on
substrates for P removal have discussed their potential as fertilizers and the plant
availability of the P (Cucarella et al., 2007). Sorptive media of different origin (natural
materials, industrial by-products and man-made products) have been also studied to
enhance P removal in constructed wetlands (Vohla et al., 2011). Similar filter bed systems
e.g. using sorptive media, but without wetland plants, have shown high treatment
5
efficiency with regard to organics, nutrients and bacteria (Jenssen et al., 2010; Paruch et
al., 2016). Nevertheless, further research is needed on the reliability of P-removal systems
using active media in full-scale applications and over prolonged periods of time, a
consideration that may limit their long-term sustainability (Bunce et al., 2018).
Alternative technologies to conventional systems, such as package plants, often include
additional biological treatment and polishing steps for pollutants removal through
filtration, absorption and adsorption (USEPA, 2002). Packaged plants use similar
conventional processes to those used in the larger wastewater treatment plants (WWTPs)
with primary, secondary and in some cases tertiary treatment, and they are often
automatized when chemical precipitation, pumping and/or aeration are present
(Libralato, 2013). Studies on package plants with sequencing batch reactors, activated
sludge or submerged biofilm reactors are found in the scientific literature (Hellström and
Jonsson, 2006; Vilpas and Santala, 2007). In general, treatment processes are well known
and they can deliver high removal efficiencies, but they are sensitive to operational
disturbances and problems often arise with changes in the wastewater load and when the
dosing equipment is not managed and refilled adequately (Hellström and Jonsson, 2006).
The overall quality of the effluent from different on-site wastewater treatment facilities
has been investigated in conventional systems with sand filters and package plants showing
varying treatment efficiencies (Vilpas and Santala, 2007; Kauppinen et al., 2014;
Heinonen-Tanski and Matikka, 2017; Martikainen et al., 2018), and in column
experiments for mineral-based sorbents for P removal (Renman and Renman, 2010;
Nilsson et al., 2013; Herrmann, et al., 2014). The number of enteric microorganisms in
the effluent of sand filters and package plants often exceed the European limit values for
good quality bathing waters (EU, 2006) or the effluent should be classified as poor quality
bathing waters at best (Heinonen-Tanski and Matikka, 2017). Furthermore, studies on
conventional OSSs in cold climates suggest that the microbial purification efficiency may
be reduced during winter but with no significant effect on nutrient removal
(Christopherson et al., 2005; Williamson, 2010; Kauppinen et al., 2014), although a
decrease of the P removal capacity in e.g. P-filters has been reported in column
experiments (Herrmann et al., 2014). Despite the importance of assessing the presence of
bacteria and viruses in the effluent of OSSs because of health implications, pathogen
removal has been less studied than nutrient removal, likely because of the existing
legislation on the latter (Martikainen et al., 2018). Further research is thus needed on
long-term removal of bacteria and viruses from OSSs, due to the risk of contamination
of surface, drinking and bathing waters.
As the simultaneous removal of N and P appears technically difficult on small scales
(Bunce et al., 2018), source-separation systems arise as a solution to remove and recover
nutrients more efficiently as the different waste fractions call for specific treatments
(Larsen et al., 2016). Source-separation systems were developed following a sustainable
water management approach and encouraged by the growing environmental concerns
and the increased focus on nutrient recycling (Larsen et al., 2013). Several studies have
shown their resource-efficiency potential in comparison to end-of-pipe technology

6
(Spångberg et al., 2014; Xue et al., 2016). Anaerobic digestion in up-flow anaerobic
sludge blanket septic tanks (Kujawa-Roeleveld et al., 2005), wet composting and urea
hygienization have been studied and implemented for blackwater treatment (McConville
et al., 2015). Moreover, alkaline dehydration of urine has emerged as an efficient and
hygienic solution to minimize the volumes when handling urine while reducing nitrogen
losses (Senecal et al., 2018; Simha et al., 2018). Aurin, a liquid fertilizer extracted from
urine and approved by the Swiss Federal Office of Agriculture was developed using the
so-called VUNA recycling technique, which stabilizes the source-separated urine and
removes pharmaceutical residues through an activated carbon filter (VUNA, 2018). More
bridges between research, policy and practice are needed to overcome the technical and
socio-economic challenges of source-separation systems, such as toilet design, blockages,
social acceptance, incentivisation and market creation (McConville et al., 2016).
Furthermore, better estimations and reliable data, e.g. emissions from blackwater
treatment, are needed to assess their overall sustainability. Other technologies for
separated treatment of toilet fractions include incinerating toilets (e.g. Cinderella®
Incineration toilet) which can be used in areas with physical restrictions in terms of soil
depths, steepness, high ground water levels or highly sensitive receiving waters (USEPA,
1999).
Besides organic compounds, nutrients and pathogen organisms, domestic wastewater
effluent also contains other unwanted pollutants. Recent screening studies on
micropollutants emissions from OSSs have identified large numbers of compounds in the
wastewater, including pharmaceuticals, personal care products, pesticides, phosphorus-
containing flame retardants (PFRs), artificial sweeteners and perfluoroalkyl substances
(PFASs) (Blum et al., 2017; Gros et al., 2017). Significant quantities of these substances
reach the aquatic environment (Gago-Ferrero et al., 2017) but no clear differences have
been found in terms of concentrations and removal rates between OSSs and WWTPs
(Gros et al., 2017; Schaider et al., 2017). When comparing different treatment systems,
including biofilters with P-filter units, biological sand filters and different types of
wetlands, the better oxygenation of the vertical flow constructed wetlands appeared to
remove micropollutants more efficiently than the rest (Matamoros et al., 2009; Vymazal,
2011). Granular activated carbon alone or in combination with Polonite® alkaline
material showed the highest removal efficiencies (above 97%) of a set of 83
micropollutants when comparing five different sorbents (Rostvall et al., 2018). Research
on WWTPs acting as potential point sources for microplastics emissions to the
environment has shown that microplastics are mainly removed during the different
settling processes (Talvitie et al., 2015), with emissions of microplastics being lower after
a tertiary treatment than after only a secondary one (Carr et al., 2016). Despite the lack
of studies on microplastics emissions from OSSs, microplastics are expected to be retained
to a large extent in the sewage sludge or in the soil infiltrations systems (Magnusson et
al., 2016).
In terms of sustainability and environmental impacts, package plants have been reported
to have greater carbon footprint and eutrophication potential when compared to soil and
source-separation systems (Lehtoranta et al., 2014). The impact of source-separation
7
systems, such as urine diversion or blackwater separation, has been shown to have lower
global warming impact (Lundin et al., 2000) and to be more energy efficient when
compared to conventional centralized options, although ammonia emissions from storage
and land application of urine and blackwater must be tackled (Spångberg et al., 2014).
The selection of the system boundaries will ultimately influence the scale of the impacts
of the systems being compared, and source-separation alternatives are often advantageous
when an extended system, e.g. replacement of mineral fertilizers, is included (Lundin et
al., 2000). The increasing demand for sustainable water management systems requires
systems analysis-thinking and the development of frameworks that integrate sustainability
indicators (Hellström et al., 2000). Environmental indicators based on life-cycle analysis
and material-flow analysis are often used when comparing different OSSs (e.g. Diaz-
Elsayed et al., 2017; Thibodeau et al., 2014; Weiss et al., 2008). However, more
knowledge is needed on how to comprehensively assess the technical robustness of the
systems or qualitative indicators such as the social acceptance, as well as on how both
qualitative and quantitative indicators can be integrated into an overall assessment.

2.2. Removal of phosphorus and microorganisms in on-site sanitation systems

Phosphorus removal from wastewater streams has primarily been implemented to

decrease the eutrophication impact in receiving waters (Yeoman et al., 1988). Moreover,
more stringent regulations on emissions and the growing limitation of P availability on
the global scale have enhanced its removal and recovery (Cordell, 2013). Phosphorus can
be removed from wastewater by physico-chemical methods or advanced biological
treatment or a combination of both (Yeoman et al., 1988). In conventional on-site
systems, which use gravel and sand for infiltration, biological removal of P is of minor
importance as the P-uptaking biota (plants and microorganisms) is generally not harvested
(Eveborn et al., 2012). Chemical processes for P are therefore of greater importance, as
phosphates are adsorbed by most sediments, often forming a wide range of stable minerals
after combination with metals cations such as iron, aluminium, manganese and calcium
(Robertson et al., 1998). The removal of P in conventional systems is a finite process
dependent on the number of sites that can react with the dissolved phosphates, the clay
and organic matter content and the pH (USEPA, 2002). Hence, P removal is limited by
the soil capacity to hold sorption processes (onto Fe and Al oxide phases) and mineral
precipitation reactions (of Al, Fe and Ca phosphates), which mainly occur in the vadose
(unsaturated) zone and are often difficult to distinguish between each other (Robertson
et al., 1998). Varying results have been reported on P removal in soil-based treatment
systems (Christopherson et al., 2005; Vilpas and Santala, 2007; Eveborn et al., 2012; Seguí
et al., 2017), likely due to the age of the systems and soil characterization, sampling
approach (e.g. flow-proportional versus grab samples) and estimation methodologies.
Nevertheless, the long term P removal in soil-base systems has been suggested to be lower
than generally reported (Eveborn et al., 2012). The removal of P from wastewater can be
enhanced by the use of reactive media with sorption properties specifically targeting P in
systems where, in contrast to conventional soil-based systems, the P is removed from the

8
wastewater by processes of sorption or direct precipitation in the reactive media (Arias et
al., 2001). The inorganic P accumulates on the surface of the reactive components of the
filter media, e.g. calcium or iron, and thus the removal capacity is dependent on the
mineral content of the reactive media (Bunce et al., 2018).
Chemical dosing of salts (e.g. Ferric/ferrous and aluminium salts) is used to promote P
precipitation in the septic tank or in the process units or reactors of package plants (e.g.
activated sludge). The P precipitates in the wastewater and the residual solids are removed
by settling or filtration, with typical removal rates proportional to the mass of added
precipitant and often achieving effluent P concentrations of 1 mg L-1 after conventional
gravity settling (Tchobanoglous et al., 2014; Bunce et al., 2018). Enhanced biological P
removal (EBPR) primarily occurs when phosphorus accumulating organisms (PAO)
accumulate more P than needed for metabolic processes (luxury uptake) through different
biochemical mechanisms (Comeau et al., 1986). EBPR has traditionally been used in
conventional activated sludge systems and recent applications implement it in membrane
bioreactors and sequencing batch film reactors (Bunce et al., 2018). However, the
metabolism and behaviour of the different PAO is not yet fully understood and their
effect on nutrient removal requires further research specifically aimed at small-scale
wastewater treatment systems (Oehmen et al., 2007).
Ultimately, more comprehensive research on full-scale facilities is needed to further
estimate the P-removal capacity as well as the performance of conventional systems in
view of new regulations for OSSs.
Wastewater contains a large range of pathogenic microorganisms that can affect public
health through direct consumption of contaminated water or recreational contact
(USEPA, 2002). Bacteria and viruses can move relatively long distances through the soil
under saturated conditions, increasing the risk of contamination of groundwater and
surface waters (Hagedorn et al., 1978; Scandura and Sobsey, 1997). Stevik et al. (2004)
reviewed the mechanisms influencing the retention and removal of bacteria in infiltration
systems, namely straining and adsorption. Physical straining is the blocking of movement
caused when the pores are smaller than the bacteria, and it is influenced by the grain size
of the porous media, bacterial cell size and shape, hydraulic loading and clogging of filter
media. Adsorption dominates in porous media where the pores are larger than the bacteria
and may be influenced by physical (e.g. presence of organic matter and biofilm), chemical
(e.g. ionic strength and pH) and microbiological (e.g. hydrophobicity and bacterial
concentration) factors (Stevik et al., 2004). Microorganisms are often attached or
associated to particulate material, and they are thus reduced in the pre-treatment phase
(e.g. septic tank) typically in the range of 50-90% (0.5-1 log) (Stenström, 2013).
Microorganisms removal in soil infiltration systems varies depending on the soil, the load
and hydraulic conditions, but it is generally in a similar range as for conventional
wastewater treatment plants (without disinfection) with reduction ranges >2 logs for
bacteria and viruses (Stenström, 2013).
In package plants with activated sludge process, the removal of microorganisms mainly
occurs due to sedimentation and adsorption or incorporation into the biological floc,
9
encapsulation within the sludge solids and inactivation by other microorganisms.
Pathogenic microorganisms removal during activated sludge treatment is highly variable
depending on the type of organism and detention times, with removal rates for all
pathogens ranging from 40% to 90% (Henze et al., 2008). Package plants with
continuous-flow and suspended-growth aerobic systems have been reported to remove
microorganisms (faecal coliform and virus) in the range of 1 to 2 logs (USEPA, 2002). A
recent study comparing five different package plants (with biorotors, biofilters and active
sludge process) and three different sand filters (with and without P precipitation) reported
higher concentrations of E. coli and enterococci in the effluents of the package plants than
in the sand filters (Heinonen-Tanski and Matikka, 2017). In a pilot plant study,
bioreactors with suspended growth with membranes immersed in an aeration tank were
reported to remove indicator microorganisms more efficiently (99.99% removal) than
e.g. activated sludge with denitrification and sand filtration and up-flow anaerobic sludge
blanket treatment (Ottoson et al., 2006). In constructed wetlands and lagoons, the
removal of pathogens is influenced by the retention time, sedimentation, sunlight and
antagonistic flora (Henze et al., 2008) and removal values between 0.5–3 log reduction
have been reported for bacteria (Stenström, 2013).
The main source of microorganisms in domestic wastewater is the faeces fraction and
hence, from a hygienic point of view, it is reasonable to treat them separately instead of
in combination with the other wastewater fractions (e.g. greywater and urine). Moreover,
separate faeces handling has gained significant attention in the new sanitation context
with regard to safe nutrient recycling (Winker et al., 2009). The inactivation of pathogens
that are present in faecal matter is determined by temperature, pH and time, with
anaerobic treatment or addition of either urea or ammonia being common treatment
processes (Larsen and Maurer, 2011).
Previous studies have discussed the effect of alkaline P-sorption media on the removal of
pathogens as the high pH (>12) contributes to the initial inactivation of bacteria (Jenssen
et al., 2005; Heistad et al., 2006; Nilsson, Lakshmanan, et al., 2013). Results from column
experiments suggest that pH significantly affects the removal of bacteria in Polonite®
alkaline filter media loaded with high concentrations of organic material (Nilsson et al.,
2013b). Additionally, pilot studies with sand filters have discussed the relation between
efficient removal of pathogens and nutrients but have not provided a comprehensive
correlation (Kauppinen et al., 2014). Such correlation between removal of nutrients and
pathogens has not been confirmed at full-scale (Martikainen et al., 2018). For this reason,
further research is needed to improve the knowledge related to the quantification of
indicator organisms for faecal contamination in wastewater effluents and removal
efficiency of OSSs, considering the hygienic risk that inadequately treated effluent may
pose to nearby water bodies.
There are no legal regulations regarding enteric microorganisms’ concentrations in
effluents in Sweden at the time of writing, despite their presence in surface waters
(Heinonen-Tanski and Matikka, 2017). Consequently, the EU bathing water directive
(2006/7/EC) is often used as a guideline (EU, 2006) with concentrations set for excellent

10
bathing quality in inland and coastal waters ranging between 200 and 100 CFU/100 mL
for intestinal enterococci and 500 and 250 CFU/100 mL for E. coli, respectively.
2.3. Sustainability assessment in on-site sanitation

Assessing the sustainability of treatment options for domestic wastewater requires a

systematic and comprehensive methodology to combine the main dimensions of
sustainability, namely economic, environmental and social (Bradley et al., 2002). Defining
and measuring the sustainability of wastewater systems has been on focus for the last two
decades, as for example shown in Lundin et al. (1999), Hellström et al. (2000) and
Balkema et al. (2001). Sustainability can be defined by a number of criteria and associated
indicators that are measurable, or at least assessable, to promote the practical use of the
sustainability concept (Hellström et al., 2000). Sustainable technology for domestic
wastewater treatment has been defined as that which does not threaten the resources’
quantity and quality, depending on different conditions (Balkema et al., 2001). Similarly,
sustainable sanitary systems have been defined as those which provide the necessary services
without compromising human health, use a minimal amount of resources, and do not
have a long-term negative environmental impact (Lundin et al., 1999).
Earlier research has aimed to assess the sustainability of sanitation systems following
different approaches and applying a wide range of methods, e.g. based on single indicators
versus multidisciplinary indicators (Molinos-Senante et al., 2014). Several studies have
discussed and proposed indicators based on sustainability principles to evaluate wastewater
treatment systems based on literature data (Hellström et al., 2000; Balkema et al., 2002;
Tjandraatmadja et al., 2013). Furthermore, life cycle approaches (Tidåker et al., 2007;
Lehtoranta et al., 2014; Xue et al., 2016; Schoen et al., 2017), environmental systems
analyses (Weiss et al., 2008; Spångberg et al., 2014) and sustainability assessments
(Molinos-Senante et al., 2014; Diaz-Elsayed et al., 2017) have been applied with the same
purpose. A review of the life cycle assessment (LCA) approach applied to wastewater
treatment systems highlighted the need to link LCA methodology to economic and social
evaluations to complete the overall picture of sustainability (Corominas et al., 2013).
Some previous studies have evaluated and compared on-site facilities with different
treatments, including e.g. source-separating options, semi-collective filter beds and
connection to the main wastewater treatment plant (Lennartsson et al., 2009). Several
studies included inspection and sampling of the facilities (Hellström and Jonsson, 2006;
Vilpas and Santala, 2007; Jenssen et al., 2010). Tidåker et al. (2007) compared three on-
site sanitation systems (urine separation, blackwater separation and chemical precipitation
in septic tank) by evaluating the energy use in a life cycle perspective, the recycling
potential and the emissions of nutrients. Weiss et al. (2008) applied LCA methodology
to study the environmental impacts and use of natural resources of four on-site sanitation
systems (infiltration, chemical precipitation and P removal using two types of filter
media). Furthermore, the potential trade-off between the reduction of emissions and the
increase of life cycle impacts of six alternatives was studied by Lehtoranta et al. (2014),
who concluded that the major causes of both impacts were the direct on-site emissions,

11
the electricity consumption and the centralization of the sludge treatment. The challenges
of assessing the sustainability of small wastewater treatment systems was thoroughly
discussed in Molinos-Senante et al. (2014), where seven different technologies for
secondary treatment in small wastewater treatment plants were compared by aggregating
system indicators into a composite sample that provided a global measure of sustainability.
The sustainability of five water and sanitation systems was compared based on normalized
metric scores by Xue et al. (2016) and Schoen et al. (2017), who included the life cycle
eutrophication potential, the energy consumption, the global warming potential, the cost
and the human health impact. In addition, a scenario with increasing variability of
climate-related factors, such as precipitation and temperature, was discussed by Kohler et
al. (2016) with regard to the fragility (the degree to which a system loses functionality)
of on-site sanitation systems. The study suggested that weather variability affected the
performance of on-site sanitation systems and their fragility, e.g. number of repairs
increased, and the consequences in terms of costs and, environmental and health impacts
should be considered in the planning, design and operation of the systems.
Multicriteria analysis techniques are widely used to systematically assess different
alternatives by combining individual criteria and incorporating weights that represent the
decision-maker’s priorities (Dodgson et al., 2009). Despite the existing literature covering
sustainability assessments of sanitation systems, there is a need for further research into
the different implications that indicators may have.
The existing literature on multi-criteria decision analysis describes numerous methods for
aggregating and weighting the criteria (Rowley et al., 2012; Kabir et al., 2014;
Tscheikner-Gratl et al., 2017). The importance of choosing the right method lies in the
fact that its very choice introduces subjectivity into the analysis (Rowley et al., 2012).
The ELECTRE III method in particular is part of the ELECTRE family of outranking
methods for multi-criteria decision analysis proposed in the mid-sixties (Roy, 1978). The
ELECTRE method family has been widely applied in sustainability-related research
(Velasquez and Hester, 2013). A review of multi-criteria decision-making methods by
Kabir et al. (2014) showed that ELECTRE method family was the second single most
used method in the literature (15.1%) between 1980 and 2012 in the field of water and
wastewater after the analytic hierarchy process (28.3%).
The novelty of ELECTRE III was the introduction of pseudo-criteria by using
preference thresholds to define a “buffer zone” between strict preference and indifference
when comparing the performance of two alternatives, in contrast to the use of true-
criteria, which has a strict preference for the best performance and does not account for
uncertainty (Rowley et al., 2012). By adding flexibility to the comparisons, uncertainty
and vagueness are thus taken into account, making the method less sensitive to variations
of the input data and parameters (Bertrand–Krajewski et al., 2002). ELECTRE III is a
non-compensatory aggregation method, hence a bad score on one indicator is not offset
by good scores on another indicators, allowing the method to provide a strong
sustainability perspective (Rowley et al., 2012).

12
3. Methods

3.1. Multi-criteria analysis

3.1.1. Sustainability indicators and alternatives for on-site sanitation

The selection of sustainability indicators was based on previous research proposing

sustainability indicators to assess wastewater systems (Balkema et al., 2002; Hellström et
al., 2000; Lennartsson et al., 2009; Molinos-Senante et al., 2014; Schoen et al., 2017).
Sustainability criteria were organized into five main categories, namely Environmental,
Economic, Socio-cultural, Technical and Health-related, and a number of assessable
indicators (qualitative and quantitative) were defined for each criterion. A brief
description of the twelve defined indicators is given below.
Nutrient removal referred to the capacity of the system to remove nitrogen (N) and
phosphorus (P) from the influent wastewater. Two sub-indicators, namely removal of N
and P, were quantified based on previous studies (Lennartsson et al., 2009; Olshammar
et al., 2015; Palm et al., 2002; Vidal et al., 2018) and considered equally important, and
thus equal weight was given to each sub-indicator.
Potential for nutrient recycling referred to the potential agricultural reuse of the different
waste fractions produced in the systems in relation to the nutrient content (N and P) in
each fraction as calculated for the nutrient removal indicator. The sludge from the septic
tanks, the sand from the sand filters, the Polonite® filter material from the P-filters, the
blackwater and the urine were considered potential sources of nutrients. This indicator
was quantified based on the current practice in Sweden when data was available, e.g.
about 34% of the generated sludge is reused as soil conditioner (Statistics Sweden, 2018),
or based on assumptions when data was not available, e.g. 100% reuse of the blackwater
and urine, 5% of Polonite® filter material, and 0% of the sand from the sand filters was
considered.
Global warming potential (GWP) accounted for the greenhouse gas (GHG) emissions in kg
CO2-equivalents (eq) released during: the production of the alternatives’ components and
materials (e.g. tanks, pipes, filter materials) and their transport (e.g. distance from the
production site), system installation and operation (e.g. electricity consumption) and
maintenance (e.g. collection of septic sludge, replacement of chemicals and P-filter), as
well as the post-treatment of the fractions that were not treated on-site (indirect nitrous
oxide emissions from ammonia emissions during storage of sludge, blackwater and urine).
The end-of-life phase was excluded from the analysis. The calculations were based on
LCA methodology standards (ISO 14040, 2006) following the global warming potential
impact assessment method [v1.0.1, January 2015] and the ELCD 3.2 database (European
reference Life Cycle Database, 2016).
Cumulative energy demand (CED) referred to the primary energy used during the
production, transport and installation of the components and materials, during operation

13
and maintenance of the sanitation alternatives and during post-treatment of the fractions
that were not treated on-site (sludge, blackwater and urine). The end-of-life phase was
excluded from the analysis. The calculations were based on the same LCA methodology
standards as described for GWP.
Energy recovery referred to the possibility to obtain energy in the form of biogas produced
from the collected septic sludge. This indicator was evaluated qualitatively with a three-
point ordinal scale that classified the energy recovery of the alternatives as low, medium
or high, and was estimated proportionally to the volume of sludge produced in each
alternative based on the composition of the different wastewater fractions according to
Jönsson et al. (2005).
Capital cost was based on the cost of the investment to purchase the different components
and the manpower required for the installation of each OSS alternative multiplied by the
annuity factor, which considers the amortization time and an interest rate of 4%
(assumed).
Operation and maintenance (O&M) cost referred to the yearly cost for the operation and
maintenance of the alternatives, which included the collection and transport of
blackwater, urine and the sludge from the septic tank; electricity use; purchase of
consumables (chemicals, P-filter) and components (change of pump); and check-up
service including effluent sampling and analysis.
Social acceptance was defined as the user-friendliness of the alternatives with regard to the
convenience, effort and degree of complexity of operating the system, from the user’s
perspective. The indicator was assessed qualitatively on a five-point ordinal scale from
very low to very high acceptance. The alternatives considered the most socially accepted,
namely the sand filter and drain field, were used as reference because they represent the
most common OSSs installed in Sweden (Olshammar et al., 2015), users are familiar with
them and they are considered convenient due to their simplicity (Lennartsson et al.,
2009). The other alternatives were assessed in comparison to the sand filter and drain
field, in terms of how the “inconvenience” for the users increased when adding different
components to the OSS.
Robustness was defined using two sub-indicators, namely the risk of failure of the system
and the adaptability to flow fluctuations. The indicator was assessed qualitatively on a
three-point scale from low to high robustness. The sub-indicator “risk of failure”
accounted for the possibility of the system to encounter a technical problem that may
hinder its treatment capacity, and the likelihood of such an incident to happen. The sub-
indicator “adaptability to flow fluctuations” accounted for the capacity of the system to
adapt to changes in the quantity and quality of the flow, e.g. increase in the average water
consumption because of greater presence of users, or periods of absence of users when
the system is not in use. The “risk of failure” was considered to be of higher importance
(2/3) than the “adaptability to flow fluctuations” (1/3) because of the more severe
implications.

14
Risk of pathogen discharge was based on a qualitative assessment of the capacity of the OSS
to remove pathogens from wastewater prior to discharge in the surrounding
environment. The indicator was assessed on a five-point ordinal scale from very low to
very high risk, based on the number of barriers included in the systems that potentially
have pathogen removal capacity and thus decrease the pathogens load (e.g. filter media,
chemical precipitants). The receiving waters groundwater/surface waters were also taken
into account.

3.1.2. The ELECTRE III method

The ELECTRE III method was selected with aid from the workflow schematic proposed
by Rowley et al. (2012) for sustainability analysts to compare the alternatives for on-site
sanitation. A flowchart summarizing the methodology followed in ELECTRE III
method is shown in Figure 2.

Problem definition Development of alternatives Performance matrix

and indicators

Definition of thresholds

Veto (v)

Meeting with Preference (p)

stakeholders
Indifference (q)
Weights

Generation of a Generation of
concordance index discordance indices

Generation of
credibility index

Comparison and ranking of

alternatives

Figure 2 Summary of the steps followed when applying the ELECTRE III method. The step described
in grey scale (Generation of discordance indices) was not used in the study.

15
The alternatives were assessed using indicators in an evaluation matrix where the best
outcome was represented by the maximum evaluation of each indicator. In ELECTRE
III, the construction of an outranking relation between two alternatives a and b (a S b) is
based on two major concepts, the concordance (the majority principle) and the
discordance (the respect of minority principle) from which a credibility index is calculated
that corresponds to a value of “outranking” of a with regard to b (Figueira et al., 2005).
To calculate the concordance index C(a, b), the alternatives were evaluated against each
indicator pairwise (Eq. 1), multiplying the partial concordance values ci (a, b) obtained
when comparing alternative a to b by the weights allocated by the reference group. The
larger C (a, b), the stronger the evidence that alternative a is preferred over b (Vincke,
1992). Preference (pi) and indifference (qi) thresholds were defined, as a constant number
or as a percentage of the performance value gi (a), for each indicator and used to calculate
the concordance values ci (a, b) (Eq. 2).

σൌͳ ™‹ ‹ ሺƒǤ„ሻ
‹ሺƒǡ„ሻൌ (Eq. 1)
σൌͳ ™

Where:
‹ˆ‰‹ሺƒሻ൅“‹൒‰‹ሺ„ሻ

‹ሺƒǡ„ሻͲ‹ˆ‰‹ሺƒሻ൅’‹൑‰‹ሺ„ሻ

୮೔ ା௚೔ ሺ௔ሻି௚೔ ሺୠሻ

otherwise (Eq. 2)
௣೔ ି௤೔

Where:
ci (a, b) is the concordance value on indicator i,
gi (a) is an individual evaluation of the alternative a for indicator i,
qi is the indifference threshold which determines if one alternative is “weakly preferred”
over another alternative with respect to a given indicator i even though its evaluation
may be (slightly) lesser in value, and
pi is the preference threshold, which determines if one alternative is “strongly preferred”
over another alternative in terms of indicator i.

The calculation of the discordance index Di (a,b) requires the definition of veto thresholds
(vi), which expresses the possibility of the alternative a to be discredited if it is exceeded
by the performance of b by an amount greater than the veto threshold, regardless of the
other indicators. The discordance index Di (a,b) was zero for all pairs of alternatives since
no veto threshold (vi) was used in this study.
An index called the degree of credibility of the statement a S b (G(a,b)), which indicates
the extent to which a outranks b, was calculated by aggregating the discordance and

16
concordance indexes. Since the discordance index was zero in this study, the credibility
of the outranking relation was equal to the concordance index C(a,b).
The ranking of the alternatives was determined by two preliminary rankings based on the
values of G(a,b), namely descending and ascending preorders or distillations. In the
descending distillation, the process ranks alternatives from the best to “less good”
alternatives, whereas in the ascending distillation the alternatives are ranked from the
“least bad” to the worst. These distillation processes represent the number of alternatives
that are outranked by Si and the number of actions that Si outranks, considering a
threshold s (λ), as described in detail in Roy (1978). A final ranking is the result of the
intersection of the two distillations.
The ELECTRE III-IV software version 3.x was used for the computations (Almeida Dias
et al., 2006). Further detailed descriptions of the methodology can be found in the
scientific literature (Roy, 1990, 1996; Vincke, 1992; Figueira and Roy, 2002).

3.1.3. Reference group and assignation of weights

The panel method was used with a reference group (Rowley et al., 2012) to weight the
importance of the indicators presented in section 3.1.1. Six representatives from a variety
of stakeholders relevant in the on-site sanitation sector in Sweden formed the reference
group, namely representatives from the responsible environmental authority for OSS
(Swedish Agency for Marine and Water Management), the Swedish Homeowners
Association, the Federation of Swedish Farmers, the Swedish Waste Management
Association and two representatives of advisors and communicators working in the small-
scale wastewater treatment sector, one from the National Platform for On-Site Sanitation
and one from the Centre for Water Development in Norrtälje.
The members of the reference group were contacted by email and provided with
information about the study, together with an online questionnaire which they
individually used to allocate weights to the indicators. During a later group meeting, the
indicators were discussed and the members of the reference group endorsed the weights
already given online or modified them, individually. The indicators were ranked from
the most to the least important and individual weights were allocated to each indicator
by giving 100 points to the most important indicator(s) and subsequently assigning less
points (from 100 to 0) to the other indicators, depending on how important they were
considered in relation to the most important one. As the group discussion did not result
in a consensus, the arithmetic mean of the normalized weights was used (Eq. 3).

ͳ ™ ȗ ‹ ήͳͲͲ
‹ ൌ σ‹ൌͳ  Eq. (3)
 ȗ

17
Where:
Wi = is the weight of indicator i,
wi* = are the points allocated by a stakeholder for each indicator, between 0 and 100,
W* = are the total points given by a stakeholder to all the indicators, and
n = is the number of stakeholders.

3.1.4. Entropy study – discriminating power of the indicators

In contrast to subjective weighting methods based on collective judgements, objective

weighting methods apply statistical models to calculate the implicit weight of the
indicators based on their quantitative values (Ibáñez-Forés et al., 2014). The entropy
method (Zeleny, 1982) is an objective weighting method that measures the distribution
of the data based on the assumption that the broader the dispersion of the performances
of an indicator for a set of alternatives, the stronger the discrimination between the
alternatives. Therefore, an indicator with a wide dispersion will be more likely to be
decisive in the ranking process than another indicator with smaller dispersion. Based on
Barraud et al. (2004), the entropy (Ei) of an indicator i was defined as:
ିଵ
‫ܧ‬௜ ൌ ή σ௡௝ୀଵ ܽԢ௝௜  ή ŽሺܽԢ௝௜ ሻ Eq (4)
୪୬ሺ௡ሻ

ܽԢ௝௜ ൌ ܽ௝௜ Ȁ σ௡௜

௜ୀଵ ܽ௝௜ Eq (5)

Where:
n = is the number of alternatives,
ni = is the number of indicators,
aji = is the performance of alternative j on indicator i, and
a’ji = is the normalized performance of alternative j on indicator i.
A dispersion parameter defined as Di = 1 – Ej , which shows the indicator’s entropy, was
used in its normalized dimensionless form wi as described in Barraud et al. (2004):
‫ݓ‬௜ ൌ  ‫ܦ‬௜ Ȁ σ௡௜
௜ୀଵ ‫ܦ‬௜ Eq (6)

For each indictor, a measure mj was calculated by the product of the dimensionless
dispersion parameter (wi) and the normalized weights given to each indicator by the
reference group (pi), as described in Barraud et al. (2004):

݉௜ ൌ  ‫ݓ‬௜  ή  ‫݌‬௜  Eq (7)

18
Where:
mi = is the measure obtained for each indicator
wi = is the dimensionless dispersion of each indicator, and
pi = is the normalized weight given by reference group to each indicator.

3.1.5. Scenarios analysis

A scenarios analysis was carried out to investigate different plausible settings based on
socio-economic and geographic factors, by changing the weights given to certain
indicators (Paper I). Scenario 0 was the baseline scenario with the initial weights allocated
by the reference group. Scenario 1 represented areas where surface waters generally do
not have special protection status according to Swedish legislation (Swedish EPA, 2006).
Furthermore, in this scenario, the possibilities to recycle nutrients into farmland were
limited as it is the case in e.g. areas in northern Sweden. Hence, the indicators related to
nutrient removal and potential for nutrient recycling were given the lowest weights.
Scenario 2 represented areas where the removal of nutrients was important due to the
presence of sensitive receiving waters, and with possibilities to recycle nutrients onto
farmland. This scenario represented typical farmland areas with eutrophicated receiving
waters. The indicators related to nutrients removal and potential for nutrients recycling
were given the highest weights.
Scenario 3 represented a context where a sociopolitical strategy has resulted in higher
demands on climate change mitigation, implicating the need for decreasing atmospheric
emissions. Hence, the indicators related to potential for energy recovery, cumulative
energy demand and global warming potential were assigned the highest weights.
The weights of the indicators mentioned in each scenario were modified using Simo´s
card method (Simos, 1990; Figueira and Roy, 2002). The method consists of placing the
indicators on cards for better visualization and organizing them in order of importance,
with the possibility of adding blank cards in between the indicators to represent larger
differences in the importance. The schematic representation of the cards’ order can be
found in the Supplementary Material of Paper I.

3.1.6. On-site sanitation alternatives

Nine on-site sanitation alternatives (A1–P2) were selected based on relevant literature
and discussions with practitioners, including conventional commonly used systems,
namely sand filters and drain fields, as well as package plants and source-separation
systems. These alternatives are briefly described below.

19
A1. Septic tank and sand filter
Wastewater from the household is collected in a three-chamber septic tank (2.2 m3) made
of fiberglass-reinforced polyester (FRP) and pumped to a sand filter constructed below
the ground surface. The wastewater is spread using perforated distribution pipes and
treated through a variety of physical, chemical and biochemical reactions and processes
during filtration. The effluent is collected at the bottom of the filter by drainage pipes. A
conventional sand filter constructed according to the Swedish standard was assumed
(Swedish EPA, 2003a), with a layer of filter media of 0.8 m (sand/gravel between from
0 to 8 mm in size) and a surface area of 30 m2. Both distribution and drainage pipes are
embedded in a 0.2 m coarse gravel layer. The sludge from the septic tank is transported
to the nearest WWTP for anaerobic treatment (valid for all alternatives with septic
sludge).
A2. Septic tank and drain field
A design similar to A1 is assumed. However, the wastewater is not collected after the
buried infiltrative material, a layer of coarse gravel of 35 cm (Swedish EPA, 2003b), but
instead continues infiltrating and percolating through the underlying soil. It was assumed
that the properties of the soil allow infiltration of the wastewater.
A3. Septic tank, sand filter and P-filter
Same design as in alternative A1 with the addition of a polishing step for phosphorus (P)
removal (alkaline P-filter), using the filter media Polonite® after the sand filter. The water
is transported from the outlet of the sand filter to the P-filter by gravity.
A4. Septic tank with chemical P removal and sand filter
Wastewater from the house is collected in a three-chamber septic tank. A chemical
precipitation unit with polyaluminum chloride is installed, in the kitchen (e.g. under the
kitchen sink) and dosed when needed. The flocculation and sedimentation occurs in the
septic tank and therefore the tank volume is larger (4 m3) than in alternatives A1-3. The
septic sludge is assumed to contain more P than in A1-3. The lifespan of the sand filter
was assumed to be longer (25 years instead of 20 years as in A1-3), because some of the
suspended solids and the BOD are removed during the flocculation process, resulting in
a lower load to the sand filter (Palm et al., 2002; Weiss et al., 2008).
A5. Septic tank with P precipitation and drain field
Same design as in A4, though in this case the wastewater is not collected after the filter
media – a layer of gravel (between 12 to 32 mm in size) of 35 cm (Swedish EPA, 2003b)
– but instead infiltrates into the soil. It was assumed that the properties of the soil allow
infiltration of the wastewater.
S1. Greywater and blackwater separation
This system is based on the separate collection of greywater and blackwater. The
greywater was assumed to be collected in a septic tank followed by a sand filter. The

20
lifespan of the sand filter was assumed to be higher than the conventional one (25 years
instead of 20), because most of the nutrients and BOD are already removed with the
blackwater. The sludge from the greywater tank is transported to a WWTP. The
blackwater is collected in a holding tank which is emptied once a year. For collection,
storage and treatment reasons, it is important that the toilet flush volume is small and
therefore the use of a low-flush toilet (about 1 L per flush) was assumed in the analysis.
The collected blackwater is transported to a central treatment facility using urea treatment
(1% urea) for hygienization.
S2. Greywater and faeces, urine diversion
This system is based on a separate collection of greywater, urine and faeces fractions from
separating toilets. The greywater and the faeces are collected together in a septic tank and
subsequently conveyed to a sand filter. The sludge from the septic tank is transported to
a WWTP. The urine is collected in a container and transported to a centralized facility
for hygienization (six-months storage).
P1. Package plant with alkaline P-filter
This package plant consists of one single unit buried underground to which the
wastewater from the household is transported by gravity. The mechanical treatment
occurs in the first three chambers of the plant, followed by two bioreactors with aerators
that provide oxygen to the water. A fraction of the water is returned to the first chamber.
The effluent infiltrates through an alkaline P-filter with Polonite® material placed in a
bag in the centre of the treatment unit. The sludge is collected from the three chambers
that act as septic tanks, and it is transported to the nearest WWTP for anaerobic treatment.
P2. Package plant with chemical P removal
This package plant consists of one single unit buried underground to which the
wastewater from the household is transported by gravity. The plant operates in a two-
phase semi-continuous regime. The main phase corresponds to a continuous process of
activated sludge, and the second phase corresponds to the discharge of excess sludge and
cleaning of the filter when the level of wastewater in the equalization tank is low. The
raw wastewater is first collected in an equalization tank and then pumped to an aerated
water-processing tank with activated sludge where primary settlement occurs and
chemical dosing equipment is installed. A small sand filter is used as a final polishing step.
All sludge is collected in a separated tank inside the unit. The sludge is collected and
transported to the nearest WWTP for anaerobic treatment.

21
3.2. Field investigation of existing sand filters and alkaline P-filters

3.2.1. Selection of facilities

The facilities were selected based on the information found in municipal databases in
cooperation with seven Swedish municipalities. When facilities with sand filter and/or
P-filter were found, the house owners of these facilities were contacted by telephone to
get their permission to visit and inspect them. Thirty-four facilities were inspected to
determine their suitability to the sampling. The inspections consisted of checking the
outlets of the sand filter to ensure they were designed in a way that would enable flow-
proportional sampling. For P-filters, inlets and outlets were both checked for their
accessibility for sampling and whether at least one of them was suitable for measuring the
flow. Twelve facilities were selected and investigated (Table 1), including three sand
filters without any downstream treatment (A–C), four sand filters with subsequent P-
filters (D–H) and four P-filters placed in package plants (I–L).
Table 1 Investigated on-site wastewater treatment facilities (extracted from Paper II). The columns Years
in use refer to the number of years the treatment units had been in use at the time of sampling.

Facility Main treatment Years in P-filter unit Years in No. of Frequency of use
unit use use users
Sand filter,
A 6-7 - - 3 Year-round
Swedish standard
Sand filter, with
B 2-3 - - 2 Year-round
biomodules
Sand filter, with
C 1 - - 5 Year-round
biomodules
Sand filter, Bag with Polonite,
D 1-2 1-2 2 Year-round
Swedish standard down-flow
Sand filter, Wavin-Labko tank
E 6-7 6-7 5 Year-round
Swedish standard with Filtra P
Sand filter, with Bag with Polonite,
F 5 1 2 Summer
biomodules up-flow
Sand filter, with Bag with Polonite,
G 3 1 2 Summer
biomodules up-flow
Sand filter, with Bag with Polonite,
H <1 <1 2 Year-round
biomodules up-flow
Biological fibre Bag with Polonite, approx. 6
I 3 2 2
material in a tank up-flow months/year
Biop®, biofilm
Bag with Polonite,
J treatment without 7 <1 7* Year-round
up-flow
aeration
Biop®, biofilm
Bag with Polonite,
K treatment without 7 <1 7* Year-round
up-flow
aeration
Activated sludge Bag with Polonite,
L <1 <1 2 Year-round
with aeration up-flow

22
* Facilities J and K served the same 14 users in total. The wastewater was pumped from the septic tank
to a distribution chamber where the flow was diverted into the two separate Biop ® facilities. Evenly
distributed flows to the two P-filters were assumed, thus serving 7 users each.

Sand filter with two

ventilation pipes

Sand filter effluent

Figure 3 Example of a sand filter with biomodules where the effluent was sampled prior to discharge
into the ditch.

Figure 4 Package plant Biop® with a P-filter (left) and P-filter placed downstream from a sand filter
(right).

Facilities A, D and E included sand filters built according to the Swedish standards,
consisting of 80 cm thick filter media (sand and gravel, particle sizes 2–8 mm) with a
design load of 30–60 L m−2 d−1 and a typical surface area of 25 m2 (Palm et al., 2012).
Slotted pipes distributed the wastewater, which subsequently filtered vertically through
the filter bed to the bottom of the system to be collected in drainage pipes. Facilities B
(Figure 3) and C included sand filters with plastic crates (biomodules), and were smaller
23
in size than the sand filters constructed in accordance with the Swedish standard. Sand
filters F, G and H had a layer of drainage baskets on the top (biomodules) with a 0.5 m
wide triangular cross section, where the distribution pipe was placed. All sand filters were
covered with an approximately 30 cm thick layer of soil.
The nine alkaline P-filters (D–L) consisted of bags filled with P-sorbing material (Figure
4). Filter media Polonite® (Supplier: Ecofiltration AB, Sweden) was used in eight of the
P-filters, placed in a container at the ground level and operated in down-flow (D) or up-
flow (F–L) mode. Alkaline P-filter media Filtra P (supplier: Wavin-Laboko Ltd) was used
in facility E, placed in two tanks in series installed at different levels, one percolating in
down-flow mode and one in up-flow mode. Four facilities (I–L) had alkaline P-filters
installed as a polishing step after a package plant. Facility I consisted of a trickling filter
installed inside a tank where the wastewater percolated through (4evergreen by
Biorock®). Facilities J and K consisted of a unit with multi-stage biological treatment
based on attached growth with no aeration (Biop®, Figure 4 left). Facility L was a unit
with aerated activated sludge (Ecobox Small by Ecotech AB).

3.2.2. Flow-proportional sampling

Sampling of the on-site wastewater treatment facilities was carried out between
September and October 2015 and between May and August 2016. Each facility was
sampled on at least in three occasions for 3-4 hours on each occasion. The samples were
taken at different times of the day depending on the specific flow and practicalities
regarding equipment and location. Grab samples were taken in the third chamber of the
septic tank or from the distribution/pumping well from where the wastewater was
transferred to the sand filter. Flow-proportional samples were taken as relevant from the
outlets of the sand filters, package plants and the P-filters (Figure 5). Volumetric flow
measurements were taken manually, using effluent flow from the sand filters and either
influent or effluent flow to/from the P-filters. Two composite samples of approximately
2–3 L each (depending on the flow) were taken from the effluent water collected from
the outlets of the sand filters, package plants and P-filters during each sampling event.

Figure 5 Sampling the effluent of a sand filter (left) and the influent of a P-filter (right).

24
3.2.3. Analyses of water samples

Total suspended solids (European Committee for Standardization, 2005) and pH (pH
meter WTW pH330 with a WTW SenTix41 pH electrode) were measured on-site.
Effluent samples were frozen and stored prior to analyses of total and dissolved P (tot-P,
diss-P), total and dissolved organic carbon (TOC, DOC) and biological oxygen demand
(BOD7). Samples for DOC and diss-P analyses were filtered through 0.45 μm filters
before freezing. BOD7 was analysed according to the European standard method CSN
EN 1899-1 (modified) (European Committee for Standardization, 1998). A Quattro
spectrometer and the device-specific method number A-031-04 was used to analyse the
phosphorus concentrations according to European standards (ammonium molybdate
method) with digestion (persulfate oxidation, SS-EN 1189 performance 6.4) (ISO
6878:2005, 2005). IR detection (based on CSN EN 1484, CSN EN 16192, and SM
5310) was used to analyse TOC and DOC (European Committee for Standardization,
1997, 2011).
Indicator organisms (bacteria) were used to assess the presence of faecal contamination in
the water samples. Samples for indicator bacteria analysis were stored in cooling bags and
transferred to the laboratory directly after the sampling was completed, or the day after if
the sampling was done in the late evening. The bacterial analyses were carried out in two
accredited laboratories using the Swedish standard methods SS 028167-2 (modified) for
E. coli and total coliforms, SS-EN ISO 7899-2 for intestinal enterococci and ISO/CD
14189/6461-2 for C. perfringens.

3.2.4. Data analyses

The overall P reduction in the studied facilities was calculated based on P concentrations
measured in the last chamber of the septic tank and the effluent from the final treatment
step (sand filter or P-filter). The estimated P reduction of the system did not include
treatment in the septic tank.
The concentrations of tot-P, diss-P, TOC and DOC, as well as the pH, were weighted
in accordance to the water flow. The flow-proportional arithmetic means (‫ݔ‬ҧ ‫ ) כ‬and
standard deviations (s) were calculated according to Eq. 8 and 9, respectively.
σ೙
೔సభ ௫೔ ௪೔
‫ݔ‬ҧ ‫ כ‬ൌ σ೙
(Eq. 8)
೔సభ ௪೔

σ‹ൌͳ ௪೔ ሺ௫೔ ି௫ҧ ‫ כ‬ሻమ

•ൌට (Eq. 9)
ሺσ‹ൌͳ ௪೔ ሻିଵ

Where:
‫ݔ‬ҧ ‫ = כ‬is the flow-proportional arithmetic mean,
n = is the number of observations,
xi = is the measured concentration,
25
wi = is the volume of wastewater making up the composite sample, and
s = is the flow-proportional standard deviation.
The influent flow rate to the filter was assumed to be the same as the measured effluent
flow rate, and vice versa when the influent flow rate was the one measured instead.
For the data analysis, the statistical software Minitab® 17 and Microsoft Excel were used.
Spearman’s rank-order correlation was used to analyse the nonparametric data (normality
test Anderson-Darling applied) such as the correlations between bacterial and phosphorus
concentrations, whereas Pearson correlation was used when the data was transformed and
normal distribution was attained, e.g. P removal or bacterial log removal. The difference
was considered significant if the p-values were <0.05. For bacterial concentrations, half
of the detection limit was used in the statistical analyses if the result was below the
detection limit (<10).
Geometric averages were used due to the skewed distribution of bacterial concentration
data, and were weighted for the flow according to Eq. 10.
భ
೙
‫ݔ‬ҧ ‫כ‬௚ ൌ ሺς௡௜ୀଵ ‫ݔ‬௜ ௪೔ ሻσ೔సభ ೢ೔ (Eq. 10)

Where:
xത* g = is the flow-weighted geometric mean,
n = is the number of observations, and
wi = is the volume of wastewater making up the composite sample.
The log removal value was determined by taking the logarithm of the ratio of pathogen
concentration in the influent and effluent water. Bacteriological data were also log10-
transformed for statistical analyses to obtain a normal distribution.

26
4. Results
The results are divided into two sections, with the first section (4.1) referring to the
sustainability assessment of on-site sanitation systems as investigated in Paper I and
including the findings from Papers II and III, and the second section (4.2) referring to
the relation between the removal of organic content, nutrients and indicator bacteria
from on-site wastewater treatment systems in a field study (Papers II and III).

4.1. Evaluating the sustainability of on-site sanitation systems

4.1.1. Evaluation of on-site sanitation systems in terms of robustness, risk of

pathogen discharge and P removal

Twelve sustainability indicators were defined to assess the sustainability of the nine on-
site sanitation systems as described in Paper I. The indicators used covered the main
aspects of sustainability and the ELECTRE III method was used to combine them.
The indicators robustness, risk of pathogen discharge and P removal are presented in this section
in more detail (Table 2), as they were also investigated in the field study described in
Papers II and III by sampling the influent and effluent of sand filters, package plants and
P-filters. Moreover, the indicators robustness, risk of pathogen discharge and nutrient removal
(including both N and P) were allocated the top three weights by the reference group in
the study presented in Paper I, and their assessment thus had a considerable impact on
the resulting ranking of the alternatives.
Table 2 Performance of the different on-site sanitation systems with respect to the indicators robustness,
pathogen discharge and P removal. The systems are listed in descending order based on the sustainability rank
obtained with the ELECTRE III method.

Indicators
Rank On-site sanitation system
Robustness Pathogen discharge P removal
1 S1. Greywater and blackwater separation Medium Very low 90%
2 A4. Chem. P removal + sand filter Medium Low 90%
2 A1. Sand filter High Medium 40%
3 A3. Sand filter + Polonite® P-filter Medium Low 90%
4 A2. Drain field High High 40%
5 A5. Chem. P removal + drain field Medium Medium 90%
6 S2. Urine diversion Low Medium 80%
7 P2. Package plant with chemical P removal Low Low 90%
8 P1. Package plant with Polonite® P-filter Low Low 90%

In the multi-criteria analysis described in Paper I, simpler systems such as A1 and A2 were
assessed as having high robustness, considering that these systems generally work well if
they are correctly designed and loaded (Palm et al., 2002), with clogging of the filter
material being the main risk (USEPA, 2002; Rolland et al., 2009). The addition of a
polishing step, as for systems A3–A5, resulted in an overall medium robustness because
of the increased number of risks when adding extra components and consumables that

27
require replacement, e.g. P-filter or chemicals for P removal (Hellström and Jonsson,
2003). Similarly, the system with greywater and blackwater separation (S1) was also
assessed as having medium robustness because the holding tank for blackwater does not
adapt to flow fluctuations in the same way as a septic tank with an outlet, and the
monitoring of an extra tank and the use of a vacuum toilet could increase the risk of
technical complications. The urine diversion system (S2) and the package plants (P1 and
P2) had lower robustness because of the increased complexity of the systems. The urine-
diverting toilets may present problems with ventilation or blockages of the urine-
conducting pipe (Udert et al., 2003), whereas the package plants generally had an
increased risk of failure due to the presence of e.g. moving parts, sensors or electrical
control systems (USEPA, 2002) and they are also often sensitive to operational
disturbances (Hübinette, 2009).
Results from the field study (Papers II and III) showed that 22 of the 34 inspected facilities
(65%) were not suitable for sampling. When looking at specific systems, 21 sand filters
could not be sampled, the reasons being the lack of flow in the outlet pipe (11), the
inexistence of an inspection chamber from which samples could be taken (5), the
inappropriate design of the inlet leading to the subsequent P-filter (6) and other reasons
(3) e.g. ponding water in the outlet, impossibility to remove the outlet pipe for sampling.
Moreover, 13 inspected P-filters could not be sampled, mainly due to the absence of flow
in the outlet pipe (8), the inappropriate design of the inlet which made it inaccessible (7),
and other reasons (2) e.g. impossibility to remove the inlet pipe for sampling. The lack
of flow in the outlet pipes was a clear sign of malfunctioning, e.g. the P-filters where
there was no flow indicated that the effluent water from the sand filters was infiltrating
into the soil instead of reaching the subsequent polishing step (Figure 6, left). The lack
of access to the effluent pipes did not necessarily imply that the system did not function
correctly. However, the inaccessibility to the system makes it difficult to evaluate their
performance and robustness.
Furthermore, three of eight sand filters were found to have dilution problems, indicating
that the sealing layer was not appropriately constructed and, hence the technical function
of these filters was different from that expected from a standard design. Six out of nine
P-filters showed adequate hydraulic function, whereas two filters had to cope with high
hydraulic loads and poorly treated effluent from the preceding package plants that lacked
aeration, hence decreasing the P-treatment capacity. One P-filter was 6-7 years old at the
time of sampling and showed signs of clogging. Only four package plants were sampled;
two showed good hydraulic function whereas the other two lacked aeration in the
biological treatment tanks and were likely under-dimensioned.

28
Figure 6 Dry P-filter where no effluent was possible to sample (left) and P-filter that had been in use for
approximately three years (right) and showed signs of needing to be replaced.

The robustness of the systems can thus vary between the theory, as assumed in Paper I,
and the practice as observed in the field study and reported in Papers II and III. When
systems are designed and built according to the guidelines, they are expected to function
hydraulically, but the opposite was found in the field.
In terms of nutrient removal, the P removal was highest in the source-separation
alternatives (S1 and S2) because most of the P (80%) is contained in the faeces and urine
(Jönsson et al., 2005). The alternatives with a P-removal step, either by the use of
Polonite® filter (A3 and P1) or the addition of chemicals (A4, A5, P2) had high P removal
(90%) together with the source-separation options S1 and S2, which had 90% and 80%
P removal, respectively. Systems A1 and A2 had the lowest P removal (40%) as they are
not designed to be a stand-alone treatment for long-term P removal (Eveborn et al., 2012;
Sinclair et al., 2014) but are instead intended to degrade carbonaceous material. The
results from the field study (Paper II) are in line with the assessment considered in the
multi-criteria analysis for the indicator P removal, since the average P-removal rates –
excluding the removal of P in septic tanks – for the facilities that functioned were 36% P
removal in sand filters (geometric mean; range 22–70% removal) and 90% in P-filters
(geometric mean; range 75–99% removal).
In terms of pathogens, the lowest risk of pathogen discharge to receiving waters was
attained by the greywater and blackwater separation system (S1) since the faeces, the
fraction that contains the largest pathogen load in wastewater, was stored in a holding
tank and collected and treated in a separate facility. The number of technical treatment
barriers included in each sanitation system, as discussed by e.g. Stenström (2013),
determined the assessment of the indicator risk of pathogen discharge (Table 2). Systems
with two barriers, such as the sand filters connected with Polonite® filter (A3) or chemical
P-removal (A4), and the package plants (P1 and P2) were considered as having low risk
of pathogen discharge. Alternative A1 with only sand filter (one barrier) and S2 (one
barrier for faecal fraction as urine is considered almost sterile), or the drain fields with
chemical P removal (but groundwater as receiving water) were estimated to carry a
medium risk. The risk of pathogen discharge was the highest for the drain field without
further treatment (A2) due to its single-barrier filter material and because the receiving

29
water body was the groundwater aquifer instead of surface water. The results from the
field study (Paper III) did not generally confirm a further reduction of the bacteria content
of the wastewater in P-filters, likely due to high data variability, although the removal of
all four indicator bacteria in nearly half of the investigated P-filters indicated the potential
of P-filters to also be a polishing step for pathogen reduction.

4.1.2. Weighting of the indicators by the reference group and results from
the entropy method

The reference group assigned the highest weights to the indicators robustness and risk of
pathogen discharge (Table 3). All members of the reference group, except for member D,
assigned the highest weight of 100 points to at least one of these two indicators.
Table 3 Weights given by the different members of the reference group for each indicator and arithmetic
mean for each indicator (normalized weight, W). O & M: Operation and maintenance.

Indicators Normalized
Members of the reference group weight
A B C D E F W
Robustness 75 85 100 80 100 90 15.5
Risk of pathogen discharge 100 100 70 50 100 100 15.5
Nutrient removal (N and P) 65 85 90 50 70 80 12.9
Capital cost 30 40 90 100 80 55 11.4
O & M cost 35 40 90 80 80 60 11.1
Potential for nutrient
55 60 50 40 80 75 10.6
recycling
Social acceptance 50 80 60 80 30 50 10.4
Cumulative energy demand 20 30 70 25 10 40 5.5
Global warming potential 25 20 70 10 30 20 4.9
Energy recovery 10 10 30 10 10 10 2.2
Total weights 100

The entropy method, which reflects the power of discrimination of an indicator without
the intervention of stakeholders, showed different results. The indicators related to the
potential for nutrient recycling, both N and P, were the most discriminating whereas the
indicator capital cost had the lowest dimensionless dispersion parameter (Table 4).

30
Table 4 Dimensionless dispersion parameter (wi) and measure parameter (mi) calculated using the entropy
method. Indicators removed during the first and second round are marked as (*) and (**) respectively.

Indicators Dimensionless dispersion Measure parameter

parameter (wi) (mi)
Robustness 0.036 0.56
Risk of pathogen discharge 0.039 0.61
Nutrient removal (N) 0.051 0.33
Nutrient removal (P) 0.019 0.12**
Capital cost 0.006 0.06*. **
O & M cost 0.04 0.44
Potential for nutrient recycling (N) 0.55 2.85
Potential for nutrient recycling (P) 0.17 0.88
Social Acceptance 0.015 0.15**
Cumulative energy demand 0.031 0.88
Global Warming Potential 0.028 0.17**
Energy recovery 0.020 0.04*. **

In a first round, the indicators energy recovery and capital cost were removed from the list of
indicators as they had the lowest value of mj (Table 4), which is the product of the
dispersion parameter (wj) and the normalized mean weight given by the reference group
(Table 3). In a second round, the indicators P removal, capital cost and energy recovery were
also removed from the analysis (the five least influential indicators). The final ranking of
the alternatives was recalculated by the ELECTRE III method, resulting in the following
changes:
Ranking baseline: S1 > A1, A4 > A3 > A2 > A5 > S2 > P2 > P1
Ranking removing 2 indicators (*): S1> A3, A4 > A1 > A2 > A5 > S2 > P1, P2
Ranking removing five indicators (**): S1 > A3, A4 > A1 > A2, S2 > P1, P2 > A5
The ranking position of the alternatives varied little between the first and second
positions. The alternative with greywater and blackwater separation (S1) remained in the
first position and the system with Polonite® P-filter material (A3) outranked the
standalone sand filter option (A1). When the five indicators were removed, the ranking
of the package plants (P1, P2) improved to the detriment of A5 (chemical P removal and
drain field). The results showed that the indicators removed from the calculations (Table
4) had little impact on the general ranking of alternatives. This does not, however, imply
that they could be eliminated from a sustainability analysis in general, as their values and
weights may vary depending on the inventory data and priorities of the stakeholders. The
analysis does lead to a better understanding of the indicators, in terms of their relevance
and usefulness but also with regard to their assessment and the possibility to simplify the
study by re-considering certain criteria.
4.1.3. Scenario analysis

The weights allocated to the different indicators by the reference group were modified
in a scenario analysis in an attempt to evaluate different plausible settings of interest based
on socio-demographic factors such as: areas where the removal of nutrients is of less
importance (Scenario 1); areas with sensitive receiving waters and presence of farmland

31
(Scenario 2); a political strategy change where importance is given to energy recovery
and climate change mitigation (Scenario 3).
The results of the scenario analysis showed that changes in the priorities of decision-
makers, by giving more or less importance to certain indicators, can affect the
sustainability ranking of the alternatives considered (Table 5).
Table 5 Ranking of alternatives for the different scenarios

Description Ranking of alternatives

Scenario 0 With original weights from reference group S1 > A1, A4 > A3 > A2 > A5 > S2 > P2 >
P1
Scenario 1 Lowest importance to nutrients-related A1, A4 > A2, S1 > A5 > A3 > S2 > P2 > P1
indicators (e.g. northern Sweden)

Scenario 2 Highest importance to nutrients-related S1 > S2 > A1, A2, A4 > A5 > A3, P2 > P1
indicators (e.g. areas with farmland)

Scenario 3 Highest importance to energy recovery, CED A4, A5 > A1 > A3, P2 > A2 > S1 > S2 > P1
and GWP (e.g. change in political strategy)

In this regard, the greywater-blackwater separation option (S1) dropped in the ranking
when the removal and recycling of nutrients was considered of less importance. In such
contexts, a simple sand filter (A1) or a sand filter in combination with chemical P removal
(A4) resulted in a more sustainable alternative. However, in areas where nutrient removal
is important both for eutrophication reasons and the possibility to recycle them to
farmland, the source separating options S1 and S2 outranked the other alternatives. When
the highest importance was given to the indicators related to energy use and recovery,
and climate change, the alternatives with chemical precipitation improved their position
(e.g. P2) and led the ranking (e.g. A5), due to their high energy recovery and moderate
GWP and CED.

4.2. Removal of organic content, P and bacteria from sand filters and P-
filters

Twelve facilities (A – L) were evaluated in terms of removal and discharge of organic

content, measured as TOC, total and dissolved P (Paper II), and indicator bacteria,
namely E. coli, total coliforms, intestinal enterococci and C. perfringens (Paper III). A
summary of the measured effluent concentrations is found in Table 6.

32
 Table 6 Flow-weighted arithmetic means of TOC, tot-P and diss-P, and flow-weighted geometric means and ranges (minimum-maximum values in brackets)
of E. coli, total coliforms, intestinal enterococci, C. perfringens, and pH (without ranges) measured in the effluents of the biological treatment units (Bio-treat) and
P-filters of the investigated facilities (A–L). When the concentrations of bacteria were under the detection limit (<10 CFU per 100 mL), half of the detection
limit was used to calculate the flow-weighted geometric mean.

pH TOC tot-P diss-P E. coli Total coliforms Intestinal enterococci C. perfringens

[] mg L-1 mg L-1 mg L-1 CFU per 100 mL CFU per 100 mL CFU per 100 mL CFU per 100 mL
Bio- Bio- Bio- Bio-
P-filter P-filter P-filter P-filter Bio-treat P-filter Bio-treat P-filter Bio-treat P-filter Bio-treat P-filter
treat treat treat treat
25 243 87 60
A 4.5 9.7 0.9 0.9
(<10-760) (60-4,200) (<10-2,900) (<10-580)
36528
94 100,000 221
B 6.9 160 8.4 7.7 (25,000-
(<10-1,500) (100,000) (30-750)
59,000)
5 7 7 15
C 4.8 9.3 1.8 0.5
(<10) (<10-330) (<10-2,500) (<10-30)
25 7 82 9 8 8 7 6
D 6.2 9.4 8.2 8.6 1.8 0.2 1.7 0.2
(<10-80) (<10-20) (60-110) (<10-50) (<10-30)) (<10-70) (<10-10) (<10-20)
5 8 5 20 35 11 19 28
E 4.2 6.0 13.2 11.9 1.6 1.5 0.8 0.8
(<10) (<10-91) (<10) (<10-720) (<10-1,000) (<10-70) (10-30) (<10-80)

33
21 5 115 5 65 5 6 5
F 7.4 12.1 19.8 12.7 14.8 0.2 14.8 0.1
(<10-400) (<10) (<10-500) (<10) (<10-120)) (<10) (<10-10) (<10)
14 255 46 77
178 30 1159 9
G 7.2 10.5 12.0 11.0 7.5 1.9 6.8 0.2 (<10- (<10- (<10- (<10-
(<10-7,500) (<10-170) (280-2,100) (<10-130)
500) 5,000) 630) 810)
14 6868 27 12
1192 81 272 37
H 6.8 9.9 20.1 14.6 10.6 0.15 11.7 0.03 (<10- (1,250- (<10- (<10-
(90-10,000) (<10-1,700) (40-550) (10-210)
460) 50,000) 120) 100)
36,979 6,291 66,712 68,726 3,990 3,687 1,375 4,011
I 7.2 9.6 39.8 31.3 10.1 0.9 9.9 0.3 (30,000- (2,400- (35,000- (10,000- (1,540- (2,500- (100- (200-
80,000) 12,000) 240,000) 1,090,000) 17,800) 5,300) 20,000) 90,000)
18,463 5,503 64,946 43,884 1,368 53,316 36,858
1,379
J 7.3 8.6 25.3 22.4 7.8 4.4 7.3 3.3 (63,000- (2,100- (20,000- (8,000- (100- (7200- (3,400-
(100-4,600)
48,000) 14,700) 127,000) 105,000) 7000) 183,000) 110,000)
22,581 2,486 85,003 31,116 3,775 74,743 13,815
1,679
K 7.5 8.8 22.0 19.0 7.5 3.0 7.4 2.7 (2,500- (1,000- (40,000- (6,200- (2,400- (6,000- (1,000-
(120-3,500)
605,000) 6,000) 200,000) 92,000) 5,300) 228,000) 45,000)
71 10,271 3,812 160
1,209 228 257 1,471
L 8 9.7 29.3 26.6 11.6 0.9 11.5 0.6 (<10- (3,000- (600- (<10-
(50-6,100) (50-700) (50-380) (300-7,200)
57,000) 27,000) 800,000) 42,000)
4.2.1 Performance of sand filters

The eight investigated sand filters (A–H) showed varying results in terms of P removal,
and the bacterial concentrations in the effluents were considerable. The average tot-P
concentrations in the effluent of four (B, F, G and H) of the eight investigated sand filters
ranged between 7.5 and 14.8 mg L-1 (Table 6), thus exceeding the Swedish
recommended concentration value of 3 mg L-1 for common receiving waters (Swedish
EPA, 2006). Of these sand filters, three (B, G and H) exceeded the criteria set for
excellent water quality by the EU bathing water directive with regard to intestinal
enterococci and one sand filter (H) with regard to E. coli. The criteria set for excellent
(coastal) bathing water quality is 100 CFU 100 mL−1 for intestinal enterococci and 250
CFU/100 mL for E. coli (EU, 2006). For example, sand filter B was only three years old
at the time of sampling, but it performed poorly in terms of P reduction (the
concentration of tot-P in the effluent was 8.4 mg L-1) and bacteria removal (the
concentration of intestinal enterococci was >100,000 CFU per 100 mL) (Table 6). The
effluent TOC (160 mg L−1) and the content of suspended solids (17mg L−1) were also
high, and the dark colour and strong odour of the wastewater indicated that this filter
was clogged. Sand filter H showed a good overall P reduction capacity (62%) although it
proved to be inefficient for reaching low effluent concentrations as it discharged 10.6 mg
L−1 tot-P, mainly due to the very high initial concentration of tot-P in the septic tank
(29.3 mg L−1). Thus, high tot-P reduction in a sand filter does not necessarily imply an
acceptable concentration of P in the effluent. Moreover, the effluent TOC
concentrations in sand filter H were at an acceptable level (20.1 mg L-1) and the
wastewater was clear although the bacterial concentrations were high (1192 CFU of E.
coli per 100 mL), showing that indicator bacteria can be found in high numbers even
though no obvious malfunctioning of the sand filter is observed.
Only one sand filter (D) was confirmed to remove P satisfactorily (70% removal) with an
effluent concentration of 1.8 mg tot-P L−1 and very low concentrations of all indicator
bacteria (Table 6). Sand filter D was relatively new and had only been in use for 1-2 years
at the time of sampling. However, its P-removal capacity is expected to decrease over
the years. Based on field observations, at least three (A, C, E) of the eight investigated
sand filters had likely dilution problems, i.e. extraneous soil, ground or rainwater
infiltrated into the filter bed and diluted the wastewater. The field observations included
high measured water flow through the sand filters even when the users were not at home
and when the pump transporting wastewater from the septic tank was not operating
during collection, as well as low effluent concentrations of TOC (below 13.2 mg L-1)
and P (below 1.8 mg L-1 tot-P) (Table 6).
No significant correlations between the effluent concentrations of P and indicator
bacteria in the sand filters were found. However, the effluents concentrations of tot-P
and dissolved-P as well as the effluents concentrations of all indicator bacteria except E.
coli, correlated positively with the concentrations of TOC and DOC in the effluents
(Spearman correlation ≥ 0.51, p < 0.009). According to Song et al. (2006), the
precipitation rate and P-removal efficiency are reduced by the presence of humic

34
substances at lower pH (e.g. pH 8) but almost no effect is seen at higher pH (≥ 9), due
to the consumption of Ca for the precipitation of the phosphates. Hence, the high TOC
concentration in the effluent of the sand filters and the low pH (below 7.4, Table 6)
could indicate that the precipitation rate of calcium phosphate and consequent P removal
in the sand filter was inhibited to some extent. Moreover, high effluent concentrations
of TOC and DOC indicate low degradation of organic matter in the filter, which may
have resulted in low bacterial removal. The high concentration of organic carbon likely
affected the capacity of the bacteria to be retained in the filters, due to competition for
adsorption sites (Sélas et al., 2003).

4.2.2 Performance of alkaline phosphorus filters

Six out of the nine investigated P-filters (D–L) removed P efficiently (75–99% P
reduction) and accounted for a considerable amount of tot-P reduced in the treatment
facilities. In contrast, the concentrations of indicator bacteria in the effluent of the P-
filters differed considerably between the filters and were in some cases high, especially in
the P-filters of facilities I–L, which consisted of package plants instead of sand filters.
Three (I, J and K) of the nine investigated P-filters exceeded excellent bathing water
quality in coastal waters with regard to E. coli and four (I, J, K and L) with regard to
intestinal enterococci. The highest effluent concentrations of tot-P, namely 4.4 and 3 mg
L-1, were found in P-filters J and K respectively (Table 6), which were highly loaded
(J=79.4 L m−2 h−1; K=58.5 L m−2 h−1) because of the number of users they provided
service to as compared to the rest of the facilities. P-filter effluent concentrations of
indicator bacteria showed moderate positive correlations with effluent concentrations of
P and organic matter, whereas the effluent pH showed moderate negative correlations
with total coliforms and C. perfringens (Table 7).
Table 7 Spearman rank correlations between the concentrations of indicator bacteria (CFU per 100 ml)
and physico-chemical parameters such as total phosphorus (tot-P), dissolved phosphorus (diss-P), total
organic content (TOC), dissolved organic carbon (DOC) and pH measured in the effluent of the P-
filters. Significances of p < 0.001 and 0.01 are marked as*** and **, respectively (n = 49–66).
E. coli Total Intestinal C. perfringens Tot-P Diss-P
coliform enterococci
E. coli - - - - - -
Total coliforms 0.85*** - - - - -
Int. enterococci 0.78*** 0.82*** - - - -
C. perfringens 0.83*** 0.79*** 0.72*** - - -
Tot-P 0.52*** 0.54*** 0.51*** 0.69*** - -
Diss-P 0.63*** 0.67*** 0.56*** 0.71*** 0.84*** -
TOC 0.68*** 0.68*** 0.65*** 0.49*** 0.39** 0.45***
DOC 0.72*** 0.70*** 0.67*** 0.52*** 0.32** 0.46***
pH -0.18 -0.47** -0.12 -0.52*** -0.43*** -0.59***

Furthermore, Pearson correlations between the average log removal of indicator bacteria
and the percentage removal of tot-P, diss-P, TOC and pH were conducted. Only the
removal of total coliforms (Figure 7) showed strong positive correlations with the pH as
well as the removal of tot-P (Pearson correlation = 0.74, p = 0.02 for both cases).

35
 2

Log10 removal of total coliforms

0
0 20 40 60 80 100

-1
Reduction of tot-P (%)

Figure 7 Relationship between the average reduction of tot-P and the average log10 removal of total
coliforms measured in the effluent of the nine P-filters studied (D–L).

The pH in many of the P-filters was possibly not high enough to support removal of
bacteria other than coliforms, and the characteristic flow-saturation of the P-filters might
have decreased the ability of the bacteria to attach to the filter particles, as discussed by
Cooper et al. (2016) who observed that increased moisture content likely reduced
bacterial attachment to soil. Nevertheless, the P-filters D, F, H and J had positive removal
of all four bacteria types, indicating a potential of the P-filters to serve as a treatment step,
not only with respect to P but also for bacteria. Possibly, the number of P-filters
investigated in this study was not large enough to prove their efficiency for bacteria
removal. Moreover, the significant positive correlation between the average P reduction
in the P-filters and the effluent pH suggests that pH could be used as an indicator of P-
filter efficacy to remove P. For example, the three P-filters that worked insufficiently,
with effluent concentrations higher than 3 mg L−1 (J, K) or 0.7% tot-P reduction (E), had
an effluent pH of 8.8 or lower (Table 6).
Paired t-tests showed no significant (α = 0.05) difference between the average influent
and effluent concentrations of the different indicator bacteria measured in facilities D–L,
possibly because of the high data variability. Consequently, the collected data did not
generally confirm a further reduction of the bacterial content of the wastewater in the P-
filters, despite their high pH (Table 6) and despite the strong positive correlation between
the average log removal of total coliform and effluent pH. Similar studies have reached
similar conclusions and were unable to confirm that the inclusion of a separate
phosphorus removal unit enhanced the removal of pathogens (Martikainen et al., 2018).
Generally, and in line with previous studies (e.g. Kauppinen et al. 2014), the P-filters that
showed good removal of P also removed indicator bacteria satisfactorily, with some
minor exceptions, e.g. facilities G and K where negative log removals of C. perfringens
and intestinal enterococci were observed.

36
5. Discussion
The discussion chapter is divided in four sections, which discuss different aspects
concerning the sustainability and function of on-site sanitation systems, specifically the
robustness of the systems and their function in terms of P and microorganism removal.
A discussion of the methodological approach used in the thesis, as well as the implications
in the Swedish context are included.
The sustainability of on-site sanitation systems can be determined by their performance
in terms of environmental, economic, socio-cultural, technical and health-related
indicators that are relevant to the local context and stakeholders, as discussed in Paper I.
To be sustainable, a solution must, for example, function hydraulically and fulfill the
environmental and health regulations, must be affordable and socially accepted.
Moreover, as suggested by Balkema et al. (2002), a sustainable system means limited use
and degradation of resources in terms of emissions, and also avoiding export problems in
time and space. Apart from the selection of indicators, which is necessary for a
comprehensive evaluation of any systems, the weighting of the indicators is essential for
the sustainability assessment (Balkema et al., 2002), although establishing the relative
importance of each indicator in the form of weights is sometimes perceived by some
analysists as adding subjectivity (Rowley et al., 2012). However, by not introducing
weights that represent the stakeholders’ priorities, the analyst must assume that each
indicator has equal importance, which is indeed an implicit, arbitrary non-transparent
valuation, and seldom recommended (Rowley et al., 2012).

5.1. Robustness and function of on-site sanitation systems

The reference group included in the multi-criteria analysis presented in Paper I allocated
the highest weights to indicators robustness, risk of pathogen discharge and nutrients removal.
These three indicators are further discussed in this thesis because of their importance and
implications in terms of eutrophication and public health concerns, and the repercussions
that the system’s robustness has on the capacity of the system to treat nutrients and
pathogenic microorganisms.
The indicator robustness was defined based on previous descriptions, e.g. Spiller et al.
(2015) and Balkema et al. (2002), and was assessed for each sanitation solution based on
earlier studies reporting monitoring and performance of on-site systems, e.g. Hellström
and Jonsson, 2003; Hübinette, 2009; Jenssen et al., 2010; Lehtoranta et al., 2014; Palm
et al., 2002. Robustness was defined as the ability of the system to deal with sources of
variability, within a range of conditions, while maintaining the functionality and an
optimal level of wastewater treatment (Paper I). In the multi-criteria analysis, simpler
systems like sand filters or drain fields were determined to be more robust than more
advanced technical solutions such as package plants, as the latter require more intensive
monitoring and maintenance than passive systems and are more exposed to process
sensitivities such as hydraulic loading. Consequently, the assessment of the indicator

37
robustness and the high weight allocated to it by the reference group contributed to the
higher ranking of soil-based systems compared to e.g. package plants.
However, the robustness of on-site sanitation systems goes beyond its theoretical
definition, and there is a gap between what is expected from the systems when correctly
designed, constructed and monitored, and what is observed and measured in existing
systems in terms of performance. Moreover, existing on-site systems often do not allow
sampling because of inappropriate design, e.g. lack of flow or inaccessibility to the effluent
pipe, which also has consequences in their robustness as they cannot be easily accessed
and monitored to detect malfunctions. Additionally, some soil-based systems can function
hydraulically in a long-term perspective, even though they do not remove phosphorus
(e.g. sand filters F and G, Paper II) or bacteria (e.g. sand filter H, Paper III) satisfactorily.
The large variations in the quality of the effluents as well as the insufficient P removal in
sand filters reported in previous field studies (Eveborn et al., 2012; Kauppinen et al., 2014;
Martikainen et al., 2018) was confirmed in Papers II and III, making downstream P
treatment advisable. The presence of intestinal enterococci and E. coli bacteria in sand
filter effluents in numbers exceeding the limit for the EU bathing water directive was
confirmed in Paper III and reported by similar studies (Kauppinen et al., 2014;
Martikainen et al., 2018). The removal of viruses was not studied in this thesis, but
previous studies have reported large variations and presence in the effluent (Martikainen
et al., 2018), and dependence on climatic conditions e.g. lower removal during cold
periods (Kauppinen et al., 2014). High pH (above 10) can have a significant effect on
virus inactivation even under cold conditions, as studied in compact filter systems
(Heistad et al., 2006). In order to assess the health risks associated to effluent discharge
from on-site sanitation systems and overall performance, direct monitoring of pathogens
at each facility has been suggested (Kauppinen et al., 2014). In terms of indicator bacteria
monitoring, flow-proportional sampling and sample handling can prove challenging, as
discussed in Paper III, especially at cold temperatures e.g. E. coli effluent concentrations
in sand filters decreased during a sampling event carried out at sub-zero temperatures.
Varying results on the effect of temperature on the removal of nutrients has been
reported, from no significant effect in the removal of nutrients in on-site sanitation
systems (Christopherson et al., 2005; Kauppinen et al., 2014) to clear effects in, for
example, the retention of P in alkaline P-filters (Herrmann et al., 2014), and soil (N. J.
Barrow, 1975). Besides, the removal of pathogenic microorganisms could be affected and
generally reduced during winter (Axler et al., 2005; Kauppinen et al., 2014).
Published reports covering the monitoring and inspection of on-site sanitation systems
endorse the idea that improving the robustness of on-site sanitation systems requires more
focus on the construction, operation and maintenance of the systems, and not just on
control. For example, Larsson et al. (2017) conducted a monitoring campaign of 40
different on-site systems for wastewater treatment in a municipality in Sweden. The visual
basic inspection of 40 facilities showed that 14 systems (1 sand filter, 8 drain fields and 5
package plants) had problems with sludge in the distribution chambers or ventilation
pipes, the systems were incorrectly installed or there were technical problems with the
phosphorus precipitation. The remaining 26 systems were visually assessed to be working
38
hydraulically well. Moreover, a more exhaustive second assessment that looked at
accessibility to important parts for inspection and sampling, showed that 22 of the 40
facilities did not meet the requirements for an appropriate full inspection (e.g. missing or
incorrectly placed pipes) (Larsson et al., 2017). A review about on-site sanitation systems
performance in Western Australia concluded that most failures of the systems are usually
associated to inappropriate design and construction, or operation and maintenance, rather
than inherent technological flaws (Gunady et al., 2015). Another study that evaluated the
performance of 21 different models of package plants used in Sweden showed that the
function of these plants was often inadequate and few of them treated the wastewater to
the same extent as described by the suppliers (Hübinette, 2009). Technical failures and
e.g. missing or incorrect dosing of chemicals were attributed to the lack of supervision
and maintenance, which resulted in deficient wastewater treatment. Four models of
package plants were able to achieve the requirements for nutrient discharge and excellent
water bathing quality (Hübinette, 2009).
These results are in line with the inspection of facilities to identify suitable systems for
sampling, carried out prior to the field study presented in this thesis (Papers II and III).
The conclusions extracted from the above-mentioned reports confirm that a follow-up
visit should be made directly after the system has been installed and put into operation,
to detect any problem related to the construction and function. Sampling points, such as
an inspection chamber, for instance, should be included in the system and accessible for
regular check-ups. Furthermore, supervision methodology needs to be developed in a
technology-oriented approach specifying, for example, the sampling points for each type
of system, so that regular inspections can be carried out uniformly. Moreover, assessment
criteria must also be described in order to be able to assess the function of the systems in
a standardized way. From these assessments, conclusions could be drawn about the
robustness of different systems.
In summary, robustness is an essential indicator when measuring the sustainability of
sanitation systems (Spiller et al., 2015), as concluded from its definition and assessment,
and discussed with the reference group who assigned it the highest weight (Paper I).
Robust systems must function hydraulically but must also perform in terms of treatment,
e.g. removal of nutrients and pathogens, and be in compliance with legislation and take
into account local conditions and requirements. Besides a robust design, maintenance and
control are needed to achieve lasting adequate performance of the systems.

39
5.2. Multi-criteria analysis methodology

Assessing the sustainability of on-site sanitation systems requires dealing with large
amounts of complex information and a broad range of criteria (Bradley et al., 2002).
Multi-criteria analysis appears as an effective approach to comprehensively and
transparently combine indicators and weights representing relative importance. The
ELECTRE III method in particular, a non-compensatory method in which all the
indicators are considered non-exchangeable capital assets, allows for a strong sustainability
perspective. Additionally, the possibility to use both qualitative and quantitative
indicators made the use of ELECTRE III suitable for the study presented in Paper I,
assuring the multidimensional character of the sustainability assessment (Balkema et al.,
2002). The use of thresholds by the ELECTRE III method makes it convenient when
dealing with data uncertainty and variation that are inherent in most assessments.
However, the use of thresholds is not exempt of implications due to the subjectivity
introduced by the analyst during their definition (Rogers and Bruen, 1998). According
to Rogers and Bruen (1998), the definition of thresholds must be specified by realistic
limits, in a rational and defendable manner, and explicitly estimated rather than arbitrarily
chosen. The thresholds used in Paper I mainly corresponded to the data uncertainty for
quantitative indicators, and to a one-point change in the scale for qualitative indicators
(e.g. medium robustness was preferred over low robustness). Their values were tested in
a sensitivity analysis in order to see if changes in their values affected the final ranking
significantly, as suggested by Roy (1990). Despite the intrinsic introduction of
subjectivity, the use of thresholds allowed the incorporation of the uncertainties inherent
in the indicators’ valuation into the assessment, which would otherwise not have been
accounted for.
The selection of indicators in a multi-criteria analysis is crucial, considering that indicators
do not always accurately reflect the performance of complex systems (Bertrand–Krajewski
et al., 2002). The entropy analysis presented in section 4.1.2 showed that certain
indicators had little influence in the final ranking of alternatives, e.g. energy recovery and
capital cost, possibly because of their low capacity to discriminate between the alternatives.
As discussed by Barraud et al. (2004), some indicators may be non-discriminatory
considering that most of the alternatives performed similarly, either because they were
implicitly taken into account during the selection of alternatives or because of the
difficulty of their assessment, in which case most of the alternatives have been assessed,
for precautionary reasons, to be the same. Simplifying the multi-criteria analysis by
removing indicators is often desired but not always recommended since certain
dimensions of the study can be lost, and the indicators’ values or weights will vary
depending on the data availability and stakeholders’ preferences if the study is further
developed (Barraud et al., 2004). The entropy analysis provides a better understanding of
certain indicators, not only of their relevance and usefulness but also of how they can be
assessed. Based on that information, further measures could be taken, such as, for
example, merging several indicators into one. The indicator energy recovery had the lowest
entropy and could be merged together with the indicator cumulative energy demand,

40
because they both deal with the energy used and (potentially) produced by the systems.
The study would likely benefit from the action, because the energy recovery was
previously assessed qualitatively and had only little discriminatory power. The merged-
indicator could better represent the net balance of energy for each system. The indicator
capital cost also had low entropy, possibly because the relative differences in terms of
investment cost among the different alternatives did not differ to a great extent. The
capital cost could also be merged with the second economic indicator, O & M cost, as
seen in similar studies e.g. Schoen et al. (2017). However, removing the indicator capital
cost from the analysis based on its low discriminatory power would not be recommended
as economic indicators are often considered decisive when selecting a specific technology
(Balkema et al., 2002).
The scenario analysis carried out in this study (section 4.1.3) showed that the
prioritization of different indicators depending on a given context could influence the
ranking of alternatives. Scenarios 1 and 2 intended to represent different reasonable
situations with regard to the protection status of the receiving water, which may have a
normal or high protection level, and the potential to reuse the nutrients in agriculture if
there are crops being cultivated nearby. Thus, sand filters or the addition of chemical P
removal resulted in more sustainable solutions in areas without eutrophication problems
in the receiving waters and little presence of farmland. However, in areas where nutrient
removal was important, the source-separating options S1 and S2 outranked the other
alternatives. These results show that the sustainability of specific systems depends to a
large extent on the local settings and priorities, and the treatment facility suitable in one
context may not be sustainable in another. In Scenario 3, where indicators related to
energy (energy recovery, CED) and climate change (GWP) were given the highest
weights, the systems with chemical precipitation were favoured. The greywater-
blackwater alternative lowered its position in the ranking, partly because of the assumed
lower potential to recycle energy (only biogas production from sludge from greywater).
These results highlight the potential for optimization of the greywater-blackwater system
with e.g. anaerobic digestion to recover biogas, which would improve the performance
in the indicator energy recovery. If the design of the system boundaries of the multi-criteria
analysis would have included the replacement of mineral fertilizers with sanitized
blackwater, the source-separating systems would likely have achieved a higher ranking in
Scenario 3. However, the addition of an extended system covering the agricultural
application of wastewater by-products was not considered in the study, to avoid adding
more complexity to the study and because the focus was placed in the treatment function
of the systems.

41
5.3. Removal of phosphorus and indicator bacteria in filters for on-site
wastewater treatment systems

The investigated sand filters had generally low P-removal capacity (Paper II) and often
exceeded the criteria set for excellent water quality by the EU bathing water directive
with regard to intestinal enterococci and E. coli (Paper III) despite the relatively short
time the systems had been in use (between <1 and 6-7 years).
Only one sand filter (D) of four was confirmed to remove P satisfactorily with a tot-P
effluent concentration below 3 mg L−1. Four sand filters (B, F, G and H) had effluent
concentrations of tot-P above 3 mg L-1, although sand filter H did reduce the influent P
to a great extent (70%). Their inefficiency indicates that a downstream treatment step is
needed in order to meet the Swedish guidelines of P discharge concentrations (Swedish
EPA, 2006).
There are several reasons explaining the low P-removal capacity of the sand filters.
Clogging, as was the case in sand filter B, is usually caused by accumulation of suspended
solids or filamentous particles such as from toilet paper, or precipitation and deposition
of compounds such as calcium carbonate, as well as an excess of extracellular bacterial
slimes in the sand pores (Kristiansen, 1981; Blazejewski and Murat-Blazejewska, 1997).
The reasons for the poor performance of sand filters F and G were not as apparent, and
their inefficiency could be generally related to the construction of the sand filter; the
inadequacy of the chemical composition of the sand used (e.g. a lack of Al or Fe
compounds) or the uneven distribution of the wastewater through the filter (e.g. creation
of preferential flows) are plausible explanations (Reneau et al., 1989; Gold et al., 1992;
Gill et al., 2009; Eveborn et al., 2012). Additionally, the particle size distribution and the
level of compaction of the sand also affects the efficiency of the sand filters in terms of
bacterial removal (Rolland et al., 2009). Bacterial adsorption in the filters, which is
dependent on the adsorption capacity of the sand (Hijnen et al., 2004), could also be
affected and determined during the construction phase of the sand filters. An
underdeveloped microbial community in the top layer of the filter (schmutzdecke) can
hinder the removal of indicator bacteria (Hijnen et al., 2004). However, considering the
age of the filters (at least a year old by the time of sampling), the schmutzdecke of the
investigated sand filters were likely fully developed and thus other factors likely
dominated the removal processes.
As three sand filters (A, C and E) were suspected to have dilution problems based on field
observations, the reduction of P was not explained by the treatment capacity of the
systems. These sand filters were both old (A and E, 6-7 years in use) and newly built (C,
<1 year in use). Correct installation and proper sealing of the filter must be ensured to
prevent mixing wastewater with other water sources in order to achieve adequate P
reduction within the treatment unit and to avoid further spreading of nutrients and
pathogens into the surrounding environment.
Six out of nine of the investigated P-filters removed P to a high extent (75–99% P
reduction) and accounted for a considerable amount of tot-P reduced in the treatment

42
facilities. Of the three remaining P-filters, one was old and clogged (P-filter E, with
Filtra-P media), and two others (J, K) had to cope with high hydraulic loads and with
influent wastewater that was treated to a lesser extent in terms of particles and
concentrations of organic compounds in the preceding package plant (Table 6).
The four P-filters not fulfilling the EU bathing water directive with respect to intestinal
enterococci (I, J, K and L) were placed in package plants with shorter residence times
than the other P-filters, which possibly affected the bacterial reduction (Sélas et al., 2003).
The removal of P was correlated positively with the pH measured in the effluent
(Spearman rank correlation coefficient=0.833, p=.005), indicating that pH could be used
as an indicator of P-filter efficiency. However, and despite their high pH (Table 6), the
results did not generally confirm a further reduction of the bacterial content of the
wastewater in the P-filters, likely because of the flow-saturation and lesser possibilities for
attachment for the bacteria (Cooper et al., 2016). Possibly, the number of P-filters
investigated in this study was not large enough to prove their bacterial reduction
efficiency.
Nevertheless, effluent concentrations of indicator bacteria showed moderate positive
correlations with effluent concentrations of P and organic matter (Table 7). Moreover,
the log removal of total coliforms specifically showed positive correlations with the
removal of tot-P (Figure 7) and the pH (Pearson correlation = 0.74, p = 0.02 for both
cases). The positive log removal of all bacteria types in P-filters D, F, H and J indicated
a potential for the P-filters to also serve as a polishing step for bacteria.

5.4. Implications of the results in the Swedish context

The results from the field study presented in Papers II and III, as well as other monitoring
reports (Hübinette, 2009; Larsson et al., 2017) showed that a considerable number of
existing on-site sanitation facilities (about 65%) are not suitable for sampling and
evidence-based assessment criteria for control need to be selected. Moreover, the
identification of on-site wastewater treatment facilities, including contacting the
operators and sampling the effluent flow-proportionally, proved to be laborious and time
consuming. Good planning and protocol are therefore indispensable, as discussed in
similar studies, e.g. Vilpas and Santala (2007). Previous studies have discussed that the
monitoring costs are much lower than the costs for mitigation measures and, for this
reason, accurate estimates of nutrients and pathogens concentrations could be prioritized
(Audet et al., 2014).
In Germany and Switzerland for example, on-site facilities need to be sampled once a
year, usually by certified companies that are hired by the operators and that report the
results to the corresponding authorities (Aicher, 2018; Schneider, 2018). However,
regular check-ups and effluent sampling are currently not required in Sweden, and a
recent governmental investigation about on-site treatment systems up to 200 PE does not
include sampling (Swedish Ministry of Environment and Energy, 2018).

43
Visual inspections are generally enough to detect obvious problems, such as misconnected
components or clogged filters, but sometimes the malfunctioning of the systems is not
obvious. In Paper II it was reported that high tot-P reduction in a sand filters (e.g. sand
filter H) did not necessarily translate into an acceptable tot-P concentration in the
effluent. Similarly, bacterial concentrations in the effluent of sand filters were high
although no obvious malfunctioning was observed (Paper III). Recent studies have
reported similar results, concluding that certain facilities can still present a risk of microbial
pollution to surrounding waters although they are able to reduce the pathogen load
(Martikainen et al., 2018). These results also indicate that visual inspection may not be
sufficient to assess the function and overall performance of on-site sanitation facilities in
terms of nutrients and pathogen discharge. A recent study suggests that the status of sand
filters can be more accurately assessed using camera inspection of distribution pipes rather
than mere visual inspection of the end of the pipes using a torch (Hedin, 2018).
Sampling of indicator bacteria such as E. coli and sample handling during sub-zero
temperatures can prove challenging, since, as discussed in Paper III, E. coli concentrations
can decrease at low temperatures. Guidelines and recommendations for acceptable levels
of indicator microorganisms may be needed for an adequate assessment of the
performance of on-site wastewater treatment systems.
Lastly, as errors in design or construction of on-site facilities are often responsible for their
malfunctioning, follow-up visits made directly after the systems have been installed and
put into operation should be prioritized to detect any problem. Accessible sampling
points should be ensured for regular check-ups and inspections, if these activities are
required. Assessment criteria and instructions should be developed to ease the control of
on-site sanitation facilities.

44
6. Conclusions
Twelve indicators were defined to assess the sustainability of nine on-site sanitation
systems and were given weights by a reference group, which considered that indicators
robustness and risk of pathogen discharge were the most important (Paper I). Assessing the
robustness of the systems proved challenging because of the large variations that on-site
sanitation systems can present in terms of performance, the barriers encountered to
inspect them in the field (Papers II and III), and the existing gap between how they work
in theory and in practice. The entropy study showed how some indicators had less
discriminatory power than others (indicators energy recovery and capital cost) and their
removal had almost no effects on the alternatives ranking. Therefore, the indicator energy
recovery could be merged into the indicator cumulative energy demand to simplify the analysis
in further studies.
A scenarios analysis showed that distinct strategies giving importance to different
indicators resulted in different rankings, indicating that systems’ sustainability -which
system is most sustainable where- will depend, to some extent, on the existing regulations
and the priorities of the decision-makers. The scenario analysis showed that in areas
without eutrophication problems in the receiving waters and little presence of farmland
(Scenario 1), simpler systems like sand filters with or without chemical P removal were
the most sustainable choices. However, conventional soil-based systems without
polishing step generally do not comply with the existing Swedish guidelines in terms of
P reduction (Paper II) especially in a long-term perspective. Source-separation systems
with blackwater or urine diversion were the most sustainable options when high removal
and recycling of nutrients was prioritized (Scenario 2). In Scenario 3, where climate
change mitigation was the priority, systems with chemical P removal were favoured
mainly due to the production of biogas. These results also indicated that blackwater
systems would benefit from including an anaerobic digestion process to the treatment
train thus increasing their energy recovery. The ELECTRE III method proved to be
appropriate for the study because both qualitative and quantitative indicators can be
combined, and the use of thresholds accounted for data uncertainty.
The field study showed that the investigated sand filters generally had low P-removal
capacity (36% removal, geometric mean) (Paper II), in line with previous research.
Moreover, the effluent wastewater exceeded, in some cases, the criteria set for excellent
water quality by the EU bathing water directive with regard to intestinal enterococci and
E. coli (Paper III). The inefficiency of the sand filters could be explained by obvious
clogging processes (sand filter B), and less apparent reasons (sand filters F and G), such as
the construction of the sand filter, the chemical composition of the sand used or the
uneven distribution of the wastewater through the filter. Furthermore, the dilution
problems observed in three sand filters (A, C and E) highlights the need of correct
construction and installation of the systems in order to prevent mixing of the wastewater
with extraneous water. The results indicate that visual inspection is not enough to assess
the treatment function of the systems as some systems may appear to work well

45
hydraulically but not in terms of treatment. A downstream polishing step for P removal
is often advisable.
Alkaline P-filters generally removed P to a high extent (75–99% P reduction), with some
exceptions, including a P-filter that was old and clogged (E), and two that received too
high hydraulic loads (J, K). The alkaline P-filters placed in package plants (I, J, K, L)
exceeded the recommended European guidelines for intestinal enterococci, possibly
because of the shorter residence times as compared to the other P-filters. The removal of
P correlated positively with the pH measured in the effluent (Spearman rank correlation
coefficient=0.833, p=0.005), indicating that pH could be used as an indicator of P-filter
efficiency. Despite the high pH (Table 6), no further bacterial reduction in the effluent
of the alkaline P-filters was confirmed, probably due to the characteristic flow-saturation
of the filters and limited attachment possibilities for the bacteria. Moderate positive
correlations of indicator bacteria with effluent concentrations of P (Spearman rank
correlations between 0.51–0.69, p <0.001) and organic matter (Table 7) suggests that
alkaline P-filters could have the potential to further reduce bacterial concentrations.

46
References
Aicher, A. (2018). Bauhaus-University Weimar. Personal communication.
Almeida Dias, J., Figueira, J. R. and Roy, B. (2006) The software ELECTRE III-IV. Methodology and
user manual (Version 3.x). Paris: University Paris - Dauphine, LAMSADE. Available at:
https://www.lamsade.dauphine.fr/spip.php?article560&lang=fr.
Arias, C. A., Del Bubba, M. and Brix, H. (2001) ‘Phosphorus removal by sands for use as media in
subsurface flow constructed reed beds’, Water Research, 35(5), pp. 1159–1168. doi: 10.1016/S0043-
1354(00)00368-7.
Audet, J., Martinsen, L., Hasler, B., De Jonge, H., Karydi, E., Ovesen, N. B. and Kronvang, B. (2014)
‘Comparison of sampling methodologies for nutrient monitoring in streams: Uncertainties, costs and
implications for mitigation’, Hydrology and Earth System Sciences, 18(11), pp. 4721–4731. doi:
10.5194/hess-18-4721-2014.
Axler, R. P., Henneck, J. R., McCarthy, B. J., Hicks, R. E. and Olson, M. R. (2005) ‘Seasonal virus
removal by alternative onsite wastewater treatment systems’, Journal of Water and Health. IWA
Publishing, 3(2), pp. 139–155. doi: 10.2166/wh.2005.0014.
Balkema, A. J., Preisig, H. A., Otterpohl, R., Lambert, A. J. D. and Weijers, S. R. (2001) ‘Developing
a model based decision support tool for the identification of sustainable treatment options for domestic
wastewater’, Water Science & Technology, 43(7), pp. 265–269.
Balkema, A. J., Preisig, H. A., Otterpohl, R. and Lambert, F. J. (2002) ‘Indicators for the sustainability
assessment of wastewater treatment systems’, Urban Water, 4(2), pp. 153–161. doi: 10.1016/S1462-
0758(02)00014-6.
Barraud, S., Miramond, M. and Le Gauffre, P. (2004) ‘A-posteriori analysis of the relevance of a
multicriteria decision-aid method for best management practice choice in urban storm drainage’, Urban
Water Journal. Taylor & Francis Group, 1(3), pp. 209–216. doi: 10.1080/15730620410001732035.
Bertrand–Krajewski, J.-L., Barraud, S. and Bardin, J.-P. (2002) ‘Uncertainties, performance indicators
and decision aid applied to stormwater facilities’, Urban Water. Elsevier, 4(2), pp. 163–179. doi:
10.1016/S1462-0758(02)00016-X.
Blazejewski, R. and Murat-Blazejewska, S. (1997) ‘Soil clogging phenomena in constructed wetlands
with subsurface flow’, Water Science & Technology, 35(5), pp. 183–188.
Blum, K. M., Andersson, P. L., Renman, G., Ahrens, L., Gros, M., Wiberg, K. and Haglund, P.
(2017) ‘Non-target screening and prioritization of potentially persistent, bioaccumulating and toxic
domestic wastewater contaminants and their removal in on-site and large-scale sewage treatment
plants’, Science of the Total Environment. Elsevier, 575, pp. 265–275. doi:
10.1016/j.scitotenv.2016.09.135.
Bradley, B. R., Daigger, G. T., Rubin, R. and Tchobanoglous, G. (2002) ‘Evaluation of onsite
wastewater treatment technologies using sustainable development criteria’, Clean Technologies and
Environmental Policy, 4(2), pp. 87–99. doi: 10.1007/s10098-001-0130-y.
Bunce, J. T., Ndam, E., Ofiteru, I. D., Moore, A. and Graham, D. W. (2018) ‘A review of phosphorus
removal technologies and their applicability to small-scale domestic wastewater treatment systems’,
Frontiers in Environmental Science. Frontiers, 6, p. 8. doi: 10.3389/fenvs.2018.00008.
Carr, S. A., Liu, J. and Tesoro, A. G. (2016) ‘Transport and fate of microplastic particles in wastewater
treatment plants’, Water Research. Pergamon, 91, pp. 174–182. doi: 10.1016/J.WATRES.2016.01.002.
Christopherson, S. H., Anderson, J. J. and Gustafson, D. M. (2005) ‘Evaluation of recirculating sand
filter in a cold climate’, Water Science and Technology, 51(10), pp. 267–272.
Comeau, Y., Hall, K. J., Hancock, R. E. W. and Oldham, W. K. (1986) ‘Biochemical model for
enhanced biological phosphorus removal’, Water Research. Pergamon, 20(12), pp. 1511–1521. doi:
10.1016/0043-1354(86)90115-6.

47
Cooper, J. A., Loomis, G. W. and Amador, J. A. (2016) ‘Hell and high water: diminished septic system
performance in coastal regions due to climate change’, PLOS ONE, 11(9), p. e0162104. doi:
10.1371/journal.pone.0162104.
Cordell, D. (2013) ‘Peak phosphorus and the role of P recovery in achieving food security’, in Larsen,
T. A., Udert, K. M., and Lienert, J. (eds) Source Separation and Decentralization for Wastewater
Management. London: IWA Publishing, pp. 29–44.
Corominas, L., Foley, J., Guest, J. S., Hospido, A., Larsen, H. F., Morera, S. and Shaw, A. (2013) ‘Life
cycle assessment applied to wastewater treatment: State of the art’, Water Research, 47, pp. 5480–5492.
doi: 10.1016/j.watres.2013.06.049.
Cucarella, V., Zaleski, T., Mazurek, R. and Renman, G. (2007) ‘Fertilizer potential of calcium-rich
substrates used for phosphorus removal from wastewater’, Polish Journal of Environmental Studies, 16(6),
pp. 817–822.
Cucarella, V. and Renman, G. (2009) ‘Phosphorus sorption capacity of filter materials used for on-site
wastewater treatment determined in batch experiments–A comparative study’, Journal of Environment
Quality, 38(2), pp. 381–392. doi: 10.2134/jeq2008.0192.
Diaz-Elsayed, N., Xu, X., Balaguer-Barbosa, M. and Zhang, Q. (2017) ‘An evaluation of the
sustainability of onsite wastewater treatment systems for nutrient management’, Water Research.
Pergamon, 121, pp. 186–196. doi: 10.1016/j.watres.2017.05.005.
Dodgson, J. S., Spackman, M., Pearman, A. and Phillips, L. D. (2009) Multi-criteria Analysis: A Manual.
London. doi: 10.1002/mcda.399.
Ek, M., Junestedt, C., Larsson, C., Olshammar, M. and Ericsson, M. (2011) På uppdrag av
Naturvårdsverket Teknikenkät -enskilda avlopp 2009 [A technical survey on on-site wastewater treatment 2009].
Swedish Environmental Emissions Data, Report 44, 1-49, Norrköping. (In Swedish).
EU (2006) Directive 2006/7/EC of the European Parliament and of the Council of 15 February 2006
concerning the management of bathing water quality and repealing Directive.
European Committee for Standardization (1997) CEN/TC 230. EN 1484:1997. Water analysis -
Guidelines for the determination of total organic carbon (TOC) and dissolved organic carbon (DOC).
European Committee for Standardization (1998) CEN/TC 230. EN 1899-1:1998. Water quality -
Determination of biochemical oxygen demand after n days (BODn) - Part 1: dilution and seeding method with
allylthiourea addition (ISO 5815:1989, modified).
European Committee for Standardization (2005) CEN/TC 230. EN 872:2005. Water quality -
Determination of suspended solids - Method by filtration through glass fibre filters.
European Committee for Standardization (2011) CEN/TC 444. EN 16192:2011. Characterization of
waste - Analysis of eluates.
Eveborn, D., Kong, D. and Gustafsson, J. P. (2012) ‘Wastewater treatment by soil infiltration: Long-
term phosphorus removal’, Journal of Contaminant Hydrology, 140, pp. 24–33. doi:
10.1016/j.jconhyd.2012.08.003.
Figueira, J., Mousseau, V. and Roy, B. (2005) Multiple Criteria Decision Analysis: State of the Art Surveys.
Edited by M. E. José Figueira, Salvatore Greco. Boston: Springer.
Figueira, J. and Roy, B. (2002) ‘Determining the weights of criteria in the ELECTRE type methods
with a revised Simos’ procedure’, European Journal of Operational Research, 139, pp. 317–326.
Gago-Ferrero, P., Gros, M., Ahrens, L. and Wiberg, K. (2017) ‘Impact of on-site, small and large scale
wastewater treatment facilities on levels and fate of pharmaceuticals, personal care products, artificial
sweeteners, pesticides, and perfluoroalkyl substances in recipient waters’, Science of The Total
Environment, 601–602, pp. 1289–1297. doi: 10.1016/j.scitotenv.2017.05.258.

48
Gill, L. W., O’Luanaigh, N., Johnston, P. M., Misstear, B. D. R. and O’Suilleabhain, C. (2009)
‘Nutrient loading on subsoils from on-site wastewater effluent, comparing septic tank and secondary
treatment systems’, Water Research. Pergamon, 43(10), pp. 2739–2749. doi:
10.1016/j.watres.2009.03.024.
Gold, A. J., Lamb, B. E., Loomis, G. W., Boyd, J. R., Cabelli, V. J. and McKiel, C. G. (1992)
‘Wastewater renovation in buried and recirculating sand filters’, Journal of Environment Quality.
American Society of Agronomy, Crop Science Society of America, and Soil Science Society of
America, 21(4), pp. 720–725. doi: 10.2134/jeq1992.00472425002100040030x.
Green, W. and Ho, G. (2005) ‘Small scale sanitation technologies’, Water Science & Technology, 51(10),
pp. 29–38.
Gros, M., Blum, K. M., Jernstedt, H., Renman, G., Rodríguez-Mozaz, S., Haglund, P., Andersson, P.
L., Wiberg, K. and Ahrens, L. (2017) ‘Screening and prioritization of micropollutants in wastewaters
from on-site sewage treatment facilities’, Journal of Hazardous Materials. Elsevier, 328, pp. 37–45. doi:
10.1016/j.jhazmat.2016.12.055.
Gunady, M., Shishkina, N., Tan, H. and Rodriguez, C. (2015) ‘A review of on-site wastewater
treatment systems in Western Australia from 1997 to 2011’, Journal of Environmental and Public Health.
Hindawi, p. 12. doi: 10.1155/2015/716957.
Gustafsson, J. P., Renman, A., Renman, G. and Poll, K. (2008) ‘Phosphate removal by mineral-based
sorbents used in filters for small-scale wastewater treatment’, Water Research, 42(1–2), pp. 189–197. doi:
10.1016/j.watres.2007.06.058.
Hagedorn, C., Hansen, D. T. and Simonson, G. H. (1978) ‘Survival and movement of fecal indicator
bacteria in soil under conditions of saturated flow’, Journal of Environment Quality. American Society of
Agronomy, Crop Science Society of America, and Soil Science Society of America, 7(1), pp. 55–59.
doi: 10.2134/jeq1978.00472425000700010011x.
Hedin, J. (2018) ‘Markbaserad rening-en studie av funktion i fält [Soil-based wastewater treatment-a
field study]’, VATTEN - Journal of Water Management and Research , 74(1–2), pp. 27–45.(In Swedish
with English abstract)
Heinonen-Tanski, H. and Matikka, V. (2017) ‘Chemical and microbiological quality of effluents from
different on-site wastewater treatment systems across Finland and Sweden’, Water. Multidisciplinary
Digital Publishing Institute, 9(1), p. 47. doi: 10.3390/w9010047.
Heistad, A., Paruch, A. M., Vråle, L., Ádám, K. and Jenssen, P. D. (2006) ‘A high–performance
compact filter system treating domestic wastewater’, Ecological Engineering, 28(4), pp. 374–379. doi:
10.1016/j.ecoleng.2006.06.011.
Hellström, D., Jeppsson, U. and Kärrman, E. (2000) ‘A framework for systems analysis of sustainable
urban water management’, Environmental Impact Assessment Review, 20(3), pp. 311–321. doi:
10.1016/S0195-9255(00)00043-3.
Hellström, D. and Jonsson, L. (2003) ‘Evaluation of small wastewater treatment systems’, Water Science
& Technology, 48(11–12), pp. 61–68.
Hellström, D. and Jonsson, L. (2006) ‘Evaluation of small onǦsite wastewater treatment systems’,
Management of Environmental Quality: An International Journal, 17(6), pp. 728–739. doi:
10.1108/14777830610702548.
Henze, M., van Loosdrecht, M. ., Ekama, G. . and Brdjanovic, D. (Eds) (2008) Biological Wastewater
Treatment. London: IWA Publishing.
Herrmann, I., Nordqvist, K., Hedström, A. and Viklander, M. (2014) ‘Effect of temperature on the
performance of laboratory-scale phosphorus-removing filter beds in on-site wastewater treatment’,
Chemosphere, 117, pp. 360–366. doi: 10.1016/j.chemosphere.2014.07.069.

49
Herrmann, I., Jourak, A., Hedström, A., Lundström, S. and Viklander, M. (2014) ‘Enhancing the
reliability of laboratory phosphorus filter tests: effect of influent properties and interpretation of effluent
parameters’, Water Air Soil Pollution, 225:1766, pp. 1–11. doi: 10.1007/s11270-013-1766-z.
Hijnen, W. A. M., Schijven, J. F., Bonné, P., Visser, A. and Medema, G. J. (2004) ‘Elimination of
viruses, bacteria and protozoan oocysts by slow sand filtration.’, Water science and technology : a journal of
the International Association on Water Pollution Research, 50(1), pp. 147–54.
Hübinette, M. (2009) Tillsyn på minireningsverk inklusive mätning av funktion [Inspection of package plants
including measurement of functioning]. Report 2009:07. Västra Götalands county. (In Swedish).
Ibáñez-Forés, V., Bovea, M. D. and Pérez-Belis, V. (2014) ‘A holistic review of applied methodologies
for assessing and selecting the optimal technological alternative from a sustainability perspective’, Journal
of Cleaner Production. Elsevier, 70, pp. 259–281. doi: 10.1016/J.JCLEPRO.2014.01.082.
ISO 6878:2005 (2005) Water quality - Determination of phosphorus - Ammonium molybdate spectrometric
method (ISO 6878:2005).
Jenssen, P. D., Mæhlum, T., Krogstad, T. and Vråle, L. (2005) ‘High performance constructed
wetlands for cold climates’, Journal of Environmental Science and Health - Part A Toxic/Hazardous Substances
and Environmental Engineering. Taylor & Francis Group, 40(6–7), pp. 1343–1353. doi: 10.1081/ESE-
200055846.
Jenssen, P. D., Krogstad, T., Paruch, A. M., Mæhlum, T., Adam, K., Arias, C. A., Heistad, A.,
Jonsson, L., Hellström, D., Brix, H., Yli-Halla, M., Vråle, L. and Valve, M. (2010) ‘Filter bed systems
treating domestic wastewater in the Nordic countries – Performance and reuse of filter media’,
Ecological Engineering, 36(12), pp. 1651–1659. doi: 10.1016/j.ecoleng.2010.07.004.
Jönsson, H., Baky, A., Jeppsson, U., Hellström, D. and Kärrman, E. (2005) Composition of Urine, Feaces,
Greywater and Biowaste for Utilisation in the URWARE Model. Report 2005:6, Urban Water Report.
Gothenburg, Sweden: Urban Water, Chalmers University of Technology.
Kabir, G., Sadiq, R. and Tesfamariam, S. (2014) ‘A review of multi-criteria decision-making methods
for infrastructure management’, Structure and Infrastructure Engineering. Taylor & Francis, 10(9), pp.
1176–1210. doi: 10.1080/15732479.2013.795978.
Kauppinen, A., Martikainen, K., Matikka, V., Veijalainen, A.-M., Pitkänen, T., Heinonen-Tanski, H.
and Miettinen, I. T. (2014) ‘Sand filters for removal of microbes and nutrients from wastewater during
a one-year pilot study in a cold temperate climate’, Journal of Environmental Management, 133, pp. 206–
213. doi: 10.1016/j.jenvman.2013.12.008.
Kohler, L. E., Silverstein, J. and Rajagopalan, B. (2016) ‘Modeling on-site wastewater treatment system
performance fragility to hydroclimate stressors’, Water Science, 74(12), pp. 2917–2926. doi:
10.2166/wst.2016.467.
Kristiansen, R. (1981) ‘Sand-filter trenches for purification of septic tank effluent: I. The clogging
mechanism and soil physical environment’, Journal of Environmental Quality, 10(3), pp. 353–357.
Kujawa-Roeleveld, K., Fernandes, T., Wiryawan, Y., Tawfik, A., Visser, M. and Zeeman, G. (2005)
‘Performance of UASB septic tank for treatment of concentrated black water within DESAR concept’,
Water Science & Technology, 55, pp. 307–313.
Larsen, T. and Maurer, M. (2011) Source Separation and Decentralization, Treatise on Water Science. Edited
by P. Wilderer. Oxford: Academic Press: Elsevier. doi: 10.1016/B978-0-444-53199-5.00083-X.
Larsen, T. A., Udert, K. M., Lienert, J. and Edit (2013) Source Separation and Decentralization for
Wastewater Management, IWA Publishing. London: IWA Publishing. doi:
10.1017/CBO9781107415324.004.
Larsen, T. A., Hoffmann, S., Lüthi, C., Truffer, B. and Maurer, M. (2016) ‘Emerging solutions to the
water challenges of an urbanizing world’, Science, 352(6288), pp. 928–933.

50
Larsson, C., Forsberg, B. and Engström, T. (2017) Uppföljande kontroll av nya små avloppsanläggningar
[Follow up control of new small wastewater facilities]. Kungsbacka. (In Swedish).
Lehtoranta, S., Vilpas, R. and Mattila, T. . (2014) ‘Comparison of carbon footprints and eutrophication
impacts of rural on-site wastewater treatment plants in Finland’, Journal of Cleaner Production, 65, pp.
439–446. doi: 10.1016/j.jclepro.2013.08.024.
Lennartsson, M., Kvarnström, E., Lundberg, T., Buenfil, J. and Sawyer, R. (2009) Comparing sanitation
systems using sustainability criteria-EcoSanRes Series, 2009-1. Stockholm.
Libralato, G. (2013) ‘The small-scale approach in wastewater treatment’, in Wastewater Reuse and
Management. Dordrecht: Springer, pp. 195–224. doi: 10.1007/978-94-007-4942-9_7.
Lundin, M., Bengtsson, M. and Molander, S. (2000) ‘Life cycle assessment of wastewater systems:
influence of system boundaries and scale on calculated environmental loads’, Environmental Science and
technology, 34(1), pp. 180–186. doi: 10.1021/es990003f.
Lundin, M., Molander, S. and Morrison, G. M. (1999) ‘A set of indicators for the assessment of
temporal variations in the sustainability of sanitary systems’, in Water Science and Technology. No longer
published by Elsevier, pp. 235–242. doi: 10.1016/S0273-1223(99)00107-9.
Magnusson, K., Eliasson, K., Fråne, A., Haikonen, K., Hultén, J., Olshammar, M., Stadmark, J. and
Voisin, A. (2016) Swedish sources and pathways for microplastics to the marine environment - A review of existing
data. Report C 183. Stockholm.
Martikainen, K., Veijalainen, A.-M., Torvinen, E., Heinonen-Tanski, H., Kauppinen, A., Pitkanen,
T., Miettinen, I. and Matikka, V. (2018) ‘Efficiency of private household sand filters in removing
nutrients and microbes from wastewater in Finland’, Water, 10(1000), p. 16. doi: 10.3390/w10081000.
Matamoros, V., Arias, C., Brix, H. and Bayona, J. M. (2009) ‘Preliminary screening of small-scale
domestic wastewater treatment systems for removal of pharmaceutical and personal care products’,
Water Research. Pergamon, 43(1), pp. 55–62. doi: 10.1016/j.watres.2008.10.005.
McConville, J., Drangert, J.-O., Tidåker, P., Neset, T.-S., Rauch, S., Strid, I. and Tonderski, K.
(2015) ‘Closing the food loops: guidelines and criteria for improving nutrient management’,
Sustainability: Science, Practice, & Policy, 11:2, pp. 33–43. doi: 10.1080/15487733.2015.11908144.
McConville, J. R., Kvarnström, E., Jönsson, H. and Kärrman, E. (2016) ‘Source separation: How ready
in [sic] the Swedish wastewater sector for technology transition?’, in 13th IWA Specialized Conference on
Small Water and Wastewater Systems and 5th IWA Specialized Conference on Resources-Oriented Sanitation.
Athens, Greece.: IWA.
Molinos-Senante, M., Gómez, T., Garrido-Baserba, M., Caballero, R. and Sala-Garrido, R. (2014)
‘Assessing the sustainability of small wastewater treatment systems: A composite indicator approach’,
The Science of the Total Environment, 497–498, pp. 607–617. doi: 10.1016/j.scitotenv.2014.08.026.
N. J. Barrow, T. C. S. (1975) ‘The slow reactions between soil and anions: 2. Effect of time and
temperature on the decrease in phosphate concentration in the soil solution’, Soil Science. Soil Science,
119(2), pp. 167–177.
Nilsson, C., Renman, G., Johansson Westholm, L., Renman, A. and Drizo, A. (2013) ‘Effect of
organic load on phosphorus and bacteria removal from wastewater using alkaline filter materials’, Water
Research, 47(16), pp. 6289–6297. doi: 10.1016/j.watres.2013.08.001.
Nilsson, C., Lakshmanan, R., Renman, G. and Rajarao, G. K. (2013) ‘Efficacy of reactive mineral-
based sorbents for phosphate, bacteria, nitrogen and TOC removal-column experiment in recirculation
batch mode’, Water research, 47(14), pp. 5165–5175. doi: 10.1016/j.watres.2013.05.056.
Oehmen, A., Lemos, P. C., Carvalho, G., Yuan, Z., Keller, J., Blackall, L. L. and Reis, M. A. M.
(2007) ‘Advances in enhanced biological phosphorus removal: From micro to macro scale’, Water
Research, pp. 2271–2300. doi: 10.1016/j.watres.2007.02.030.

51
Olshammar, M., Ek, M., Rosenquist, L., Ejhed, H., Sidvall, A. and Svanström, S. (2015) På uppdrag av
Havs-och vattenmyndigheten Uppdatering av kunskapsläget och statistik för små avloppsanläggningar [The state of
knowledge and statistics for small sewage plants. Swedish Agency for Marine and Water Management]. Report
166. Norrköping. (In Swedish).
Ottoson, J., Hansen, A., Björlenius, B., Norder, H. and Stenström, T. A. (2006) ‘Removal of viruses,
parasitic protozoa and microbial indicators in conventional and membrane processes in a wastewater
pilot plant’, Water Research, 40(7), pp. 1449–1457. doi: 10.1016/j.watres.2006.01.039.
Palm, O., Elmefors, E., Moraeus, P., Nilsson, P., Persson, L., Ridderstolpe, P. and Eveborn, D. (2012)
Läget inom markbaserad avlopsvattenrening - Samlad kunskap king reningstekniker för små och enskilda avlopp
[State of the art in soil-based wastewater treatment - Review of treatment techniques for small-scale wastewater
treatment]. Swedish EPA, Report 6484, 1-35, Stockholm. (In Swedish with English summary).
Palm, O., Malmén, L. and Jönsson, H. (2002) Robusta, uthålliga små avloppssytem. En
kunskapssammanställning [Robust and durable small wastewater systems. A knowledge compilation]. Report 5224.
Stockholm. (In Swedish with English summary).
Paruch, A. M., Mæhlum, T., Haarstad, K., Blankenberg, A.-G. B. and Hensel, G. (2016) ‘Performance
of Constructed Wetlands Treating Domestic Wastewater in Norway Over a Quarter of a Century –
Options for Nutrient Removal and Recycling’, in Natural and Constructed Wetlands. Cham: Springer
International Publishing, pp. 41–55. doi: 10.1007/978-3-319-38927-1_3.
Pell, M. and Nyberg, F. (1989) ‘Infiltration of Wastewater in a Newly Started Pilot Sand-Filter System:
I. Reduction of Organic Matter and Phosphorus’, Journal of Environment Quality. American Society of
Agronomy, Crop Science Society of America, and Soil Science Society of America, 18(4), p. 451. doi:
10.2134/jeq1989.00472425001800040009x.
Reneau, R. B., Hagedorn, C. and Degen, M. J. (1989) ‘Fate and transport of biological and inorganic
contaminants from on-site disposal of domestic wastewater’, Journal of Environment Quality. American
Society of Agronomy, Crop Science Society of America, and Soil Science Society of America, 18(2),
pp. 135–144. doi: 10.2134/jeq1989.00472425001800020001x.
Renman, A. and Renman, G. (2010) ‘Long-term phosphate removal by the calcium-silicate material
Polonite in wastewater filtration systems’, Chemosphere, 79(6), pp. 659–664. doi:
10.1016/j.chemosphere.2010.02.035.
Robertson, W. D., Schiff, S. L. and Ptacek, C. J. (1998) ‘Review of phosphate mobility and persistence
in 10 septic system plumes’, Ground Water. Wiley/Blackwell (10.1111), 36(6), pp. 1000–1010. doi:
10.1111/j.1745-6584.1998.tb02107.x.
Rogers, M. and Bruen, M. (1998) ‘Choosing realistic values of indifference, preference and veto
thresholds for use with environmental criteria within ELECTRE’, European Journal of Operational
Research, 107, pp. 542–551.
Rolland, L., Molle, P., Liénard, A., Bouteldja, F. and Grasmick, A. (2009) ‘Influence of the physical
and mechanical characteristics of sands on the hydraulic and biological behaviors of sand filters’,
Desalination, 248, pp. 998–1007. doi: 10.1016/j.desal.2008.10.016.
Rostvall, A., Zhang, W., Dürig, W., Renman, G., Wiberg, K., Ahrens, L. and Gago-Ferrero, P. (2018)
‘Removal of pharmaceuticals, perfluoroalkyl substances and other micropollutants from wastewater
using lignite, Xylit, sand, granular activated carbon (GAC) and GAC+Polonite®in column tests–Role
of physicochemical properties’, Water Research, 137, pp. 97–106. doi: 10.1016/j.watres.2018.03.008.
Rowley, H. V., Peters, G. M., Lundie, S. and Moore, S. J. (2012) ‘Aggregating sustainability indicators:
Beyond the weighted sum’, Journal of Environmental Management. Academic Press, 111, pp. 24–33. doi:
10.1016/J.JENVMAN.2012.05.004.
Roy, B. (1978) ELECTRE III: Un algorithme de classement fondé sur une représentation floue des préférences en
présence de critères multiples [A ranking algorithm based on a fuzzy representation of preferences in the presence of
multiple criteria]. Cahiers du Centre d’Ét. (In French).

52
Roy, B. (1990) The Outranking Approach and the Foundations of Electre Methods, Readings in Multiple
Criteria Decision Aid. Edited by Bana e Costa C.A. Berlin, Heidelberg: Springer Berlin Heidelberg. doi:
10.1007/978-3-642-75935-2_8.
Roy, B. (1996) Multicriteria Methodogy for Decision Aiding. 1st edn. Edited by S. S. & B. Media. Springer
US. doi: 10.1007/978-1-4757-2500-1.
Scandura, J. E. and Sobsey, M. D. (1997) ‘Viral and bacterial contamination of groundwater from on-
site sewage treatment systems’, in Water Science and Technology, pp. 141–146. doi: 10.1016/S0273-
1223(97)00249-7.
Schaider, L. A., Rodgers, K. M. and Rudel, R. A. (2017) ‘Review of Organic Wastewater Compound
Concentrations and Removal in Onsite Wastewater Treatment Systems’, Environmental Science and
Technology. American Chemical Society, 51(13), pp. 7304–7317. doi: 10.1021/acs.est.6b04778.
Scheiner, M. (2018). Swiss Federal Institute of Aquatic Science and technology. Personal communication.
Schoen, M. E., Xue, X., Wood, A., Hawkins, T. R., Garland, J. and Ashbolt, N. J. (2017) ‘Cost,
energy, global warming, eutrophication and local human health impacts of community water and
sanitation service options’, Water Research, 109, pp. 186–195. doi: 10.1016/j.watres.2016.11.044.
Seguí, M. del M. B., Hess, T., Sakrabani, R. and Tyrrel, S. (2017) ‘Long-term phosphorus removal in
land treatment systems: Evaluation, experiences, and opportunities’, Critical Reviews in Environmental
Science and Technology. Taylor & Francis, 47(5), pp. 314–334. doi: 10.1080/10643389.2017.1318617.
Sélas, B., Lakel, A., Andres, Y. and Le Cloirec, P. (2003) ‘Wastewater reuse in on-site wastewater
treatment: bacteria and virus movement in unsaturated flow through sand filter’, Water Science and
Technology. IWA Publishing, 47(1), pp. 59–64. doi: 10.2166/wst.2003.0016.
Senecal, J., Nordin, A., Simha, P. and Vinnerås, B. (2018) ‘Hygiene aspect of treating human urine by
alkaline dehydration’, Water Research. Pergamon, 144, pp. 474–481. doi:
10.1016/J.WATRES.2018.07.030.
Simha, P., Senecal, J., Nordin, A., Lalander, C. and Vinnerås, B. (2018) ‘Alkaline dehydration of
anion–exchanged human urine: Volume reduction, nutrient recovery and process optimisation’, Water
Research. Pergamon, 142, pp. 325–336. doi: 10.1016/J.WATRES.2018.06.001.
Simos, J. (1990) Évaluer l’impact sur l’environnement. Une approche originale par l’analyse multicritère et la
négotiation [Environmental impact assessment. An original approach for multi-criteria analysis and negociation],
Géographie physique et Quaternaire. Laussane.: Presses Polytechniques et Universitaires Romandes. doi:
10.7202/032939ar. (In Fench).
Sinclair, A., Jamieson, R., Gordon, R. J., Madani, A. and Hart, W. (2014) ‘Modeling Phosphorus
Treatment Capacities of On-Site Wastewater Lateral Flow Sand Filters’, Journal of Environmental
Engineering, 140(2), p. 04013002. doi: 10.1061/(ASCE)EE.1943-7870.0000796.
Song, Y., Hahn, H. H., Hoffmann, E. and Weidler, P. G. (2006) ‘Effect of humic substances on the
precipitation of calcium phosphate’, Journal of Environmental Sciences, 18(5), pp. 852–857. doi:
10.1016/S1001-0742(06)60004-1.
Spångberg, J., Tidåker, P. and Jönsson, H. (2014) ‘Environmental impact of recycling nutrients in
human excreta to agriculture compared with enhanced wastewater treatment’, Science of The Total
Environment. Elsevier, 493, pp. 209–219. doi: 10.1016/J.SCITOTENV.2014.05.123.
Spiller, M., Vreeburg, J. H. G., Leusbrock, I. and Zeeman, G. (2015) ‘Flexible design in water and
wastewater engineering--definitions, literature and decision guide.’, Journal of environmental management,
149, pp. 271–81. doi: 10.1016/j.jenvman.2014.09.031.
Steffen, W., Richardson, K. and Rockström, J. (2015) ‘Planetary boundaries: Guiding human
development on a changing planet’, Science, 347(6223), p. 12.
Stenström, T. A. (2013) ‘Hygiene, a major challenge for source separation’, in Larsen, T. A., Udert, K.
M., and Lienert, J. (eds) Source Separation and Decentralization for Wastewater Management. London: IWA
Publishing, pp. 151–161.

53
Stevik, T. K., Kari Aa, Ausland, G. and Fredrik Hanssen, J. (2004) ‘Retention and removal of
pathogenic bacteria in wastewater percolating through porous media: a review’, Water Research, 38(6),
pp. 1355–1367. doi: 10.1016/j.watres.2003.12.024.
Swedish EPA (2006) Naturvårdsverkets författningssamling. Naturvårdsverkets allmänna råd om små
avloppsanordningar för hushållsspillvatten [General advice about small-scale household wastewater treatment
facilities], ISSN 1403-8234. Sweden. (In Swedish).
Swedish Ministry of Environment and Energy (2018) Vägar till hållbara vattentjänster [The way to
sustainable water services]. SOU 2018:34. Stockholm. (In Swedish).
Talvitie, J., Heinonen, M., Paakkonen, J.-P., Vahtera, E., Mikola, A., Setala, O. and Vahala, R. (2015)
‘Do wastewater treatment plants act as a potential point source of microplastics? Preliminary study in
the coastal Gulf of Finland, Baltic Sea’, Water Science and Technology. IWA Publishing, 72(9), pp. 1495–
1504. doi: 10.2166/wst.2015.360.
Tanık, A. and Çomakoğlu, B. (1996) ‘Nutrient removal from domestic wastewater by rapid infiltration
system’, Journal of Arid Environments. Academic Press, 34(3), pp. 379–390. doi:
10.1006/JARE.1996.0118.
Tchobanoglous, G., Stensel, D. H., Tsuchihashi, R., Burton, F., Abu-Orf, M., Bowden, G. and
Pfrang, W. (2014) Wastewater Engineering: Treatment and Resource Recovery. 5th edn. Edited by Metcalf &
Eddy I AECOM. New York: McGraw-Hill.
Thibodeau, C., Monette, F., Bulle, C. and Glaus, M. (2014) ‘Comparison of black water source-
separation and conventional sanitation systems using life cycle assessment’, Journal of Cleaner Production,
67, pp. 45–57. doi: 10.1016/j.jclepro.2013.12.012.
Tidåker, P., Sjöberg, C. and Jönsson, H. (2007) ‘Local recycling of plant nutrients from small-scale
wastewater systems to farmland—A Swedish scenario study’, Resources, Conservation and Recycling, 49(4),
pp. 388–405. doi: 10.1016/j.resconrec.2006.05.004.
Tjandraatmadja, G., Sharma, A. K., Grant, T. and Pamminger, F. (2013) ‘A decision support
methodology for integrated urban water management in remote settlements’, Water Resources
Management, 27(2), pp. 433–449. doi: 10.1007/s11269-012-0195-x.
Tscheikner-Gratl, F., Egger, P., Rauch, W. and Kleidorfer, M. (2017) ‘Comparison of multi-criteria
decision support methods for integrated rehabilitation prioritization’, Water. Multidisciplinary Digital
Publishing Institute, 9(2), p. 68. doi: 10.3390/w9020068.
Udert, K. M., Larsen, T. A. and Gujer, W. (2003) ‘Biologically induced precipitation in urine-
collecting systems’, Water Science and Technology: Water Supply, 3(3), pp. 71–78.
USEPA (1999) ‘Water Efficiency Technology Fact Sheet - Incinerating Toilets, EPA 832-F-99-072’.
US Environmental Protection Authority.
USEPA (2002) ‘Onsite Wastewater Treatment Systems Manual (EPA/625/R-00/008)’, p. 367.
USEPA (2014) Decentralized Wastewater Program. Annual Report 2013 EPA-832-R-140006.
Velasquez, M. and Hester, P. T. (2013) ‘An analysis of multi - criteria decision making methods’,
International Journal of Operations Research, 10(2), pp. 56–66.
Vilpas, R. and Santala, E. (2007) ‘Comparison of the nutrient removal efficiency of onsite wastewater
treatments systems: applications of conventional sand filters and sequencing batch reactors (SBR)’,
Water Science & Technology, 55(7), pp. 109–117. doi: 10.2166/wst.2007.134.
Vincke, P. (1992) Multicriteria Decision - Aid. Chichester: John Wiley & Sons.
Vohla, C., Kõiv, M., Bavor, H. J., Chazarenc, F. and Mander, Ü. (2011) ‘Filter materials for
phosphorus removal from wastewater in treatment wetlands-A review’, Ecological Engineering, 37(1), pp.
70–89. doi: 10.1016/j.ecoleng.2009.08.003.
VUNA (2018) VUNA ltd, an Eawag spin-off. Available at: http://vuna.ch/index_en.html.

54
Vymazal, J. (2011) ‘Constructed wetlands for wastewater treatment: Five decades of experience’,
Environmental Science and Technology. American Chemical Society, 45(1), pp. 61–69. doi:
10.1021/es101403q.
Weiss, P., Eveborn, D., Kärrman, E. and Gustafsson, J. P. (2008) ‘Environmental systems analysis of
four on-site wastewater treatment options’, Resources, Conservation and Recycling, 52(10), pp. 1153–1161.
doi: 10.1016/j.resconrec.2008.06.004.
Williamson, E. (2010) ‘Cold climate performance analysis of on-site domestic wastewater treatment
systems’, Water Environment Research, 82(6), pp. 512–518. doi: 10.2175/106143009X12529484815557.
Wilson, J., Boutilier, L., Jamieson, R., Havard, P. and Lake, C. (2011) ‘Effects of hydraulic loading rate
and filter length on the performance of lateral flow sand filters for on-site wastewater treatment’, Journal
of Hydrologic Engineering, 16(8), pp. 639–649. doi: 10.1061/(ASCE)HE.1943-5584.0000359.
Winker, M., Vinnerås, B., Muskolus, A., Arnold, U. and Clemens, J. (2009) ‘Fertiliser products from
new sanitation systems: Their potential values and risks’, Bioresource Technology, 100(18), pp. 4090–4096.
doi: 10.1016/j.biortech.2009.03.024.
Xue, X., Hawkins, T., Schoen, M., Garland, J. and Ashbolt, N. (2016) ‘Comparing the life cycle
energy consumption, global warming and eutrophication potentials of several water and waste service
options’, Water. Multidisciplinary Digital Publishing Institute, 8(4), p. 154. doi: 10.3390/w8040154.
Yeoman, S., Stephenson, T., Lester, J. N. and Perry, R. (1988) ‘The removal of phosphorus during
wastewater treatment: A review’, Environmental Pollution. Elsevier, 49(3), pp. 183–233. doi:
10.1016/0269-7491(88)90209-6.
Zeleny, M. (1982) Multiple Criteria Decision Making. New York: McGraw-Hill.

55