Table : General operational assumptions

Coupling Urine Diverting Flush Toilets with Anaerobic Co-Digestion in the Global South Modelling Climate Benefits in Kpong, Ghana

Bachelor’s Thesis In partial fulfillment of the academic requirements for the degree B.Sc. Environmental Engineering Chair of Urban Water Systems Engineering School of Engineering and Design Technical University Munich Supervisors M.Sc. Mohammad Bardi Chair of Urban Water Systems Engineering PD Dr.-Ing. Konrad Koch Chair of Urban Water Systems Engineering Submitted by Felix Giglberger Matriculation number 03710215 felix.giglberger@tum.de Date of submission 2022-06-12

Abstract

In the Global South, urine diverting dry toilets have been successfully tested as feasible tools to increase sanitation coverage. However, many people desire flush toilets. Some approaches to treat wastewater from flush toilets have been implemented successfully, yet many sanitation projects have failed because operation and maintenance costs were too high. Biogas production from wastewater can be seen as a tool that benefits operators directly and tangibly, reducing the likelihood of sanitation projects to fail due to economic reasons once in operation. However, decentralized treatment systems in the Global South that can cope with larger quantities of flushwater do not provide reliable gas retrieval mechanisms. Fixed-dome digesters (FDDs) can provide a simple retrieval mechanism without the need for electricity or moving parts, allowing for sustainable gas collection. However, they are not used for treatment of diluted wastes, as this would make them unreasonably large. Urine diverting flush toilets (UDFTs) can save large quantities of water due to reduced flushwater consumption for urinations and could therefore enable the combination of flush toilets with biogas collection.

For a specific trade center with a size of 120 person-equivalents in Ghana, the economic and greenhouse gas (GHG) mitigation benefits of using this combination were compared against a standard anaerobic wastewater treatment approach without biogas collection. It was analyzed how flushwater consumption and co-digestion with food waste affect operational benefits, increased investment costs, and space consumption. This was done by comparing the influences of the different treatment elements used in the two approaches, namely the FDD, Imhoff tank, anaerobic baffled reactor, urine storage and composting of food waste as an alternative to co-digestion in the FDD. A key underlying assumption was that all water and nutrients would be properly reused. Theoretical calculations were performed using URWARE wastewater data and treatment modelling approaches that were combined with DEWATS engineering guidelines. An emission modelling approach was set up to investigate how different flushwater and food waste quantities affect emissions from the compensation tank of the FDD. Additionally, effects of varying LPG and carbon prices as well as effects of leaks and lack of flaring were shown and it was investigated how results could be translated to differing user populations.

It was found that about 700 USD can be saved per year but additional initial investment cost would be as large as 13,600 USD. Still, 32% of initial investment costs were attributed to urine storage, which may not be needed in some cases. GHG emissions from construction would amortize already after 1.7 years and 200 t_CO2-eq could be mitigated after 20 years of operation. Adjustment with a carbon price of 100 USD t_CO2-eq^-1 would allow for economic amortization after 8.6 years, resulting in economic savings of 20,400 USD after 20 years of operation. Food waste dosage did not affect this result significantly, showing that co-digestion is not needed if this carbon price would be in effect. In conclusion, UDFTs are a powerful tool to mitigate GHG emissions and ensure continued treatment system operation in the Global South, however technology diffusion will require adequate funding mechanisms that can cope with the increased investment costs.

Table of Content

Abstract II

List of Figures V

List of Tables VI

Abbreviations VII

1 Introduction 1

1.1 The Sanitation Problem in the Global South 1

1.1.1 Cost of sanitation 1

1.1.2 The issue with flush toilets 2

1.2 The case study 2

1.2.1 The Ghanaian background 3

1.2.2 Economic incentives and environmental framework 3

1.2.3 Social factors 4

2 Theoretical Background 5

2.1 Urine diverting flush toilets 5

2.2 The DEWATS approach 6

2.2.1 Anaerobic digestion 6

2.2.2 Fully mixed fixed-dome digester 6

2.2.3 Imhoff tank and anaerobic baffled reactor 8

3 Research questions 9

4 Methodology 10

4.1 Toilet characteristics 12

4.2 Substrate characterization 13

4.3 Digestion and Emission Modelling 14

4.4 Fixed-Dome Digester 15

4.5 Imhoff tank 18

4.6 Anaerobic baffled reactor 19

4.7 Urine treatment 20

4.8 Composting 21

5 Results 23

5.1 Gas production and emission 23

5.2 Flushwater consumption 24

5.3 Food waste dosage 25

5.4 Gas losses 26

5.5 LPG and Carbon Price 27

5.6 Person-equivalents 27

6 Discussion 28

6.1 Result validation 28

6.2 Model limitations 28

6.3 Applicability in Ghana and the Global South 29

7 Conclusion and outlook 30

8 References Error! Bookmark not defined.

Appendix A – Design drawings vii

Appendix B – Contributions of different construction materials viii

Appendix C.1 – Flushwater effects ix

Appendix C.2 – Effects of food waste dosage xi

Appendix C.3 – Effects of gas losses xiii

Appendix C.4 – Effects of general system size xiv

List of Figures

Figure 1: Scheme of fixed-dome digester (Vögeli et. al. 2014) 7

Figure 2: Imhoff tank design - (Gutterer et al. 2009) 8

Figure 3: Design of an anaerobic baffled reactor - (Gutterer et al. 2009) 8

Figure 4: System 1 configuration 10

Figure 5: System 2 configuration 10

Figure 6: Adjusted substrate characterization based on URWARE data (H Jönsson et al. 2005). Q-Discharge, UDFT - Urine diverting flush toilet, SDFT – Standard dual flush toilet, WM – Wet mass, TS – Total Solids, VS – Volatile Solids, COD – Chemical Oxygen Demand, BOD, Biochemical oxygen demand, sbCOD – soluble biodegradable COD, nbsCOD – non biodegradable soluble COD, pbCOD – particulate biodegradable COD, nbpCOD – non biodegradable particulate COD, TP – Total Phosphorus, TN – Total Nitrogen 14

Figure 7: CH4 emissions compared between system 2 components 23

Figure 8: Comparison of systems regarding CH4 emissions 24

Figure 9: Effects of flushwater consumption on amortization time of system 2 24

Figure 10: Effects of food waste ratio on greenhouse gas emission mitigation potential of system 2 over system 1 26

Figure 11: Effects of food waste ratio on amortization time of system 2 compared to system 1 26

Figure 12: Effect of external gas losses on amortization time of system 2 compared to system 1 26

Figure 13: Effect of LPG price on Economic savings 27

Figure 14: Effects of LPG Price on amortization time of system 2 over system 1 27

Figure 15: Effect of carbon pricing on amortization time of system 2 over system 1 27

Figure 16: Imhoff tank design figures (DEWATS Chapter 9 Technical Components 2009) vii

Figure 17: Anaerobic baffled reactor design figures (DEWATS Chapter 9 Technical Components 2009) vii

List of Tables

Table 1: General operational assumptions 11

Table 2: Building material properties 11

Table 3: Toilet Parameters 13

Table 4: Physical constants regarding LPG substitution 15

Table 5: Fixed-dome parameters 16

Table 6: Calculated parameters for fixed-dome modelling 17

Table 7: Imhoff tank design parameters 18

Table 8: Anaerobic baffled reactor design parameters 19

Table 9: Urine treatment parameters 21

Table 10: Assumptions regarding composting 22

Abbreviations

ABR – Anaerobic Baffled reactor

ADM1 – Anaerobic Digestion Model 1

BOD – Biochemical oxygen demand

BORDA – Bremen Overseas Research and Development Association

bpCOD – biodegradable particulate COD

bsCOD – biodegradable soluble COD

COD – Chemical Oxygen Demand

DEWATS – Decentralized wastewater treatment systems

FDD – Fixed-dome digester

GHG – Greenhouse gas

nbpCOD – non biodegradable particulate COD

nbsCOD – non biodegradable soluble COD

O&M – Operation and maintenance

PE – Person-equivalents

SDFT – Standard dual-flush toilet

UDDT – Urine diverting dry toilet

UDFT – Urine diverting flush toilet

URWARE – Urban water research model

Introduction

The Sanitation Problem in the Global South

To achieve sustainable development goal number 6 aiming for full sanitation coverage by 2030, the rate at which sanitation coverage is increasing will need to quadruple. Half of the world’s population use sanitation services that leave human waste untreated, threatening human and environmental health as well as economic development. 26% do not have basic access to sanitation facilities, causing 830,000 deaths annually. In an environmental domain, inadequate management of human excreta is associated with pollution of ecosystems and contributes significantly to global greenhouse gas emissions and resource depletion. (UNICEF and WHO 2020)

Cost of sanitation

The economic cost of decreased productivity from death and disease in Sub-Saharan Africa has been estimated to be on average 3.2% of its Gross Domestic Product, which is why estimates of economic return generated from investments in the sanitation sector are generally beyond five times the investment. However, with an estimated annual budget of 105 billion US-dollars (USD) needed to reach universal sanitation by 2030, many countries have complained about a lack of possibilities for financing their goals. Yet, sanitation as a sector has been largely underprioritized and suffers from various institutional failures. This includes financial mismanagement, with operation and maintenance costs often not included in budgets. (UNICEF and WHO 2020).

Many governments focus on keeping investment costs to a minimum, but this can easily result in failures. In Ghana, many people have abandoned pit latrines co-financed by the government, as emptying would cost them more than paying for privately owned public toilets, with some reverting to open defecation (Appiah-Effah et al. 2014). Providing people with sanitation does make economic sense, but the return generated is largely intangible to the individual. With approaches that can generate tangible profits from wastewater treatment, donors and lenders could be less reluctant to provide the needed funding and projects would be less likely to fail if operation & maintenance costs are lower than the tangible benefits generated at the user level. However, such approaches will necessarily facilitate higher investment costs. This thesis aims to investigate one such approach and attempts to quantify the relevant economic parameters in order to enable decision-making on the feasibility of adoption.

The proliferation of cheap sanitation technologies also has an environmental cost. Shaw et al. (2021) showed that achieving the sanitation targets of the United Nation’s Sustainable Development Goals will increase greenhouse gas emissions by 7 to 68 Tg_CO2-eq yr^-1, which can amount to 0.2% of global anthropogenic emissions. This was mainly attributed to septic tanks and pit latrines, as these systems are anaerobic and subsequently emit more methane than natural decomposition. Capturing and burning the methane would therefore significantly reduce emissions caused by increasing sanitation coverage.

The issue with flush toilets

Cheap sanitation technologies may face problems of social acceptance. Narain (2002) concluded in grey literature that with increasing economic development people would discard cost-effective technology solutions and increasingly use flush toilets that convey wastewater without treatment, even if that would lead to more environmental problems. Flush toilets are seen as desirable ambition (J Behrendt et al. 2006, p. 6), while dry toilets are often seen as an inferior and unjust solution for the poor (Reynaud 2014; Larsen et al. 2021). However, flush toilets lead to surface water pollution and losses in biodiversity while also increasing reliance on mineral fertilizers (Larsen et al. 2021). They facilitate water stress (Larsen et al. 2013) and the high energy demand of treating water to drinking water quality for toilet flushing significantly contributes to the environmental impact of sanitation, even in the Global South (Wissing 2020). As the demand for flush toilets will grow, strategies need to be found to mitigate the negative impacts of flush toilets, so that the environmental problems caused in the Global North are not repeated in the same degree or worse in the Global South. Many of the negative impacts could be largely reduced by water reuse, however treatment technologies will need to adapt and still prevail within socioeconomic conditions. This thesis aims to quantify the economic and ecological differences of two treatment approaches towards diluted wastewater using established technologies for the Global South and to explore their sensitivity to the amount of flushwater in order to quantify effects of water-saving.

The case study

There are many input parameters needed for the analysis performed that can fluctuate and are subject to regional differences. In order to avoid complexity, some parameters will be fixed to standard values only valid for the specific case to be introduced. However, some of the parameters will be sensitized to enable decision making for similar cases.

The investigation focusses on a trade center near Kpong in the Eastern region of Ghana. It is expected to host 120 person-equivalents (PE) of workers and visitors. The center is built and operated by a NGO that has the goals of providing good working conditions for underprivileged women and revitalizing Ghana’s textile industry.

The trade center produces multiple sources of wastewater, but this thesis will not include any wastewater produced by the manufacturing processes. This decision has been made as (i) the wastewater remains poorly characterized as laboratories in Ghana lack capabilities to analyze parameters needed in the model used in this thesis, (ii) laboratories in Ghana have not always managed to provide realistic values for some analyzed parameters, (iii) industrial wastewater will likely be treated separately as it may not be treatable to satisfactory levels for agricultural reuse using local technologies, (iv) models used in this thesis have only been validated with municipal sources and (v) better comparability to other projects can be achieved as complexity is reduced.

The Ghanaian background

Ghana largely suffers from the same problems as the Global South in general. Set targets of the Millenium Development Goals have largely not been met by wide margins (Appiah-Effah et al. 2019). A survey conducted by the Ghana Statistical Survey (2018) found that only 21% of the population uses basic sanitation facilities, with next to none being safely managed. It established that pit latrines are the main sanitation technology, which is used by 40% of the population. 2% of the population has access to sewers and 19% have their excreta removed for treatment elsewhere (Ghana Statistical Service 2018). However, fecal sludge treatment facilities and conventional sewer-based treatment systems have largely failed in Ghana, with decentral methods of treatment being more wide-spread (Appiah-Effah et al. 2014). One example used in Ghana is the Biofil-toilet treatment system, yet it has been shown that it is not sufficient as a standalone technology to achieve acceptable effluent levels (Kuwornu 2014). Biogas technology has been applied mainly at the institutional level and mainly to prevent problems from excessive firewood usage like air pollution and deforestation (Arthur et al. 2011, p. 1514). Multiple companies have realized its financial benefit and treat fecal sludge collected by sewage trucks, also sometimes with wastewater reuse of the effluent to mitigate reliance on chemical fertilizers as predicted by Arthur et al. (2011, p. 1516). Co-digestion with various industrial wastes is also performed, however post-treatment infrastructure is still lacking at two observed treatment sites.

Problems of the slow progress in the sanitation sector have been associated with institutional failures like poor working conditions, understaffing of technical personnel and unclear guidelines, as well as social provisions against fecal management. Financially, sanitation is not prioritized compared to drinking water provision. Economic losses due to poor sanitation were found to be higher than what was invested. Donor funding is dwindling and the government now increasingly relies on private sector participation and tariffs. (Appiah-Effah et al. 2019, pp. 406–409)

Economic incentives and environmental framework

The center and wastewater system are built with donor funds. Although decreasing, donor grants still make up 42% of non-household sanitation investments in the Global South (UNICEF and WHO 2020). Therefore, an incentive is given to maximize benefits of the wastewater treatment systems, even though investment costs may be higher. Environmental benefits of non-standard wastewater treatment facilities could also provide economic advantages in the form of corporate social responsibility marketing, which may be applicable in this case as target markets are mainly in the developed world.

The lack of a sewer system, drainage canals, and nearby waterways, as well as limited space availability and poor soil conductivity, facilitate the need for water reuse on production crops planted on the compound. Fertigation, a combination of irrigation and fertilization, is limited by the quantities of crops harvested. To avoid overfertilization without the need for denitrification, either the nutrients of urine need to be extracted separately and applied elsewhere, the treatment sludge cannot be reused or no food waste can be added to the treatment process for enhanced biogas production. Therefore, UDFTs also present the opportunity to gain a concentrated fertilizer product that could be sold to surrounding farmers.

Social factors

Support for wastewater reuse and biogas production is high with only three of 53 sampled people expected to work at facility having reservations against use of biogas and two against wastewater reuse. Four workers were undecided in the two respective cases. The facility is planned to have 200 workers. In 83% of worker households food waste was not utilized. In total, workers were willing to bring on average 1.96 PE worth of food waste from their homes, counting children as 0.5 PE. It is to be investigated how this would influence the choice of treatment system provided. Due to experience with previously used Biofil-toilets, workers can be trusted to use toilets correctly and bins for disposing toilet paper can be used. Anal cleansing with water is not practiced.

Theoretical Background

Urine diverting flush toilets

Urine diverting flush toilets (UDFTs) generally collect urine undiluted or diluted with little flushwater separately from feces, which are flushed as in a conventional toilet. Various toilet designs have been tested with varying degrees of separation, dilution and success (González-Salgado 2017). However, many institutional challenges and problems regarding social acceptance of toilet design have been overcome in over three decades of research, which makes large scale marketing possible (Larsen et al. 2021). So far, UDFTs have largely been used only in sewered areas in Europe, while the Global South has only seen proliferation of urine diverting dry toilets (Larsen et al. 2021).

The main argument provided for use of UDFTs in literature is resource recovery, as urine contains a large amount of excreted nutrients in a concentrated form. This makes them more easily extractable and reusable (Peter-Fröhlich et al. 2007) and reduces nitrogen pollution, mitigating ecological problems like eutrophication (Larsen et al. 2021, p. 1165). It moreover replaces mineral fertilizers without compromises in effectiveness (Peter-Fröhlich et al. 2007; Münch and Winker 2011). This alleviates various environmental, geopolitical and socioeconomic burdens associated with mineral fertilizers (Münch et al. 2009; Larsen et al. 2013).

Another major advantage of UDFTs that is not as often mentioned is their ability to save large amounts of water, up to 80% compared to conventional toilets (González-Salgado 2017). This results from their ability to use no or much less water for urine-only flushes, which are about 5.5 times more common than flushes with feces (Rose et al. 2015). It especially holds true when toilet paper used after urination is disposed of separately and does not need to be flushed (Münch and Winker 2011). The benefits of water saving are generally considered to be hydrological, environmental, associated with the cost of treating toilet water to drinking water standards and water-price-related (Münch and Winker 2011). However, treatment infrastructure costs and emissions are also affected, which will be further investigated in this thesis. While back-end applications for treating urine have been and are further developed, less attention has so far been paid towards application for treatment of diluted feces (Larsen et al. 2021). In a notable exception, Reynaert et al. (2020) have recently developed a toilet system that can independently treat urine, feces and greywater while recovering nutrients and recycling water internally. For treatment of diluted feces, hydrothermal oxidation and an aerated bioreactor have been suggested. It may be suited for the Global South and can solve the problems associated with flush toilets, but chemical and electricity consumption as well as complexity regarding maintenance problems could hamper technology diffusion.

Lastly, exclusion of urine from the main treatment can lead to more favorable conditions for biogas generation, as it shifts the carbon to nitrogen ratio in a favorable direction (Dahlmann 2009).

While resource recovery will be a major argument for UDFTs in the Global North, water saving and its effects on biogas generation further explained in the next section could be the prevailing drivers for proliferation of UDFTs in the Global South. This thesis investigates how influential these respective factors are, so that feasibility of spreading UDFTs in the Global South can be assessed.

The DEWATS approach

Decentralized Wastewater Treatment Systems (DEWATS) is an approach to wastewater treatment specialized for the Global South that was pioneered by the German NGO Bremen Overseas Research and Development Association (BORDA). It is a modular concept that encompasses multiple technologies, which need to be chosen according to local circumstances. It constitutes low-technology, self-manageable combinations of space-efficient, independent technology for small to medium sized communities. The DEWATS components consist of a pretreatment step, which is usually a septic tank or Imhoff tank, followed by an anaerobic baffled reactor (ABR) as a main stage. Post-treatment is usually carried out by ponds, wetlands, or both. Its systems have increasingly gained recognition as easy to operate with low operating costs. (Wissing 2020; Gutterer et al. 2009)

Although DEWATS is a general concept, BORDA has outlined and tested specific approaches for engineering the different system components in a standardized way, which makes comparability between different systems within the scope of this thesis possible. This work refers specifically to the approaches outlined by BORDA (Gutterer et al. 2009).

Anaerobic digestion

The pretreatment units and main stage of the DEWATS system are anaerobic because it can treat high strength wastes without the need for expensive and maintenance-prone aeration equipment (Gutterer et al. 2009). Anaerobic digestion is a natural process that takes place in areas rich in organic substrate where free oxygen is not available or has already been depleted (Murphy and Thamsiriroj 2013). Although utilized since the mid-1800s, with respect to wastewater treatment, it has gained popularity in the last decades mainly due to its ability to produce biogas, which can be used as energy (Kythreotou et al. 2014, p. 701).

There are two basic approaches used with regards to anaerobic treatment. In the first approach, biomass is suspended. The waste stream and degrading biomass have the same retention time in the reactor, meaning the hydraulic retention time (HRT) and the solid retention time (SRT) are equal. In the second approach, biomass is sessile, meaning that SRT is significantly lower than HRT, which allows for smaller treatment elements. (Larsen et al. 2013, p. 326).

Fully mixed fixed-dome digester

The fully mixed digester functions as a continually stirred tank reactor with suspended biomass. Mixing is generally provided by the biogas generated from anaerobic digestion (Larsen et al. 2013, p. 326). There exist different types of fully mixed digesters, but in the Global South mainly fixed-dome digesters (FDDs) and floating-drum digesters are used. FDDs comprise of a main digestion cylinder, a rigid gas holder above and a compensation tank. Gas is retrieved batch-wise, while liquid continuously flows into the digester. As gas is produced, liquid is pressed into the compensation tank. Upon retrieval, the compensation tank empties out and provides the necessary pressure, so that neither energy nor moving parts are required. This makes the system durable and fit for application in the Global South. Figure 1 shows the design of a typical FDD. (Vögeli et al. 2014)

Biogas collection is beneficial because it can replace other fuel sources and their associated environmental and social impacts, which can lead to a reduction in greenhouse gas emissions, women empowerment, health benefits, reforestation, and a reduction in dependencies at the personal and national level (Arthur et al. 2011).

Moreover, collection and subsequent combustion converts the collected methane to carbon dioxide. This has environmental benefits as methane has a 100-year Global Warming Potential 28 times greater than carbon dioxide (Stocker 2014), so if the waste would have otherwise partially undergone aerobic digestion and methane is not collected and burned, biogas production contributes to Global Warming. Carbon dioxide emissions from anaerobic digestion of wastewater do not contribute to Global Warming, as this process only returns what has previously been extracted from the atmosphere. (Vögeli et al. 2014)

The amount of gas collected can be increased with the co-digestion of other wastes, for example food waste. Combining human feces with food waste has been shown increase process stability and gas production by improvement of the buffering capacity, dilution of inhibitory compounds, and facilitation of synergistic effects by providing a more favorable carbon to nitrogen ratio. (Kim et al. 2019)

The addition of food waste also concentrates the digestate, which can justify the use of suspended biomass systems like the fixed dome digester. However, this will necessarily also facilitate more emissions of non-captured methane that will not be converted to carbon dioxide afterwards. Another disadvantage is the large HRT needed, which makes fixed dome digesters big and expensive if substrate is diluted with flushwater (J Behrendt et al. 2006). Although these cause-effect relationships have been known for decades, little emphasis has been placed on their quantification beyond individual case studies (Mathieu Dumont et al. 2013, p. 263). While greenhouse gas emissions from mainly agricultural plants in the Global North have been extensively studied (Paolini et al. 2018), various emission aspects of the digestion process and digestate application are not translatable to treatment infrastructure in the Global South, where emission estimates vary significantly (Bruun et al. 2014).

Imhoff tank and anaerobic baffled reactor

The Imhoff tank is a standard solution in the DEWATS approach that works with sessile biomass, meaning SRT and HRT are decoupled. This allows for a compact design because dilution does not have major effects on the treatment system size. It is used for pretreatment and tries to minimize odors by phase separation. The anaerobic baffled reactor is the main element of DEWATS. It consists of a settler and usually four to six upflow chambers. It also operates with sessile biomass. It can effectively reduce organic matter content without electricity, odors or large footprint. Figure 2 and 3 show general designs. (Gutterer et al. 2009)

Biogas is produced in both systems but not collected as the simple designs would have to cope with leakages and no simple collection mechanism can be used for developing countries due to the dilution and minimal gas production of typically fed substrate (Wissing 2020).

Research questions

Based on the literature review, multiple research gaps have been identified. Effects of UDFTs on treatment infrastructure for diluted feces have not been sufficiently studied, especially for treatment technology of the Global South. Economic and environmental effects of dilution rates on waste treatment in FDDs were only studied on a case-by-case basis for small systems and no theoretical generalization approach was found that could quantify how these effects differ by changes in substrate composition, dilution of the substrate and general size of the system.

According to problems identified in Introduction and Research Gaps the following research questions were formulated:

• What are the most significant changes in treatment structure, the initial investments required and operational benefits due to UDFT introduction in the Global South?

• How are they quantified within the project-framework?

• How do they differ in their economic, environmental and spatial impact?

• How important is the coupling of food waste co-digestion to UDFTs to ensure their applicability?

• How do greenhouse gas emissions in fixed dome digesters change with the respective degree of dilution and co-digestion?

• Could UDFTs be viable option for the Global South in general? How sensitive are the results to differing input parameters?

The final objective of this thesis is to provide data called for by UNICEF and WHO (2020) that can inform decisionmakers to optimally select treatment systems and therefore tackle the current and future issues of sanitation in the Global South.

Methodology

Two treatment scenarios were compared regarding their differences in upfront investment and operational benefits in an economic, environmental and spatial domain. This was done based on the combination of the URWARE model (Jeppsson et al. 2005), DEWATS dimensioning guidelines (Gutterer et al. 2009) and a scalable model of FDD dimensioning and emission modelling. Microsoft Excel was used for modelling of the anaerobic digestion process in the different reactors and for sensitivity analysis regarding key parameters affecting the economic and environmental differences investigated.

In the following, the treatment approach outlined in Figure 4 referred as System 1 is a standard DEWATS approach combined with composting as a treatment for food waste. It serves as a baseline for comparison.

Figure : System 1 configuration

System 2 shown in Figure 5 is the system modified with UDFTs and co-digestion with food waste in a fixed dome digester.

Figure : System 2 configuration

Comparing these two approaches simplifies the analysis, as both approaches satisfy the framework condition that nutrients and water are both fully reused. Therefore, effects of nutrient pollution, synthetic fertilizer, landfill emissions, water resource depletion and drinking water treatment are similar in both systems, which is why they are neglected in this analysis and environmental effects can be reduced to greenhouse gas emissions. This approach does have some limitations discussed in chapter 6 but serves as a useful approximation to reach conclusions regarding the research questions.

The differences in the approaches, namely the difference in primary treatment and the resulting effects on secondary treatment as well as composting and urine storage, were investigated as described in the following sections. The effects of these differences were then quantified over time by estimating the different material quantities needed for the upfront investment as well as the benefits and downfalls arising from the systems. How the different treatment elements are dimensioned is described in their respective sections, but some general assumptions needed for dimensioning and estimation of economic and GHG-related investment costs that are relevant to multiple components are laid out in Tables 1 and 2. Where available, site-specific data was obtained locally via the consultation of construction companies, however, due to the lack of available data, sometimes general reference values with potentially high uncertainty margins had to be used.

Table : Building material properties

Parameter	Unit	Set value	Comment	References
Person-equivalents (PE)	-	120	-	-
Operational hours	h	10	Estimates peak flow	-
Temperature (T)	°C	25	Nighttime average	World weather online (2022)
Exchange rate	USD GHS-1	0.13	Last 5 years falling trend from 0.23 to 0.13	Xe.com Inc. (2021)
LPG price	USD kg-1	1	Fluctuates between 0.25 and 1.3 $ kg-1 in Ghana	Asante et al. (2018)

Table : Toilet Parameters

Parameter	Unit	Set value	Comment	References
Brick dimensions	cm-cm-cm	20-10-10	Local reference value	-
Brick density	kg m-3	1890	Medium brick density	Amin Al-Fakih et al. (2018)
Brick cost	USD	0.09	Local reference value	-
Brick carbon footprint	kgCO2-eq kg-1	0.195	Can vary significantly, for clay fired bricks value can be as high as 2.7 and for sun-dried bricks as low as 0.12	Dabaieh et al. (2020); Kulkarni and Rao (2016)
Mortar between bricks	cm	1	Local reference value	-
Mortar cost	USD m-3	190	Local reference value	-
Mortar carbon footprint	kgCO2-eq m-3	300	207-574	Onn et al. (2019)
Concrete price	USD m-3	260	Local reference value	-
Concrete carbon footprint	kg CO2-eq m-3	420	Value for South Korea, no estimates available for Ghana	Paik and Na (2019)
Rebar in concrete	kg m-3	70	Local reference value	-
Rebar cost	USD kg-1	0.43	Local reference value	-
Rebar carbon footprint	kgCO2-eq kg-1	3.84	Value for South Korea, no estimates available for Ghana	Baek et al. (2016)
Cast iron cover plate thickness	cm	2.5	-	-
Cast iron density	kg m-3	7100	Found to vary between 6969 and 7242 kg m-3	Fatahalla et al.(1996)
Cast iron plate cost	USD	19.5	0.36 m2 plate, local reference value	-
Cast iron carbon footprint	kgCO2-eq kg-1	2.0	1.0-3.9	Lisienko et al. (2016)
Additional piping length System 2	m	150	Pipes for Blackwater and Urine respectively	-
Additional piping cost System 2	USD	700	Pipes for Blackwater and Urine; Local estimate	-
Additional piping carbon footprint System 2	kg CO2-eq m-1	27.1	Pipes for Blackwater and Urine	Baldasano Recio et al. (2005)
Construction cost relative to material cost	-	1.3	Calibrated with local data on 20m3 digester and within upper boundary of factor 2 determined by the reference	J Behrendt et al. (2006)

Piping effects were fixed to the project estimate, which is why they won’t change when input parameters are sensitized. Differences in operation and maintenance (O&M) costs strongly depend on the skill level of operators, the availability of specialized companies, the type of UDFTs used, wage levels and management style. It was not attempted to quantify O&M differences within the scope of this thesis, as this would have required calibration data and a more in-depth socioeconomic analysis. It was assumed that O&M differences are small compared to the factors investigated but further work is needed to properly embed this factor into decision-making.

Toilet characteristics

UDFTs in system 2 are compared against a standard dual-flush toilet (SDFT) in system 1. Along with price differences, Table 3 shows toilet flushwater use for urine and feces. Urination and stool frequencies are assumed to be independent. In the case of system 2, it is assumed that no crossflow of water occurs, meaning flushwater from urinations fully ends up with the urine and flushwater from defecations only goes to the FDD. Standard figures for flushwater consumption and UDFT cost correspond to the only model available in Ghana, which is the Laufen Save! toilet. This toilet does have some crossflow in order to be able to flush toilet paper, however it can be largely reduced by pushing the button only shortly, which is why this crossflow was not considered.

Table : Physical constants regarding LPG substitution

Parameter	Unit	Set value	Comment	References
Flushwater for flushing of feces (UDFT and Standard toilet)	L	6	Depending on UDFT model 2-6	González-Salgado (2017
Stool frequency	d-1	1.10	Can range from 0.74 to 1.97, 1.1 is average value	Rose et al. (2015)
Urination frequency	d-1	6.0	Suggested design value	Rose et al. (2015)
Flushwater diluting Urine (UDFT)	Liters per flush	0.15	Depending on UDFT model 0.2-3; Laufen Save! 0.15	González-Salgado (2017)
Flushwater for flushing of urine (SDFT)	Liters per flush	3		Özel and Baykal (2013)
Price of UDFT	USD	950	Between 346 and 780€; Laufen Save! 950$ (Ghana)	González-Salgado (2017)
Price of standard dual flush toilet (SDFT)	USD	300	Local suppliers charge between 2000 and 4000 GHS (260-520 USD)	-
PE per toilet	-	25	Should not be higher than 25	Gutterer et al. (2009)
Cost of biogas utilization	USD	2200	Gas storage in bags (15m3), desulfurization, dehydration, measurement equipment, based on project data	-

Substrate characterization

Wastewater composition is modelled according to data used by the URWARE model (H Jönsson et al. 2005). The model did not contain volumetric estimates for feces, biowaste and toilet paper. Therefore the density of feces was added with 1.06 mg L^-1 (Brown et al. 1996), food waste density was measured with samples from the case study to be 1.13 mg L^-1 when submerged underwater and toilet paper volume was neglected. However, both of the resulting volumes are small with respect to the other sources, as seen in Figure 6. Any solutes and particulates in toilet water were neglected as well. Figure 6 shows how different wastewater sources contribute to the relevant wastewater parameters.

Figure : Adjusted substrate characterization based on URWARE data (H Jönsson et al. 2005). Q-Discharge, UDFT - Urine diverting flush toilet, SDFT – Standard dual flush toilet, WM – Wet mass, TS – Total Solids, VS – Volatile Solids, COD – Chemical Oxygen Demand, BOD, Biochemical oxygen demand, sbCOD – soluble biodegradable COD, nbsCOD – non biodegradable soluble COD, pbCOD – particulate biodegradable COD, nbpCOD – non biodegradable particulate COD, TP – Total Phosphorus, TN – Total Nitrogen

The URWARE model based its input figures on data from Swedish households, however they generally align with similar data collected by (Rose et al. 2015) and (Larsen et al. 2013, p. xii). Data for the Global South has not been collected broadly regarding some of the parameters such as the fractionation of chemical oxygen demand (COD) used later on. (Rose et al. 2015) identified that the wet mass of fecal matter differs significantly between high- and low-income countries, however as this difference is small compared to the flushwater used, the model was not adjusted. The input fractionation demonstrates that feces, toilet paper, biowaste and toilet flushwater for feces constitute 69% of COD while only occupying 15.8% of volume in the case of SDFTs and 5.0% when UDFTs are used. Compared to a SDFT, the UDFTs can save 70% of water by reducing flushwater consumption for urine.

Digestion and Emission Modelling

Many models for anaerobic digestion have been put forward with varying degrees of complexity. Simpler models often directly estimate the quantity of biogas produced for specific substrates based on volatile solids, rendering them useless for analysis of the digestion process. Complex models like the Anaerobic Digestion Model 1 (ADM1) require many input parameters, of which data are often not or hardly obtainable. (Kythreotou et al. 2014, p. 712)

As complexity is not of much use when input data are generalized, such as in this analysis, and simpler models lack information to perform this analysis due to inabilities of determining effluent quality and gas composition, the medium complexity model URWARE (Jeppsson et al. 2005) was chosen as the main digestion model. However, this model had to be enhanced further to also model gas emissions from FDDs as it lacks the complexity to estimate emissions from compensation tanks as a function of substrate composition, dilution and PE served. Furthermore, the model needed to be merged with the DEWATS design guidelines on dimensioning the different treatment components. As the URWARE model is only valid for fully mixed digesters, the DEWATS model needs to provide necessary information for adjustment regarding the other reactors used.

The URWARE model estimates gas production based on how the different COD fractions are degraded. Using its standard parameters in its mode 1, 95% of the biodegradable soluble COD (bsCOD) and 60% of the particulate biodegradable COD (bpCOD) is converted to methane. Via the combustion equation CH₄ + 2O₂ → CO2 + 2H₂O, it can be determined that CH4 has a COD of nearly 4 kg. CO2 cannot use oxygen as an electron acceptor and therefore does not contain any COD. COD of other trace gases is neglected. This way, CH4 production could be estimated via the input COD fractionation. Using other standard assumptions that set the protein content to 30%, the lipid content to 35% and the fraction of carbohydrates to 35%, CO2 production can be estimated based on CH4 production. The assumptions result in a ratio of 1.47 moles of CH4 per mol of CO2. (Jeppsson et al. 2005)

Emissions and gas collected were then converted using data from Table 4 to estimate climate impact by replacement of LPG. This assumed that gas conversion efficiencies of stoves are independent of the type of gas used.

Table : Fixed-dome parameters

Parameter	Unit	Set value	References
CH4 net calorific value	MJ kg-1	55.6	Dincer et al. (2018)
LPG net calorific value	MJ kg-1	47.3	Dincer et al. (2018)
LPG emission factor	kg CO2-eq GJ-1	63.1	Juhrich
CH4 Global Warming potential (100 yr)	kgCO2-eq kgCH4-1	28	Stocker (2014)

Fixed-Dome Digester

Feces, toilet paper, the toilet water from UDFTs used for feces, and biowaste enter the FDD in system 1. The different input parameters outlined in the previous section were compiled accordingly. Fixed parameters used for calculations regarding the fixed-dome digester are shown in Table 5.

Table : Calculated parameters for fixed-dome modelling

Parameter	Notation	Unit	Set value	Comment	References
HRT	tHRT	d	30	Generally ranges between 15 to 40 days in the Global South depending on Temperature and degree of mixing	Gutterer et al. (2009)
Gas retrieval rate	rretrieval	d-1	1	-	-
Maximum relative gas pressure	prel	Pa	15,000	gives height of compensation chamber above lower end of dome	Vögeli et al. (2014)
Maximum absolute gas pressure	pmax	Pa	116,300	assuming ambient pressure 101,300 PA	-
Average absolute gas pressure	p	Pa	108,800		-
Gas losses due to leaks, incomplete oxidation and gas releases	Lother	-	0.05	Between 0 and 10% for leaks, overall up to 40%	(Bruun et al. 2014)
Length & witdh of connection to compensation tank	lC	m	1	Local reference value	-
Height of connection to compensation tank	hC	m	0.5	Local reference value	-
Henry constant CH4	KH,CH4	mol (m3)-1 Pa-1	1.37494E-05	Temperature adjusted	Sander (2015)
Henry constant CH4	KH, CO2	mol (m3)-1 Pa-1	3.31287E-04
ideal gas constant	R	m3 Pa K-1 mol-1	8.3145
Peak flow factor	PQ	-	2.4	based on assumption of uniform distribution during operating hours

The following system of nonlinear equations was set up to determine the fractions of CH4 and CO2 that leave the system dissolved, are collected in the dome, and leave the system through the compensation tank respectively. The system was solved iteratively. The geometry of the FDD had to be generalized for any input parameters to allow for sensitivity analysis. Volume in the digester for accumulated sludge was not considered as it is assumed to be fully mixed. The bottom of the digester is meant to be flat. Table 6 shows and explains the parameters calculated.

Table : Imhoff tank design parameters

Parameter	Notation	Unit	Parameter	Notation	Unit
Volume of cylinder (main digestion part)	VL = \(t_{HRT}Q\)	m3	Gas fractionation	fCH4 / CO2	-
Volume of dome and compensation tank	VD = VCT	m3	Gas produced	GP,CH4/CO2	mol d-1
Cylinder height	hCylinder	m	Gas in dome	GDome,CH4/CO2	mol d-1
Cylinder radius	rCylinder	m	Gas dissolved	Gdiss, CH4/CO2	mol d-1
Dome height	hDome	m	Gas in compensation tank	GCT, CH4/CO2	mol d-1
Time until compensation tank filled up again after withdrawal	tCT	d	Average fraction of compensation tank occupied	fCT	-

• \(\frac{f_{CH4}}{f_{CO2}} = \frac{G_{Dome,\ CH4}}{G_{Dome,CO2}}\), The equation provides that methane content in the dome equals the fraction of CH4 after losses were accounted for.

• \(G_{Dome,\ CH4/CO2} = \ G_{P,CH4\text{/CO2}} - G_{diss,\ CH4\text{/}CO2} - \ G_{CT,\ CH4\text{/CO2}}\) This represents the mass balance of gas produced, dissolved and escaped through the compensation tank.

• \(t_{CT} = \ \frac{V_{CT}}{QP_{Q} + V_{CT}r}\); \(f_{CT} = 1 - 0.5t_{CT}r\) Assuming withdrawals happen in the morning, these equations estimate how fast the compensation tank fills up due to new wastewater entering the digester and gas produced and therefore what fraction is on average occupied with digesting liquid.

• \(V_{D} = \frac{\left( G_{Dome,CH4} + G_{Dome,CO2} \right)\frac{RT}{p}}{r}\) This calculates the volume of the dome and compensation tank based on withdrawal rate and how much gas is accumulating in the dome according to the ideal gas law.

• \(G_{CT,CH4/CO2} = \ G_{P,CH4/CO2}\frac{f_{CT}V_{CT} + h_{c}l_{c}^{2}}{f_{CT}V_{CT} + h_{c}l_{c}^{2} + V_{L}}\ \)Based on the assumption that the digester is fully mixed, the equation estimates gas losses through the compensation tank depending on how much liquid is on average in it as well as the connector to it.

• \(G_{diss,CH4/CO2} = K_{H,CH4/CO2}Qpf_{CH4/CO2}\) This formula estimates how much gas is dissolved based on Henry’s law.

• \({(h_{Cylinder} + h_{Dome})}^{2} = r_{Dome}^{2} + h_{Cylinder}^{2}\) With this equation the geometry of the digester is determined based on the requirement that the dome can easily be constructed with a rotatable stick placed in the bottom-center of the digester.

• \(V_{L} = \pi r^{2}h_{Cylinder}\) This formula relates the geometry requirement to the design HRT.

• \(V_{Dome} = \ \pi\left( h_{Cylinder} + h_{Dome} \right)h_{Dome}^{2} - \frac{1}{3}\pi h_{Dome}^{3}\) This geometric equation interlinks dome height with dome volume

Effluent parameters were determined according to the URWARE model with 100% of original non-biodegradable soluble COD (nbsCOD) in addition to 5% of bsCOD converted to nbsCOD. Additionally, 100% of non-biodegradable particulate COD (nbpCOD) plus 15% of bpCOD, converted to nbpCOD, is discharged. Finally, 20% of bpCOD exits the system as biomass. The effluent biochemical oxygen demand (BOD) is modelled to be 65% of the effluent bsCOD and bpCOD. (Jeppsson et al. 2005)

For the material quantity estimation, according to local standards, all horizontal parts are made of reinforced concrete with a thickness of 15 cm while vertical walls and the dome are made of brick. Bricks of walls are arranged to have wall thicknesses of 20 cm, while for the dome construction bricks are arranged longitudinally resulting in a thickness of 10 cm. However, to seal the system gastight, an additional 10cm thick layer of mortar is applied to the dome.

Imhoff tank

The Imhoff tank was designed according to standard DEWATS guidelines (Gutterer et al. 2009). Table 8 shows the external input values used, which are default values in the DEWATS guidelines.

Table : Anaerobic baffled reactor design parameters

Parameter	Unit	Set value
HRT	H	1.5
Desludging interval	Yr	1
Inner flow tank width	m	1.30
Space beside flow tank	m	0.55

Feces, toilet paper, urine, greywater and flushwater from SDFTs were added to form the influent according to system 1. The fraction of COD contained by settleable solids needed for this approach was assumed to be equal to the fraction of particulate COD to total COD. The COD/BOD₅ ratio was determined via the URWARE input values to 1.86. The URWARE model however uses BOD₇, but the value obtained is sufficiently close to the suggested standard value of 1.9 that the DEWATS guidelines use for municipal wastewater. Effluent COD was estimated via the empirical DEWATS formula relating HRT and settled COD. The bsCOD and nbsCOD were presumed to pass through the Imhoff tank unaffected and output values of bpCOD and nbpCOD were assumed to have equal settling properties. Both these assumptions are also used in the URWARE model for settling applications. Conversion from particulate to soluble COD during digestion of the settled solids was neglected. BOD5 effluent was determined via the empirical formula provided by DEWATS guidelines relating BOD removal to COD removal depending on the input COD/BOD fraction. The COD producing biogas was estimated to be 60% of settled bpCOD according to URWARE, which amounted to 2.1 kg COD converted to methane.

The geometry of the tank was designed according to the exemplary drawing provided in Chapter 9 of the DEWATS guidelines using the parameters set earlier and adjusting to the calculated sludge height, total inner length and width, sludge height and depth of the outlet based on design results. It can be found in the appendix.

Anaerobic baffled reactor

The anaerobic baffled reactor was designed according to standard DEWATS guidelines (Gutterer et al. 2009). Table 9 shows the external input values used, which are default values in the DEWATS guidelines.

Table : Urine treatment parameters

Parameter	Unit	Set value	Comment
Desludging interval	yr	1	Corresponding to FD
Peak factor	-	2.4	Corresponds to 10 h of operation
Desired effluent BOD	Mg L-1	30	Value set as highest practically achievable based on DEWATS experience
HRT settler	h	1.5	Standard DEWATS value
Depth at outlet	m	1.5
Settler depth	m	1.5
Maximum upflow velocity	m h-1	1.8
Length of one downflow shaft	m	0.25

Substrate input values were calculated according to the determined FDD and Imhoff tank effluents respectively. In scenario 2, greywater was added to the FD effluent. The fraction of COD contained by settleable solids needed for this approach was assumed to be equal to the fraction of particulate COD to total COD as in the calculations for the Imhoff tank.

Firstly, BOD and COD degradation was estimated in the integrated settler using DEWATS guidelines and the parameters listed in Table 9. It was assumed that bsCOD and nbsCOD pass through the settler unaffected and bpCOD and nbpCOD have equal settling properties as suggested by URWARE model assumptions. Moreover, COD used for gas production was calculated as 60% of settled bpCOD.

Secondly, the main part of the ABR was dimensioned with the obtained effluent of the internal settler. This was done with DEWATS guidelines and the values provided. The estimation of COD degraded had to be modified as the guidelines contained an error leading to possible removal rates over 100% and a non-continuous removal function based on the number of chambers added. The case differentiation “less than 7 chambers” was replaced with “less than 4 chambers” to fix this error. Calculations were performed for zero to ten upflow chambers to determine the lowest number of chambers necessary to achieve a certain BOD set as an input value. Both systems can therefore be compared regarding their efficiency since they can be assumed to be similarly effective. Additionally, calculated values were not rounded up as done in the guidelines. Upflow chamber length was set to half the water depth as the highest recommended value.

Degradation of bsCOD was set to equal BOD degradation, but if the BOD value was higher, then the difference was added as bpCOD. COD converted to CH4 in the baffles was assumed to be 95% of bsCOD and 60% of bpCOD as per the URWARE model.

The geometry of the tank was designed in accordance with the exemplary drawing provided in Chapter 9 of the DEWATS guidelines found in the appendix. The drawing was adjusted with design results, namely the calculated inner settler length, the number of chambers, and chamber width. Settler width was chosen to be the upflow chamber width calculated to allow for a more compact design. Wall strength of the brickwork was reduced from 23 cm to 20 cm to meet Ghanaian brick standards and allow comparability towards the FDD. Gas space over the inlet was set to 0.25 m and thickness of the cover was assumed to be as strong as the foundation.

Urine treatment

To be properly reused, urine must undergo some form of treatment to eliminate pathogens. As it is assumed that all nutrients are reused this could be done in the post-treatment of the ABR effluent. However, then it would need to be designed larger, which is not accounted for in this analysis. Moreover, while the main wastewater is designed to irrigate crops for production, some food crops could also be irrigated safely with treated urine. Extended storage of urine has been identified as a simple and effective tool for pathogen control (Münch and Winker 2011). The authors suggested a storage time of 180 days for urine to be safely applied to food crops, although storage time can be reduced significantly if less strict pathogen control is required. Table 10 shows the assumptions from which the additional needed investment was calculated. Values regarding tank dimensions are based on data from local suppliers. Urine was assumed to be stored in polyethylene tanks locally known as polytanks. Space consumption was estimated with tanks assembled in an array with no space in between.

Table : Assumptions regarding composting

Parameter	Unit	Set value
Storage time	d	180
Polyethylene carbon footprint	kg CO2-eq kg-1	3.64
Polyethylene density	kg m-3	930
Polytank cost	USD (m3)-1	117
Polytank height	m	1.5
Polytank radius	m	1.0
Polytank thickness	m	0.01

Greenhouse gas emissions from urine storage are poorly quantified in literature, but simulation data from Udert et al. (2003) revealed that even if all nitrogen lost was escaping as N2O, emissions would still be at least one order of magnitude below the total of other emissions calculated and therefore emissions from storage were not further considered. The carbon footprint of polyethylene has not been estimated for production in Ghana, however Dormer et al. (2013) found that life cycle emission rates were between 1.538 and 3.64 kg_CO2-eq kg_Polyethylene^-1 depending on the fraction of polyethylene recycled. Crawford and Throne (2002) found that polyethylene densities are generally in the range of 925 to 940 kg m^-3.

Composting

In Ghana, food waste is landfilled, used as animal feed, composted, utilized in anaerobic digestion or processed to bio-char briquettes (Galgani et al. 2014). As the comparison between systems is based on the effective reuse of all nutrients, composting of food waste in system 1 is one option that allows for an effective comparison to the food waste utilization in wastewater treatment in system 2. Composting causes significantly more emissions of N₂O, which is why it negatively affects the greenhouse gas balance of system 1. Galgani et al. (2014) investigated processes for food waste composting in Northern Ghana and found that composting additionally causes 0.17 kg_CO2-eq kg^-1, comparing it with other studies putting the value between 0.06 and 0.99. As emissions from anaerobic digestion are already accounted for, absolute emissions from food waste treatment by composting estimated in the study to be 0.19 kg_CO2-eq kg^-1 were used. This could be done on the assumption that N₂O emissions from anaerobic digestion are negligible. It was an assumption taken in the study, although it has been acknowledged that large uncertainties exist for both anaerobic digestion and composting. Fobil et al. (2005) found that the density average for different income levels vary between 410 and 540 kg m^-3, with which emissions could be calculated for the food waste input substrate. Composting time largely depends on the operation and maintenance work invested and can range between twelve weeks and two years (Nardo 2011). Table 10 shows the parameters set for calculation of space and climate impact. Operation and maintenance costs were assumed to be similar between systems and were not included in the analysis. Capital investment and initial carbon footprint were assumed to be negligible.

Parameter	Unit	Set value
Composting time	d	90
Food waste density	kg m-3	475
Compost height	m	1.5
GHG emissions from composting	kg CO2-eq kg-1	0.19

Results

Compared over system 1, system 2 using UDFTs comes with a higher investment cost of 113 USD PE^-1. The most significant factor was identified to be urine storage (32.1%) followed by the higher cost of UDFTs compared to SDFTs (23.9%) and substitution of the Imhoff tank with a FDD (22.2%). The costs due to biogas utilization equipment (16.2%), additional piping needed (5.2%), and ABR enlargement (4.5%) were comparably lower. Regarding greenhouse gas emissions, system 2 had and added carbon footprint of 154 kg_CO2-eq PE^-1, which was mainly attributed to additional piping needed (44.0%) and the FDD replacing the Imhof tank (37.8%). Emissions from construction of a larger ABR (16.7%) and urine treatment (12.1%) were less significant. The additional space consumption excluding gas storage was determined to be 0.28 m² PE^-1 due to urine storage (67.6%), the larger first treatment stage (27.6%), and larger ABR (4.8%).

Use of biogas would constitute 5.8 USD PE yr^-1 in savings. With regards to GHG emissions, 91.4 kg_CO2-eq PE^-1 yr^-1 would be saved by the use of system 2. These savings consist of reduced emissions during treatment (64.4%), avoided emissions by LPG replacement with biogas (18.9%), and reduced emissions from food waste treatment (16.7%).

Accounting for the higher investment but also for the operational benefits provided by system 2, it was determined that financial investment would be returned after 19.5 years in contrast to initially higher GHG emissions being amortized in just 1.7 years. Over a lifespan of 20 years, 1.7 t_CO2-eq PE^‑1 are saved by the use of system 2 instead of system 1. With a hypothetical carbon price of 100 USD t_CO2^-1, this would combine to an amortization time of 8.6 years and 170 USD PE^-1 saved, which scales up to 20400 USD for 120 PE.

These findings are illustrated in the Appendix. Part of the following sections give an overview on how these findings would change with respect to flushwater consumption, food waste dosage, gas losses, PE as well as LPG and carbon prices.

Gas production and emission

Figure 7 shows that 92% of gas produced in the FDD can be captured. Most losses occur through the compensation tank, with dissolved CH4 only playing a minor role. In the ABR, 68% of gas produced could theoretically be captured if more complicated capturing technology is available. 63% of gas produced was modelled to be produced in the FDD but 88% of emissions would be caused by the ABR.

In Figure 8 it can be seen that 34% of emissions can be mitigated by system 2. Gas production was modelled to be 46% higher in system 2. However, only 55% of total gas produced could be recovered. Even if all gas not dissolved could be captured, system 1 would still have emissions 2.2 times higher.

Flushwater consumption

Using the standard assumptions laid out in section 4.1, UDFTs have a flushwater consumption of 6.6 L PE^-1 d^-1 compared to SDFTs with 24.6 L PE^-1 d^-1. The differences in flushwater mainly had effects on the size of the FDD compared over the Imhoff tank, but smaller differences in ABR design were also observed. The settler of the ABR needs to be larger to compensate for the inorganic part of food waste added to the FDD. This effect is directly correlated to flushwater consumption. In low flush scenarios, one additional baffle is needed to achieve the same effluent quality. For flushwater of more than 8 L PE-1 d-1, the additional chamber is no longer needed due to dilution effects.

From an economic point of view, using no flushwater via urine diverting dry toilets (UDDTs) comes with the benefit of 24.2% lower additional capital investment needed. However, profits from LPG replacement would be 35.7% lower. This is because if there is very little or no water, the volume of the compensation tank increases, compared to total digester volume. This leads to a higher percentage of gas generated in the compensation tank that cannot be collected and utilized. Moreover, the additional piping would not be needed, which was accounted for in the values provided above. Regarding GHG emissions, UDFTs could reduce additional initial investment by 57.1%, causing the amount of saved GHG emissions compared to system 1 to drop by 42.3%. GHG amortization time would decrease by 25.6% at the cost of an increase in economic amortization time by 15.2%, which is illustrated in Figure 9. Spatially, UDDTs would result in 19.8% lower additional space consumption.

On the other hand, along with an increased additional space consumption of 33.8%, increasing flushwater consumption leads to more emissions, as more CH4 leaves the system dissolved. However, this effect was found to be negligible compared to the additional investment that is needed due to the over-proportional growth of the FDD compared to the Imhoff tank. UDFTs like the Laufen Save! toilet that allow toilet paper flushing after urination can have a similar flushwater consumption as SDFTs, especially at the trade center where mainly women are employed. Collecting toilet paper separately and hence saving flushwater was therefore found to reduce additional initial capital investment by 16.8% and additional initial GHG emissions by 22.8%. However, it was found that operational savings would be reduced by 5.1% economically and 7.2% regarding greenhouse gas emissions, as the ratio of compensation tank volume to total digester volume would increase. Combined, this resulted in amortization times getting reduced 12.2% and 16.7% respectively. UDFTs therefore have a flushwater consumption that is ideal with regards to amortization time, while UDDTs are most favorable towards investment costs and high flushwater consuming toilets have the most operational benefits, provided that all the water can be effectively reused.

Figures in the appendix A illustrate these findings in more detail and show how the relative influences of single parameters differ with different flushwater consumption.

Food waste dosage

As workers largely do not utilize food waste at home, it would be possible to bring household waste to the production site, increasing the amount of food waste in the digester. Dosage ratio is defined as PE of food waste per PE of wastewater producers. Assuming 1 PE of food waste can additionally be provided by the average worker and 200 workers are employed, 320 PE of food waste would be treated. This equates to a dosage ratio of 2.67, which would require an increased additional capital investment of 3.9%, an upfront GHG emission increase of 4.4%, and an additional space consumption of 2.0%. These increases can be mainly attributed to enhanced gas production and the subsequent need to enlarge the FDD, with minor contributions of the actual food waste volume added to the digester and ABR enlargement. Profit due to biogas collection would increase by 73.4%, but emissions would also increase by 30.3%. Although larger quantities of methane are emitted via the compensation tank, this is partially made up for by saved emissions from LPG replacement and from not composting. As a result, economic amortization could be reduced by 40.0% at the expense of increasing amortization time of GHG emissions by 49.9%. The net present value for 20 years after commissioning would increase from 2.7 USD PE^-1 to 83.4 USD PE^-1, which translates to 6730 USD. However, at the same time only 1113 kgCO2-eq PE^-1 could be mitigated compared to system 1, which constitutes a reduction of 33.5%.

On the contrary, not using co-digestion would severely limit chances of project implementation because investment costs would only pay off after 33.7 years. However, GHG emissions would amortize already after 1.4 years compared to 1.6 years for a dosing ratio of 1.0, mainly due to mitigated CH4 emissions by biogas usage.

As shown in Figure 10, in case of dosage ratios of five or more, operational GHG emissions of system 2 are higher than for system 1. However, using Figure 11, the combined amortization time using a carbon price of 100 USD tCO2-eq-1 stays relatively constant around 10 years, indicating that should this carbon price come into effect in the future the dosage ratio would not be important.

Gas losses

The fraction of collectable gas that is not combusted and released into the atmosphere due to leakages, lack of flaring and incomplete combustion does have a significant impact on operational cost benefits and especially GHG emission. A fraction of 5% was set as the default value for this project, but digesters are known to operate across the whole spectrum in the Global South. A maximum of 6.1 USD PE-1 can be achieved with the value declining linearly with increasing losses. For fractions of 46% of avoidable gas emissions, methane emissions of the FDD and ABR in system 2 outweigh those in system 1. Above 61%, net emissions from system 2 are higher than for system 1, which is also illustrated by the amortization time in Figure 12. In that case, additional emissions are greater than the carbon mitigation by avoidance of LPG and composting

LPG and Carbon Price

Ghana has comparably medium to high LPG prices at around 1 USD kg-1, but price volatility is high (Asante et al. 2018, p. 98). Figures 13 and 14 show that even small changes in gas price have significant effects on economic amortization time.

Carbon pricing currently ranges between 0 and 140 USD tCO2-eq-1 (Parry et al. 2021), however a combination of physical and economic models by van der Ploeg and Rezai (2021) showed that the optimal carbon price may be as high as 2.039 USD kgCO2-eq-1 if losses in productivity due to climate change impacts are taken into account. Figure 15 shows that carbon prices as high as this would essentially negate the high economic return periods. Interrelations between LPG and carbon price were not investigated.

Person-equivalents

Smaller systems have to deal with higher per capita investment costs and get lower operating profits, as a higher fraction of gas is lost through the compensation tank. This is because in small systems the volume of the connector between main structure and compensation tank also plays a non-negligible role in emissions (Hamad et al. 1983). Only systems with 115 PE or more have economic amortization times of less than 20 years. However, with carbon prices of 0.1 USD kgCO2-eq-1 even systems as small as 30 PE would have 20-year amortization times. The Appendix shows the relevant figures outlining how the discussed parameters apply to systems of different sizes.

Discussion

Contributions to practice have been made towards the quantification of economic feasibility and climate impact of UDFTs in the context of FDDs as a treatment technology. A model was developed that can inform on optimal treatment selection given certain socioeconomic framework conditions like water reuse. Theoretical contributions were put forward in the form of gas emission modelling in FDDs and how this is affected by dilution.

Result validation

Recent studies regarding gas emissions from fixed dome digesters are not available. Khoiyangbam et al. (2004) measured GHG emissions from small size biogas digesters treating cattle dung and found emissions to be between 43.1 and 83.1 g_CH4 m^-3 d^-1, corresponding to 5-8% of CH₄ production. The digester for the trade center was determined to have a size of 28.8 m³ and emissions of 1.23 kg_CH4 d^-1, which translates to 42.7g_CH4 m^-3 d^-1. This estimate being in the lower end would confirm model predictions of larger digesters emitting less GHGs, however comparability is limited because of the different substrates used. Hamad et al. (1983) investigated how digester pressure affects gas losses and set up a similar model that estimated the amount of slurry in the compensation tank with its subsequent emissions. Depending on digester pressure, between 5.8% and 14.5% of CH₄ was lost, which is in accordance with the 8% estimated for the trade center. In this case, the digester had slightly higher pressure but its geometry was optimized to cause less emissions in the compensation tank. Empirical measurements would be needed for validation of the model. It also remains to be investigated how conditions of less-than-ideal mixing would affect results and how this could be used as a calibration parameter.

Bruun et al. (2014) did a theoretical assessment on GHG emissions from small scale fixed dome and floating drum digesters. They calculated the break-even point where the GHG benefits of replacing various fuels are set off by emissions. For LPG, they found that gas losses of 16% or more would already make biogas technology a net-emitter, which could in total lead to 1% of global emissions for an assumed maximum of 40% of gas not used. The study did however not account for external emissions that would occur if the different wastes were treated in a different way. This explains the difference to the break-even point of 61% of gas losses obtained by this thesis, which also factored in emissions from food waste composting and wastewater treatment in anaerobic systems without biogas collection but without food waste, as they emit significantly more GHGs than open defecation (Shaw et al. 2021).

Model limitations

The approach outlined in the thesis faces different limitations in a technical, empirical and socioeconomic domain. It is based on simplified models that were not designed to be fully interoperable and may not be equally applicable to the different wastewater compositions at the various treatment stages. Input parameters fluctuate on a daily basis and some input parameters like emissions from brick production have multiple orders of magnitude uncertainty. Furthermore, externalities like economic feedback loops or political drivers like decreasing reliance on imported fertilizers or enabling independent energy supplies to individuals cannot be accurately quantified. As a main constraint, effects could only be quantified in a relative system comparison, which excludes comparability with absolute costs and benefits. Other treatment approaches and basis values like the benefit of providing sanitation at all were therefore not evaluated. The models used did not account for effects like organic underloading (Reynaud 2014), changes of the carbon to nitrogen ratio (Kim et al. 2019), or moisture content of the digester (Fujishima et al. 2000), all of which are influenced by the quantity of flushwater and food waste added, and were shown to have effects on treatment performance.

Full reuse of water and nutrients was used as an assumption, but this may not be realistic in many cases. The different treatment approaches lead to unequal fertilization potentials, as ammonia is volatilized in urine storage (Etter et al. 2011) and plant growth is different based on the type of fertilizer applied (Cai et al. 2017). N₂O emissions also differ significantly with the type of fertilizer applied, as human urine only causes about 1/3 of the N₂O emissions the application of composted household waste causes (Gómez-Muñoz et al. 2017).

Wissing (2020) provided a full life cycle assessment for a typical DEWATS system. He found that reducing flushwater volume by 18% could reduce the overall impact score by a factor of 10 and mainly attributed the cost of potable water production to it. He showed that categories like marine aquatic ecotoxicity and eutrophication contribute more significantly to the ecological impacts than global warming. Regarding global warming, emissions during operation constituted over 95% of total emissions, mainly due to CH₄ production in the primary treatment and the ABR. It remains to be investigated how these other impact categories change if UDFTs and FDDs are used, even when all water is reused properly.

Applicability in Ghana and the Global South

If UDFTs can diffuse into markets depends on many more technical, social and political circumstances. People need to be willing to use urine diversion components and biogas (Arthur et al. 2011), biogas use can also present safety hazards (Vögeli et al. 2014), and the digesters need to be constructable with the right soil conditions and skilled labor available (Monvois et al. 2010).

As it is questionable if higher investments can be justifiable, institutional developments will need to evolve. Currently, sanitation budgets in developing countries are typically less than 2 USD PE^-1. However, with the rise of Private Public Partnerships and microfinance, there is the prospect of additional funds that could be mobilized. This funding could be especially effective with financial management tools that can ensure these funds can be paid back later on, for example by economic savings facilitated by the wastewater system. (UNICEF and WHO 2020)

Conclusion and outlook

Urine diverting flush toilets (UDFTs) combined with FDDs (Fixed dome digesters) and co-digestion with food waste can be a powerful tool to mitigate greenhouse gas emissions in the sanitation sector of the Global South, primarily due the possibility to collect biogas produced by wastewater treatment with robust, simple and proven technology, without relying on often undesired UDDTs. Problems of flush toilets in general can be mitigated by water reuse, so that UDFTs can be utilized as a potentially more socially acceptable tool that allows for less emissions and more biogas collection than UDDTs could. Economic benefits of biogas production are appealing but the comparatively high additional investment cost presents a barrier that could be weakened by either considering carbon costs or increasing the amount of food waste digested. Funding tools like private-public partnerships as well as donor funding could be used, as confidence that the systems will last is increased due to the benefit of biogas collection without failure-prone components. A certain degree of gas leaking or not being used can be tolerated, but in extreme cases more greenhouse gases could be released by the system with UDFTs and a FDD. This would also be the case if the food waste ratio is excessively increased.

UDFTs have a major potential to diffuse into markets in the Global South, especially when its GHG mitigation potential is taken into account and solutions for dealing with the higher investment costs are successful. However, among many factors, subsidies on natural gas, social provisions and lack of funding opportunities can significantly hamper this potential. Many simplifications had to be made in this model and it remains to be investigated how other externalities like different reuse strategies, sanitation policies and economic growth in the Global South can affect decision-making. Baseline values like the benefits of providing sanitation at all and the benefits of water reuse need to be incorporated into the model to investigate the relative effect of UDFT contributions on overall sanitation cost benefits. A full life cycle assessment of all the different sanitation tools needs to be conducted to allow for a modular tool that can, together with economic models, investigate how large the future market for UDFTs can be. Case studies will need to validate the model and check assumptions. As the global need to mitigate climate change and resource depletion increases, and economies in the Global South develop, governments, individuals, and donors will need to have access to reliable data that allows for sustainable decision-making regarding sanitation-related issues. Otherwise, UDFTs may not be able to live up to their potential.

References

Amin Al-Fakih; Bashar Mohammed; M.s. Liew; Wesam Alaloul (2018): Physical properties of the rubberized interlocking masonry brick. In International Journal of Civil Engineering and Technology 9 (6), pp. 656–664. Available online at https://www.researchgate.net/profile/amin-al-fakih/publication/326317521_physical_properties_of_the_rubberized_interlocking_masonry_brick.

Appiah-Effah, Eugene; Duku, Godwin Armstrong; Azangbego, Naziru Yakubu; Aggrey, Ransford Kojo Aduafo; Gyapong-Korsah, Barbara; Nyarko, Kwabena Biritwum (2019): Ghana's post-MDGs sanitation situation: an overview. In Journal of Water, Sanitation and Hygiene for Development 9 (3), pp. 397–415. DOI: 10.2166/washdev.2019.031.

Appiah-Effah, Eugene; Nyarko, Kwabena Biritwum; Gyasi, Samuel Fosu; Awuah, Esi (2014): Faecal sludge management in low income areas: a case study of three districts in the Ashanti region of Ghana. In Journal of Water, Sanitation and Hygiene for Development 4 (2), pp. 189–199. DOI: 10.2166/washdev.2014.126.

Arthur, Richard; Baidoo, Martina Francisca; Antwi, Edward (2011): Biogas as a potential renewable energy source: A Ghanaian case study. In Renewable Energy 36 (5). DOI: 10.1016/j.renene.2010.11.012.

Asante, Kwaku Poku; Afari-Asiedu, Samuel; Abdulai, Martha Ali; Dalaba, Maxwell Ayindenaba; Carrión, Daniel; Dickinson, Katherine L. et al. (2018): Ghana's Rural Liquefied Petroleum Gas Program Scale Up: A case study. In Energy for Sustainable Development 46, pp. 94–102. DOI: 10.1016/j.esd.2018.06.010.

Baek, Cheonghoon; Tae, Sungho; Kim, Rakhyun; Shin, Sungwoo (2016): Life Cycle CO2 Assessment by Block Type Changes of Apartment Housing. In Sustainability 8 (8), p. 752. DOI: 10.3390/su8080752.

Baldasano Recio, José María; Guerrero, Pedro Jiménez; Ageitos, María Gonçalves; Narváez, René Parra (2005): Estimate of energy consumption and CO2 emission associated with the production, use and final disposal of PVC, HDPE, PP, ductile iron and concrete pipes. Available online at https://pipa.com.au/wp-content/uploads/2021/05/estima1.pdf.

Brown, D. M.; Butler, D.; Orman, N. R.; Davies, J. W. (1996): Gross solids transport in small diameter sewers. In Water Science and Technology 33 (9). DOI: 10.1016/0273-1223(96)00366-6.

Bruun, Sander; Jensen, Lars Stoumann; van Khanh Vu, Thi; Sommer, Sven (2014): Small-scale household biogas digesters: An option for global warming mitigation or a potential climate bomb? In Renewable and Sustainable Energy Reviews 33, pp. 736–741. DOI: 10.1016/j.rser.2014.02.033.

Cai, Yanjiang; Du, Ziyin; Yan, Yan; Wang, Xiaodan; Liu, Xiuping (2017): Greater stimulation of greenhouse gas emissions by stored yak urine than urea in an alpine steppe soil from the Qinghai-Tibetan Plateau: A laboratory study. In Grassl Sci 63 (3), pp. 196–207. DOI: 10.1111/grs.12164.

Crawford, Roy J.; Throne, James L. (2002): ROTATIONAL MOLDING POLYMERS. In R. J. Crawford, James L. Throne (Eds.): Rotational molding technology. Norwich, N.Y.: Plastics Design Library/William Andrew Pub, pp. 19–68.

Dabaieh, Marwa; Heinonen, Jukka; El-Mahdy, Deena; Hassan, Dalya M. (2020): A comparative study of life cycle carbon emissions and embodied energy between sun-dried bricks and fired clay bricks. In Journal of Cleaner Production 275, p. 122998. DOI: 10.1016/j.jclepro.2020.122998.

Dahlmann, Kristina (2009): Modeling sanitation scenarios in developing countries. Available online at http://www.w-program.nu/filer/exjobb/kristina_dahlman.pdf.

DEWATS Chapter 9 Technical Components (2009).

Dincer, Ibrahim; Rosen, Marc A.; Khalid, Farrukh (2018): 3.16 Thermal Energy Production. In : Comprehensive energy systems. San Diego: Elsevier Science, pp. 673–706.

Dormer, Aaron; Finn, Donal P.; Ward, Patrick; Cullen, John (2013): Carbon footprint analysis in plastics manufacturing. In Journal of Cleaner Production 51, pp. 133–141. DOI: 10.1016/j.jclepro.2013.01.014.

Etter, B.; Tilley, E.; Khadka, R.; Udert, K. M. (2011): Low-cost struvite production using source-separated urine in Nepal. In Water Research 45 (2), pp. 852–862. DOI: 10.1016/j.watres.2010.10.007.

Fatahalla, Nabil; Abd Al Hakim, Hani; Abo-El-Ezz, Aly; Mohamed, Moenes (1996): Effect of the percentage carbon equivalent on the nodule characteristics, density and modulus of elasticity of ductile cast iron. In JOURNAL OF MATERIALS SCIENCE 31 (18), pp. 4933–4937. DOI: 10.1007/BF00355883.

Fobil, Julius N.; Carboo, Derick; Armah, Nathaniel A. (2005): Evaluation of municipal solid wastes (MSW) for utilisation in energy production in developing countries. In IJETM 5 (1), p. 76. DOI: 10.1504/IJETM.2005.006508.

Fujishima, S.; Miyahara, T.; Noike, T. (2000): Effect of moisture content on anaerobic digestion of dewatered sludge: ammonia inhibition to carbohydrate removal and methane production. In Water Science and Technology 41 (3), pp. 119–127. DOI: 10.2166/wst.2000.0063.

Galgani, Pietro; van der Voet, Ester; Korevaar, Gijsbert (2014): Composting, anaerobic digestion and biochar production in Ghana. Environmental-economic assessment in the context of voluntary carbon markets. In Waste management (New York, N.Y.) 34 (12), pp. 2454–2465. DOI: 10.1016/j.wasman.2014.07.027.

Ghana Statistical Service (2018): Multiple Indicator Cluster Survey (MICS2017/18), Survey Findings Report. Accra, Ghana.

Gómez-Muñoz, Beatriz; Magid, Jakob; Jensen, Lars Stoumann (2017): Nitrogen turnover, crop use efficiency and soil fertility in a long-term field experiment amended with different qualities of urban and agricultural waste. In Agriculture, Ecosystems & Environment 240, pp. 300–313. DOI: 10.1016/j.agee.2017.01.030.

González-Salgado, Irene (2017): Optimization and monitoring of two pilots plants: a mesophilic anaerobic digester and urine diversion toilets in a context of source separation wastewater management. Available online at https://uvadoc.uva.es/handle/10324/25390.

Gutterer, Bernd; Sasse, Ludwig; Panzerbieter, Thilo; Reckerzügel, Thorsten (2009): Decentralised wastewater treatment systems (DEWATS) and sanitation in developing countries. A practical guide: Water, Engineering and Development Centre. Available online at https://www.pseau.org/outils/ouvrages/borda_wedc_decentralised_wastewater_treatment_systems_dewats_and_sanitation_in_developing_countries_2009.pdf.

H Jönsson; A Baky; Ulf Jeppsson; D Hellström; E Kärrman (2005): Composition of Urine, Feaces, Greywater and Biowaste for Utilisation in the URWARE Model. Available online at https://portal.research.lu.se/en/publications/composition-of-urine-feaces-greywater-and-biowaste-for-utilisatio.

Hamad, M. A.; Abdel Dayem; A. M., El-Halwagi (1983): Rural biogas technology: Effect of digester pressure on gas rate and composition. Available online at https://www.solarcities.eu/sites/solarcities.eu/files/documents/rural_biogas_technology.pdf.

J Behrendt; St Deegener; F Meinzinger; D R Gajurel; R Otterpohl (2006): Appropriate de-central wastewater technologies for low income regions. Available online at https://www.researchgate.net/profile/claudia-wendland/publication/228661366_appropriate_de-central_wastewater_technologies_for_low_income_regions.

Jeppsson, Ulf; Baky, A.; Hellström, D.; Jönsson, H.; Kärrman, E. (2005): The URWARE Wastewater: treatment Plant Models. Chalmers University of Technology. Available online at https://portal.research.lu.se/files/4541059/588063.pdf.

Juhrich, Kristina: CO2 Emission Factors for Fossil Fuels. Umweltbundesamt.

Khoiyangbam, R. S.; Kumar, Sushil; Jain, M. C.; Gupta, Navindu; Kumar, Arun; Kumar, Vinod (2004): Methane emission from fixed dome biogas plants in hilly and plain regions of northern India. In Bioresource technology 95 (1), pp. 35–39. DOI: 10.1016/j.biortech.2004.02.009.

Kim, Jaai; Kim, Jinsu; Lee, Changsoo (2019): Anaerobic co-digestion of food waste, human feces, and toilet paper: Methane potential and synergistic effect. In Fuel 248, pp. 189–195. DOI: 10.1016/j.fuel.2019.03.081.

Kulkarni, Nitin G.; Rao, Anand B. (2016): Carbon footprint of solid clay bricks fired in clamps of India. In Journal of Cleaner Production 135, pp. 1396–1406. DOI: 10.1016/j.jclepro.2016.06.152.

Kuwornu, Isaac (2014): Assessing the performance of the filtering unit of the biofil toilet system. Master thesis. Available online at https://saniup.org/wp-content/uploads/2018/01/isaac-msc-thesis-4-may-2014.pdf.

Kythreotou, Nicoletta; Florides, Georgios; Tassou, Savvas A. (2014): A review of simple to scientific models for anaerobic digestion. In Renewable Energy 71, pp. 701–714. DOI: 10.1016/j.renene.2014.05.055.

Larsen, Tove; Udert, Kai; Lienert, Judit (2013): Source Separation and Decentralization for Wastewater Management: IWA Publishing. Available online at https://library.oapen.org/handle/20.500.12657/24364.

Larsen, Tove A.; Gruendl, Harald; Binz, Christian (2021): The potential contribution of urine source separation to the SDG agenda – a review of the progress so far and future development options. In Environ. Sci.: Water Res. Technol. 7 (7), pp. 1161–1176. DOI: 10.1039/D0EW01064B.

Lisienko, V. G.; Chesnokov, Yu N.; Lapteva, A. V.; Noskov, V. Yu (2016): Types of greenhouse gas emissions in the production of cast iron and steel. In IOP Conf. Ser.: Mater. Sci. Eng. 150, p. 12023. DOI: 10.1088/1757-899X/150/1/012023.

Mathieu Dumont, N. L.; Luning, Luchien; Yildiz, Ismail; Koop, Klaas (2013): Methane emissions in biogas production. In The Biogas Handbook, pp. 248–266. DOI: 10.1533/9780857097415.2.248.

Monvois, Jacques; Gabert, Julien; Frenoux, Clement; Guillaume, Marie (2010): How to select appropriate technical solutions for sanitation.

Münch, E.; Dahm; P. (2009): Waterless urinals: a proposal to save water and recover urine nutrients in Africa. Available online at https://sswm.info/sites/default/files/reference_attachments/muench%202009%20waterless%20urinals%20%20proposal%20to%20save%20water%20and%20recover%20nutrients%20in%20africa.pdf.

Münch, Elisabeth von; Winker, Martina (2011): Technology review of urine diversion components. Overview of urine diversion components such as waterless urinals, urine diversion toilets, urine storage and reuse systems. Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ) GmbH. Eschborn, Germany.

Murphy, Jerry D.; Thamsiriroj, Thanasit (2013): Fundamental science and engineering of the anaerobic digestion process for biogas production. In The Biogas Handbook, pp. 104–130. DOI: 10.1533/9780857097415.1.104.

Narain, Sunita (2002): The flush toilet is ecologically mindless. Available online at https://waterfund.go.ke/watersource/downloads/016.%20the%20flush%20toilet.pdf.

Nardo, Federico de (2011): Composting from A to Z.

Onn, Chiu Chuen; Mo, Kim Hung; Radwan, Mohammed K. H.; Liew, Wen Hong; Ng, Chee Guan; Yusoff, Sumiani (2019): Strength, Carbon Footprint and Cost Considerations of Mortar Blends with High Volume Ground Granulated Blast Furnace Slag. In Sustainability 11 (24), p. 7194. DOI: 10.3390/su11247194.

Özel, B. D.; Baykal, B. B. (2013): Awareness raising and educated activities for effective water savings: a case study with dual flush toilets. In WIT Transactions on Ecology and the Environment 178, pp. 255–261. DOI: 10.2495/WS130211.

Paik, Inkwan; Na, Seunguk (2019): Comparison of Carbon Dioxide Emissions of the Ordinary Reinforced Concrete Slab and the Voided Slab System During the Construction Phase: A Case Study of a Residential Building in South Korea. In Sustainability 11 (13), p. 3571. DOI: 10.3390/su11133571.

Paolini, Valerio; Petracchini, Francesco; Segreto, Marco; Tomassetti, Laura; Naja, Nour; Cecinato, Angelo (2018): Environmental impact of biogas: A short review of current knowledge. In Journal of environmental science and health. Part A, Toxic/hazardous substances & environmental engineering 53 (10), pp. 899–906. DOI: 10.1080/10934529.2018.1459076.

Parry, Ian; Dohlman, Peter; Hillier, Cory; Kaufman, Martin; Florian Misch; James Roaf; Waerzeggers, Christophe (2021): Carbon Pricing: What Role for Border Carbon Adjustments? International Monetary Fund. Available online at https://policycommons.net/artifacts/1819301/carbon-pricing/2556762/.

Peter-Fröhlich, Anton; Bonhomme, Alexandre; Oldenburg, Martin (2007): Sanitation Concepts for Separate Treatment of Urine, Faeces and Greywater (SCST)-Results. With assistance of Regina Gnirß, Boris Lesjean. Berliner Wasserbetriebe. Available online at https://sswm.info/sites/default/files/reference_attachments/peter%20froehlich%202007%20sanitation%20concepts%20for%20separate%20treatment%20of%20urine%20faeces%20and%20greywater.pdf.

Reynaert, Eva; Greenwood, Esther E.; Ndwandwe, Bonginkosi; Riechmann, Michel E.; Sindall, Rebecca C.; Udert, Kai M.; Morgenroth, Eberhard (2020): Practical implementation of true on-site water recycling systems for hand washing and toilet flushing. In Water research X 7, p. 100051. DOI: 10.1016/j.wroa.2020.100051.

Reynaud, Nicolas (2014): Operation of Decentralised Wastewater Treatment Systems (DEWATS) under tropical field conditions. Available online at https://core.ac.uk/download/pdf/236372775.pdf.

Rose, C.; Parker, A.; Jefferson, B.; Cartmell, E. (2015): The Characterization of Feces and Urine: A Review of the Literature to Inform Advanced Treatment Technology. In Critical Reviews in Environmental Science and Technology 45 (17), pp. 1827–1879. DOI: 10.1080/10643389.2014.1000761.

Sander, R. (2015): Compilation of Henry's law constants (version 4.0) for water as solvent. In Atmos. Chem. Phys. 15 (8), pp. 4399–4981. DOI: 10.5194/acp-15-4399-2015.

Shaw, Kelsey; Kennedy, Christopher; Dorea, Caetano C. (2021): Non-Sewered Sanitation Systems’ Global Greenhouse Gas Emissions: Balancing Sustainable Development Goal Tradeoffs to End Open Defecation. In Sustainability 13 (21), p. 11884. DOI: 10.3390/su132111884.

Stocker, Thomas (2014): Climate Change 2013. The physical science basis. New York: Cambridge University Press.

Udert, Kai M.; Larsen, Tove A.; Biebow, Martin; Gujer, Willi (2003): Urea hydrolysis and precipitation dynamics in a urine-collecting system. In Water Research 37 (11), pp. 2571–2582. DOI: 10.1016/S0043-1354(03)00065-4.

UNICEF; WHO (2020): State of the world’s sanitation: UNICEF: United Nations Children's Fund. Available online at https://policycommons.net/artifacts/1427024/state-of-the-worlds-sanitation/2041623/.

van der Ploeg, Frederick; Rezai, Armon (2021): Optimal carbon pricing in general equilibrium: Temperature caps and stranded assets in an extended annual DSGE model. In Journal of Environmental Economics and Management 110, p. 102522. DOI: 10.1016/j.jeem.2021.102522.

Vögeli, Yvonne; Lohri, Christian Riuji; Gallardo, Amalia; Diener, Stefan; Zurbrügg, Christian (2014): Anaerobic digestion of biowaste in developing countries. Rractical information and case studies.

Wissing, Gareth Christian. (2020): A life cycle analysis for the application of decentralised sanitation technologies at Newlands Mashu in the Ethekwini Municipality. Available online at https://ukzn-dspace.ukzn.ac.za/handle/10413/19875.

World Weather Online (2022): Kpong Climate Weather Averages, updated on 2022, checked on 4/13/2022.

Xe.com Inc. (2021): Xe Currency Converter, updated on 4/13/2021, checked on 4/13/2021.

Appendix A – Design drawings

Figure : Imhoff tank design figures (DEWATS Chapter 9 Technical Components 2009)

Figure : Anaerobic baffled reactor design figures (DEWATS Chapter 9 Technical Components 2009)

Appendix B – Contributions of different construction materials

Appendix Figure : Economic contributons of different construction materials

Appendix Figure : Greenhouse gas contributions of different construction materials

Appendix C.1 – Flushwater effects

All figures give only relative values on system 2 compared over system 1 as described in the main part of the thesis.

Appendix Figure : Flushwater effects on amortization time

Appendix Figure : Flushwater effects on additional capital investment

Appendix Figure : Flushwater effects on economic savings

Appendix Figure : Flushwater effects on initial greenhouse gas emissions

Appendix Figure : Flushwater effects on greenhouse gas savings

Appendix Figure : Flushwater effects on space consumption

Appendix C.2 – Effects of food waste dosage

All figures give only relative values on system 2 compared over system 1 as described in the main part of the thesis.

Appendix Figure : Food waste effects on amortization time

Appendix Figure : Food waste effects on additional initial investment

Appendix Figure : Food waste effects on economic savings

Appendix Figure : Food waste effects on initial greenhouse gas emissions

Appendix Figure : Food waste effects on greenhouse gas savings

Appendix Figure : Food waste effects

Appendix C.3 – Effects of gas losses

All figures give only relative values on system 2 compared over system 1 as described in the main part of the thesis.

Appendix Figure : Effect of gas losses on amortization time

Appendix Figure : Effect of gas losses on economic savings

Appendix Figure : Effect of gas losses on greenhouse gas savings

Appendix C.4 – Effects of general system size

All figures give only relative values on system 2 compared over system 1 as described in the main part of the thesis.

Appendix Figure : Effects of person-equivalents on amortization time

Appendix Figure : Effects of person-equivalents on economic savings

Appendix Figure : Effects of person-equivalents on additional economic investment

Appendix Figure : Effects of person-equivalents on initial greenhouse gas emissions

Appendix Figure : Effects of person-equivalents on greenhouse gas savings

Appendix Figure : Effects of person-equivalents on space consumption

Affidavit

I hereby declare that I have developed and written the enclosed bachelor’s thesis entirely on my own and have not used outside sources without declaration in the text. Any concepts or quotations applicable to these sources are clearly attributed to them. This bachelor thesis has not been submitted in the same or a substantially similar version, not even in part, to any other authority for grading and has not been published elsewhere. I am aware of the fact that a misstatement may have serious legal consequences. I also agree that my thesis can be sent and stored anonymously for plagiarism purposes.

Munich, June 12^th, 2022

Felix Giglberger