Preference for and characteristics of an appropriate sanitation technology for the slums of Kisumu, Kenya

ABSTRACT

Slums are faced with several challenges including lack of sanitation facilities. This lack is further compounded by the difficulty in determining an appropriate sanitation technology due to the challenging living conditions. An appropriate technology should, among others, consider the conditions of each settlement as well as views of the slum residents. Through interviews and participatory discussions, residents of the slums in Kisumu city identified sanitation technologies they preferred, and the reasons for or against these technologies. These responses were pointers to characteristics of an appropriate technology for the slums. Pit latrines, ecological sanitation and flush toilets were identified, but residents selected pit latrines as appropriate for their settlements. The results mean that technically, an ideal technology should be easy to construct and use, be socioculturally acceptable, environmentally friendly and economically affordable. This article calls for cooperation among stakeholders, including landlords and tenants in decision-making for sanitation in slums.

KEYWORDS:

• EcoSan
• landlord
• participatory
• Pit latrine
• sanitation decision-making
• sanitation preference
• tenant

Introduction

Slum settlements are faced with several challenges, including lack of sanitation facilities, and the consequent use of improper methods of human waste disposal such as open defecation. Provision of sanitation facilities in urban slums is, however, complex due to a number of factors such as insecure land tenure that discourages investment in sanitation, lack of clear-cut responsibilities among the many stakeholders involved in urban slum development, differences in residents’ social backgrounds which limit social harmony and difficulty in determining an appropriate sanitation technology (Lüthi et al. 2010). In an effort to determine an appropriate sanitation technology, a number of authors have shed light on characteristics of a sanitation technology that would be appropriate in poor urban areas. Some highlight that for low-income urban areas in Africa, sanitation technologies need to be socioculturally acceptable to users, economically affordable and environmentally friendly (Mara 2003; Semiyaga et al. 2015). Others highlight factors such as availability of space, effect on human health, feasibility in the local area, water availability, institutional arrangements and possibility of nutrient reuse, as being important indicators (Mara et al. 2007; Katukiza et al. 2012; Nansubuga et al. 2016).

Based on some of these indicators, a number of sanitation technologies have been developed, tested and used in various slums. These technologies range from traditional/simple pit latrines, ventilated improved pit latrines, toilets connected to septic tanks and the urine diversion dehydrating (UDD) technologies such as ecological sanitation (EcoSan) (Mara et al. 2007; Paterson et al. 2007; Tumwebaze et al. 2011; Katukiza et al. 2012). With this variety, there are various suggestions of technologies that would be considered appropriate. Paterson et al. (2007) propose that a simplified sewer system is technically feasible and economically appropriate for the urban poor, especially after its successful implementation in Pakistan as detailed in Hasan (2006, 2008) and Mcgranahan (2015). In Uganda, urine diversion dry toilets and biogas latrines are recommended as appropriate for the slums (Katukiza & Ronteltap 2010; O’Keefe et al. 2015b). These varied recommendations signify that each slum has different conditions and therefore an appropriate sanitation technology ought to be specific to each slum.

These slums are common in most African countries. In East Africa, it is noted that there is a dearth of research on a suitable sanitation technology for the slums (Letema et al. 2014). This shortage in research on sustainable sanitation technologies is further complicated by multiple service actors in the sanitation sector including governmental and non-governmental organisations, community-based organisations, informal service providers and slum residents (Letema et al. 2014). These actors are often times unregulated or lack coordination among themselves (Tukahirwa et al. 2013; Galli et al. 2014; O’Keefe et al. 2015a). In addition, sometimes the views of all stakeholders are not factored into decision-making, as shown by O’Keefe et al. (2015b) who recommend that in addition to views from scientific experts, evaluation of sanitation technologies by end users is crucial for the success of sanitation interventions. It is therefore crucial that attention is given to the beneficiaries and end users of sanitation systems (Tilley et al. 2014).

These end users, who are slum residents, would best understand conditions in their settlements and can provide information that is important in decision-making by all relevant stakeholders concerning appropriate sanitation technologies. Therefore, realising the important role played by end users and the lacuna in suitable sanitation technologies, the aim of this study was to highlight characteristics of an appropriate sanitation technology for the slums of Kisumu city in Kenya by understanding the sanitation technology preferences of residents in the slums. A description of the study area, including data collection methods will be presented, after which results, a discussion, a conclusion and policy implications will follow.

Methods

Study area

Kisumu is a city in Kenya’s western region. It is in Kisumu County and has an estimated population of 420,000 people (Republic of Kenya 2013), with more than half of this population living below the poverty line (UN-Habitat 2005). There are a number of slum settlements around the city, namely Manyatta A, Manyatta B, Nyalenda A, Nyalenda B, Bandani, Obunga Manyatta Arab, Kibos and Kaloleni. These slums host approximately 60% of the city’s population (Syrjänen 2008).

The slum settlements are situated on land that is registered as freehold land, whose owners obtained through inheritance (Huchzermeyer 2009; Smith 2017). As population increased, some landowners constructed rental units on their pieces of land to cater for the increasing demand for housing, while others moved and settled in other areas. Consequently, there are resident landlords (with no tenants), landlords with tenants, as well as absentee landlords. These residents live in compounds/plots, which are a group of households who share basic services such as water and sanitation.

Further estimates reveal that approximately half of the population in the settlements lacks sanitation facilities (Karanja 2010). There are a few toilets connected to septic tanks, but the majority of sanitation facilities are pit latrines (Letema et al. 2014; Simiyu et al. 2017). These pit latrines are usually unlined and are constructed by semi-skilled masons, who are from the same slums. The poor construction quality of the latrines, coupled with environmental factors in the slums such as high water tables and loose soils, reduces the longevity of these pit latrines as most of them collapse during the rainy season (UN-Habitat 2003, 2005). In the recent past, however, ecological sanitation facilities and communal biogas latrines have been introduced in the slums by stakeholders from the non-governmental sector (Letema et al. 2014).

Theory and data collection methods

Due to the challenges faced by residents of slum settlements such as lack of adequate sanitation facilities, research on sanitation preferences may be challenging due to the sensitivity of the subject. A few studies, such as that by Lagerkvist et al. (2014), have been conducted to understand sanitation preferences in slum settlements. Due to the sensitivity of the subject, the current study employed simple methods to understand sanitation technology preferences. This study borrowed concepts from Van Ittersum et al. (2007) who discuss that the importance of attributes can be measured in three dimensions: salience (importance of an attribute in memory which is measured by free elicitation methods), relevance (importance based on personal values and desires, measured by rating, ranking and point allocation methods) and determinance (importance in judgement and choice, measured by multi-attribute methods, trade-off methods, swing weight methods and conjoint methods). Two of these attributes, salience and relevance, were incorporated in this study.

The study was conducted in Nyalenda A, Nyalenda B and Obunga slums. The slum settlements are divided into clusters, commonly referred to as ‘units’. Together with local leaders, transect walks were taken through the settlements to identify cluster boundaries. Data collection methods were mainly household interviews and group discussions with landlords, tenants and caretakers.¹

For the household interviews, one cluster was selected from each of the three slum settlements. Interviewers selected compounds by walking through each cluster from one end to the other. One household per compound was selected for interviewing. Using a semi-structured interview guide, respondents were asked to mention sanitation technologies that they knew of and/or used (thereby measuring salience dimension since the responses were from the interviewees’ memory). The respondents were then asked to classify the technologies they had identified from the most preferred to the least preferred, giving reasons for their classification. Selection and interviewing of households continued until the cluster had been combed through, and it was felt that sufficient information had been gathered from the respondents. A total of 34 interviews were conducted with respondents who were mainly landlords, tenants and caretakers, and some of whom lacked sanitation facilities within their compounds.

After household interviews, a different criterion was employed to select respondents for group discussions. This criterion was based on type of respondent (landlord/tenant) and availability of a toilet in the compound, and it led to eight categories of respondents:

1. Resident landlords, with tenants, but who lack a sanitation facility.

2. Resident landlords with tenants, and have a sanitation facility.

3. Landlords without a sanitation facility.

4. Tenants with resident landlords, and have a sanitation facility.

5. Tenants with resident landlords, but lack a sanitation facility.

6. Tenants with absentee landlords, and have a sanitation facility.

7. Tenants with absentee landlords, and lack a sanitation facility.

8. Tenants living on compounds with caretakers, and have a sanitation facility.

These respondents were selected from all clusters of each settlement. The same sampling strategy used to select respondents for household interviews was adopted to select adult household heads. These selected participants were invited to a common venue, on a specified day and time, for group discussions.

Each group had their meeting on a separate day and time from another group. In each session, there were a minimum of 8 and a maximum of 15 participants. The meeting began with participants being in smaller groups of at most four participants in order to ensure that there was participation from all members, and that those with dominant voices did not lead to others being silent (Nieuwenhuis 2010, p. 91; Cohen et al. 2011, pp. 436–437).

Within these small groups, just like it was during household interviews, participants were asked to individually identify sanitation technologies that they knew and/or used, and classify their preferences on provided sheets of paper from the most to the least preferred. After this exercise, participants then held a discussion within the small group to discuss the identified sanitation technologies.

After the discussion, all the participants converged, and the technologies that had been identified from each of the smaller groups were presented to all participants. The participants were then asked to rank the technologies in order of most viable to least viable within their slum settlements. Participants therefore had an opportunity to collectively discuss the advantages and disadvantages of the identified sanitation technologies, and to finally arrive at a consensus of what was most viable (in their own view) in their settlements. The participatory group discussions measured the dimension of relevance through ranking of sanitation technologies, which was an expression of the residents’ desires. These desires, expressed as advantages/disadvantages, were pointers to characteristics of an appropriate sanitation technology for the slums.

To assess that adequate and valid data had been collected, a number of strategies were applied. First, the study used multiple methods to ensure adequacy and validity. In the first stage, household interviews stopped after saturation was achieved. The second stage of small group discussions was used to validate the data that had been collected during the household interviews, and to collect any more information that may have been missed during the interviews. Finally, the study also used the group discussions to further validate and assess the depth of the data.

Ethical considerations

Research authorisation was granted by the Kenya National Commission for Science and Technology, research permit letters were obtained from Kisumu County Education office and permission was granted by chiefs in the slums. Before any interview, respondents were informed of the aims of the study and given time to make a decision on whether to participate or not. The interviews only began after respondents consented. During selection for group interviews, the participants were also informed of the aims of the research and their role in the discussions. After the discussions, participants were compensated for their time and participation.

Data management and analysis

Data from both interviews and participatory group discussions were recorded on audio devices, after which they were transferred to computers. Preference of sanitation technologies from household interviews and the small groups was summarised descriptively using the Statistical Package for Social Science (Vs.21). Qualitative data analysis began during data collection and involved listening to participants and probing for clarification. After data collection, recorded interviews were transcribed verbatim into Microsoft Word. These transcripts were then read and re-read in order to gain familiarity with the data. Analysis of the data followed a thematic content approach. The main themes were the sanitation technologies that had been summarised descriptively. Codes that explained the reasons for or against the respondents’ choices of sanitation technologies were developed from the household and group interview transcripts. Coding was an iterative process that required reading and re-reading the transcripts again to ensure that the transcripts were adequately coded. The codes were then compared and, where necessary, two or more codes that explained the same concept were merged into one main code. The codes were then linked to the identified sanitation technologies, in order to explain reasons for or against the sanitation technologies. These reasons were pointers to characteristics of an appropriate sanitation technology for the slums.

Results

Identified sanitation technologies

Four main technologies were identified: pit latrines, flush toilets (connected to a sewer system), ecological sanitation (EcoSan) facilities and pour flush toilets. According to respondents, a pour flush toilet has a sunken bowl instead of a drop hole as is the case with a pit latrine. It does not have a cistern and requires that water is poured into the bowl after use.

Preferred sanitation technologies

Data from household interviews and small group exercises, shown in Tables 1 and 2, revealed that most respondents preferred pit latrines and flush toilets.

Table 1. Preferred sanitation technologies from household interviews.

	Pit latrine	Sewer system	EcoSan	Total
First option	16 (47%)	16 (47%)	2 (6%)	34
Second option*	17 (71%)	5 (21%)	2 (8%)	24
Third option*	1 (14%)	1 (14%)	5 (72%)	7

*Second and third options have less count because respondents were allowed to give at least one technology depending on what they knew.

Table 2. Preferred sanitation technologies from group exercises.

	Pit latrine	Sewer system	EcoSan	Pour flush	Total
First option	38 (48%)	26 (32%)	16 (20%)	0	80
Second option	29 (43%)	22 (32%)	13 (19%)	4 (6%)	68
Third option	7 (18%)	13 (33%)	18 (46%)	1 (3%)	39

During the participatory discussions, participants debated over the feasibility of the technologies they had identified and finally agreed on the most viable option. In most of the categories, the pit latrine was selected as the most viable and preferred technology (Table 3).

Table 3. Preferred sanitation technologies from participatory discussions.

Category of respondents	First option	Second option	Third option
Resident landlords, with tenants, but lack a sanitation facility	Pit latrine	Flush toilet	EcoSan
Resident landlords with tenants, and with a sanitation facility	Pit latrine	EcoSan	Flush toilet
Landlords without a sanitation facility	EcoSan	Pit latrine	Flush toilet
Tenants with resident landlords, and have a sanitation facility	Pit latrine	Flush toilet	EcoSan
Tenants with resident landlords, but lack a sanitation facility	Pit latrine	EcoSan	Flush toilet
Tenants with absentee landlords, and have a sanitation facility	Pit latrine	Flush toilet	EcoSan
Tenants with absentee landlords, and lack a sanitation facility	Pit latrine	Pour flush	EcoSan*
Tenants from compounds with a sanitation facility, and with caretakers	Pit latrine	EcoSan	Flush toilet

*This group ranked the flush toilet as the fourth option.

Reasons for sanitation technology preferences

Respondents mentioned various reasons for or against the sanitation technologies.

Pit latrines

Pit latrines were voted as ideal because of the advantages they offered. They could be shared among several households, and compared to EcoSan facilities, they took longer to fill up. Compared to flush toilets, they were easier to maintain as shared sanitation facilities, as expressed by one tenant who said that ‘we are many here, so pit latrine is the best …’

However, the disadvantage with pit latrines was that most of them collapsed during the rainy season and they required emptying. Respondents of group discussions acknowledged that the collapse of pit latrines was partly due to poor workmanship, which could be reduced by proper construction of the latrines. To deal with emptying, the respondents acknowledged that the manual pit latrine emptiers were employed to empty filled-up latrines. Participants felt that pit latrines were ideal for their settlements because the slums were very congested, thereby limiting the municipality trucks from access in order to empty the filled-up latrines. In addition, manual emptying provided an opportunity for the pits to be reused again, thereby cutting down costs of constructing new latrines. Pit latrines did not require a constant water supply like the flush toilet, and since the slums did not have a constant water supply, pit latrines were thought of as ideal.

Participants also acknowledged that due to poor solid waste management in the slums, residents dumped other waste material such as sanitary pads and diapers in the pits:

‘You will find everything dumped in the pits … baby’s diapers, sanitary pads, and a lot of other solid waste material’.

Although participants acknowledged that such behaviour was not proper, they noted that compared to the other sanitation technologies, pit latrines would work best in the slums because they could contain such waste.

EcoSan

Generally, most respondents did not know about the EcoSan or its operation. Respondents who knew about the EcoSan preferred it because it was raised from the ground, giving it a good-looking appearance compared to the traditional pit latrine. EcoSan was lauded for being easier to empty since human waste was contained in a bucket that was to be emptied every time it was full. Landlords who had farms particularly preferred EcoSan because of manure that could be used on their farms. These landlords lived in flood-prone areas, where pit latrines collapsed almost every year during the rainy season.

On the other hand, many respondents lived in high-density areas where they shared sanitation facilities and lacked farmlands where manure could be used. These respondents disliked EcoSan toilets because of the high fill-up rate and frequent emptying process. Due to congestion in the settlements, sanitation facilities were often shared, and the EcoSan was considered not ideal as it was likely to have maintenance challenges among the users, especially related to use and emptying. A tenant, describing the challenges that the EcoSan might pose due to sharing with other tenants, said that ‘it is difficult to take care of it’.

One landlady further expressed the financial challenges related to the frequent emptying of EcoSan by saying:

‘I will not be able to empty it every so often … how many times will I need to empty it and how much will I use?’

Participants expressed that although the EcoSan toilet had been introduced in the slums and was being used by some residents, they noted that it was still not ideal because of these challenges it presented. They also acknowledged that the cultural beliefs are likely to hinder full adoption of EcoSan since many residents would shy away from emptying (culturally, the dominant community in the slums believed that people should not handle and/or touch faecal matter, except that for children).

Flush toilet/sewer system

The flush toilet was preferred because, just like the EcoSan, it looked presentable. Respondents expressed their preference by saying that the flush toilet was ‘nice and clean’. However, most respondents preferred the flush toilet only if it were a household sanitation technology that was not shared. As a household facility, ‘it was easier to maintain’ and ‘one can keep it clean’.

Despite the preference for a flush toilet, respondents acknowledged its limitations. The slums lack a sewer system and a regular supply of water, and thus the flush toilet was unrealistic. It was also costly, and not appropriate for sharing because it would pose maintenance and cleaning challenges. Consequently, most respondents wished for a flush toilet, but realistically, they acknowledged that they may still use the pit latrine. For example, one landlady stated as follows:

‘I prefer a flush toilet … but I would still construct a pit latrine because it is helpful when there are many people using the toilet’.

By saying so, the landlady expressed that she would prefer a flush toilet as her household facility, but for sharing among tenants, a pit latrine was more ideal. These sentiments were also expressed by participants in group discussions who acknowledged that these limitations make the flush toilet not practical in the settlements. Participants further noted that because many residents in the slums were ignorant/lacked knowledge, the flush toilet may be difficult to operate, and consequently it may easily break down due to improper usage. As an example, participants noted that some residents may dispose other types of waste into flush toilets leading to blockages.

Pour flush toilet

The pour flush toilet was mentioned during the group discussions. Generally it was preferred because it presented some advantages of the pit latrine and the flush toilet. For instance, participants in the group discussions noted that the pour flush toilet was presentable and it did not require a constant supply of water like the flush toilet. They also acknowledged that it could be shared among tenants. However, participants also noted that the pour flush toilet may not be ideal because residents may not use it appropriately, thereby posing challenges. They also noted that the pour flush may be costly for most residents compared to the pit latrine. Finally, since masons in the slums mostly constructed pit latrines, the pour flush toilet may not be ideal because of lack of technical expertise for its construction.

Discussion

Four sanitation technologies were mentioned by respondents, with the pit latrine being the most commonly used and preferred technology. Studies have shown that use of pit latrines is common in slum settlements in East Africa (Kulabako et al. 2010; Omambia 2010; Szántó et al. 2012; Isunju et al. 2013; Tumwebaze et al. 2013; Pieter Van Dijk et al. 2014; Jenkins et al. 2015; Nakagiri et al. 2015; Tsinda et al. 2015) and in other African countries such as Ghana (Adubofour et al. 2013; Obeng et al. 2015), Senegal (Gulyani et al. 2010; Scott et al. 2013), Zimbabwe (Chinyama et al. 2012) and Malawi (Chunga et al. 2016). In spite of their common use, pit latrines present a couple of shortcomings. In settlements like those of Kisumu, they collapse during rainy seasons due to the weak soil structure and high water table, they attract flies and the pits can be breeding grounds for insects such as mosquitoes which pose public health risks and are a nuisance to users (Nakagiri et al. 2016). In addition, they often need to be emptied when they fill up which is a challenge in slums because municipal trucks cannot access the congested slums (Katukiza et al. 2012; Tsinda et al. 2013; Kwiringira et al. 2014; Semiyaga et al. 2015).

Nevertheless, residents in slums, including those in Kisumu, have devised ways of handling such challenges, hence the popularity of pit latrines. For instance, they employ manual pit latrine emptiers who empty filled-up latrines. Manual pit latrine emptying, however, poses health risks to those who empty the pits, especially if they do not use protective clothing. These risks include exposure to helminths, metals and gases that may be present in the pits (Thye et al. 2011; Ahamada et al. 2016). In other slum areas in Uganda, studies show that pit latrines are sometimes raised from the ground during construction as a way of dealing with weak soils and flooding, and for emptying, manual or mechanical means are used (Niwagaba et al. 2008; Isunju et al. 2013). In some cases, when households have space, a filled-up latrine is closed and a new one is constructed as is the case in Mongolia (Sigel et al. 2012),but when the space is not available, the filled-up pit is emptied. In Kisumu, the popularity of the pit latrine in spite of its shortcomings may be due to various factors: residents have used the pit latrine for a long time and have therefore learnt to deal with the challenges associated with it. It may also be that residents lack knowledge of other sanitation technologies, also evidenced by the lack of knowledge of the EcoSan toilet. This latter reason is equally posited by Niwagaba et al. (2008) as a reason for continued use of pit latrines in the urban areas of Kampala. Similarly, the preference for the pit latrine may be due to challenges with alternative sanitation technologies, a challenge also experienced in Malawi (Chunga et al. 2016). Coupled with these challenges, living conditions in slums make the pit latrine more practical; for instance, high population density and lack of space to construct individual household sanitation facilities force a number of households to share one pit latrine. The high number of users means high loading rates, which may not be ideal for technologies such as EcoSan.

The EcoSan (sometimes referred to as the Urine Diversion Dehydrating toilet) was not voted as ideal due to the disadvantages it poses, more often related to the living conditions of the slums. Its adoption was still low, a finding similar to other studies that revealed that the EcoSan toilets in Kisumu were not fully in use (Letema 2012; Simiyu 2015). Low preference of the EcoSan in the present study was related to factors such as cultural and social acceptability (fear of/shame in handling human waste), lack of knowledge of their existence and use, and difficulties in operation and maintenance of shared sanitation facilities. In other areas where EcoSan has been introduced, its uptake has been linked to similar challenges: in Bangladesh and South Africa, challenges were mainly related to sociocultural acceptance, costs, maintenance problems and the consequent breakdown of some parts (Roma et al. 2013; Uddin et al. 2014). In Malawi and Zimbabwe, there was low knowledge, and uptake required user education, as well as adequate monitoring and maintenance (Banana et al. 2015). Further, just like in this study, in Malawi, the EcoSan was costly, posed operation and maintenance challenges, and was not suitable among shared households (Chunga et al. 2016). In Mexico, challenges were related to phobia and sensitivity for human faeces (Davies-Colley & Smith 2012).

However, there are other areas where EcoSan has had successful uptake. In Kabale Municipality of Uganda, which has similar environmental characteristics as Kisumu, Tumwebaze et al. (2011) report that EcoSan was the second most used technology after the pit latrine and its success was attributed to the area’s terrain that did not favour pit latrines, the small space taken by EcoSan compared to pit latrines, economic benefit from manure, availability of masons and high level of awareness. EcoSan has therefore been lauded as the most appropriate in the slums of Uganda (Katukiza & Ronteltap 2010). These studies from different countries provide learning points for appropriate sanitation technologies for Kisumu. First, phobia for human waste is often experienced with EcoSan toilets, and it is important that users have a positive attitude towards sanitation technologies. Changing users’ attitudes and behaviour may entail strategies such as education and awareness raising. Behaviour change and adoption of new sanitation technologies happen gradually over a period of time implying that adoption of such technologies will take some time. Second, the acceptance of a sanitation technology varies from one context to another (Tilley et al. 2014) and is often times not guaranteed. It is therefore important to use local resources, including human resources, in construction and maintenance as this will ensure that they could be replicated, and finally, regular monitoring and evaluation should be carried out to identify possible barriers and devise specific strategies to counteract the barriers, as was the case in Mexico (Davies-Colley & Smith 2012).

The other identified sanitation technology was the flush toilet connected to a sewer system, which was praised for its appealing appearance as residents noted that it looked ‘presentable’. It was considered the best option for individual households. However, its limitations were grave, since the slums lacked a sewer system, and experienced frequent water shortages. High costs of installation were also noted as barriers to the acquisition of the flush toilet. Similar reasons for preference and challenges of the flush toilet were noted in the urban areas of Accra in Ghana (Nimoh & Poku 2014), Uganda (Niwagaba et al. 2008) and Botswana (Bolaane & Ikgopoleng 2011). These limitations of a sewer system, mainly costs and an unreliable water supply are also reiterated by Paterson et al. (2007) who advocate for a simplified sewerage system, otherwise referred to as condominial sewerage. Such limitations mean that residents often resort to sanitation technologies that are affordable in terms of installation and maintenance. Moreover, many slum residents buy water for use at home, since they are not connected to a water supply system. It is estimated that poor people (in comparison to their wealthier counterparts) pay more for water, which accounts for a large percentage of households’ monthly spending (UNDP 2006, pp. 51–54). From an economic viewpoint, therefore, it is highly unlikely that the urban poor would buy water to flush it down a toilet as this would mean spending more on their already stretched budgets. Water is likely to be flushed down if it has been used for other purposes such as cleaning and laundering.

The preference of the flush toilet by many respondents, however, indicates a level of knowledge and/or experience with it. Apart from such knowledge and experience, residents were attracted to the flush toilet because of the natural human desire for something better. It could be deduced that the flush toilet is an embodiment of something better than the norm; the norm in this case being the pit latrine. The challenges of the flush toilet, however, proved that it was unattainable given the challenges. Respondents, for instance, acknowledged that due to lack of space, sanitation facilities were shared. As such, challenges of maintenance of shared facilities led users to resort to the pit latrine as the most practical. Since the flush toilet was unattainable, the pour flush toilet was mentioned as a preference, as it was a hybrid between the desired which was unattainable, and the less desired but attainable and practical.

In summary, therefore, Table 4 presents an analysis of strengths, weaknesses, opportunities and threats (SWOT analysis) of each of the identified technologies.

Table 4. Strengths, weaknesses, opportunities and threats (SWOT) analysis of identified sanitation technologies in Kisumu’s slum settlements.

Technology	SWOT analysis
EcoSan	Strengths	Weaknesses
	Recycling of waste products	It requires proper management structures among users, especially if shared
		Lack of management structures can easily lead to conflicts and mismanagement
		It has a high fill-up rate if shared
		Many residents lack knowledge about it
		Lack of locally trained masons to construct it
		High installation costs
	Opportunities	Threats
	Emptying presents a business opportunity	Social and cultural rejection
	Can be used in flood-prone areas with weak soils
	Market for human waste as manure
Flush system	Strengths	Weaknesses
	Is aesthetically presentable	Lack of a sewer system in the slums
	Residents have some level of knowledge about it	High financial requirements for its installation
		Requires constant water supply
		It is not ideal for sharing by many households
	Opportunities	Threats
	Residents may require further knowledge on management	May break down easily due to high number of users
Pit latrine	Strengths	Weaknesses
	Does not need constant water supply	Requires emptying
	Residents have knowledge and experience on its functionality	Inaccessibility by municipality trucks
	Can work better with proper management structures
	Can be shared by several households
	Easier to manage if shared by several households
	Opportunities	Threats
	Emptying can be a business to locals	May collapse easily if poorly constructed due to poor soil structure
	Proper construction can promote its use	Manual emptying poses health risks if done without protective clothing and equipment
Pour flush	Strengths	Weaknesses
	It does not require constant water supply	Lack of a sewer system in the slums
	Can work well with proper management structures	Lack of skilled persons to construct it
	Can be shared by several households
	Opportunities
	It need not be connected to a sewer system
	It may be easily accepted if modified to adopt the advantages offered by the pit latrine

Having considered the four technologies mentioned by residents in Kisumu’s slums, it is important to note that other sanitation technologies and services have been implemented in various slum settlements which also provide vital lessons. In Cape Town (South Africa), for instance, toilets that do not have an outlet, but rather fill up and are emptied regularly (otherwise referred to as container toilets), as well as buckets, have been used for human waste disposal (Mels et al. 2009). Such market-based approaches have lately been adopted in urban slums where there is no space to construct sanitation facilities. Private stakeholders provide alternative sanitation technologies such as the UDD technologies to households. Waste is picked on a regular basis and transported to a decomposing/waste recovery site, and households have to pay for these services. These approaches have been tested in East Africa (Kenya and Uganda) (O’Keefe et al. 2015a) and Haiti (Russel et al. 2015). These examples provide important lessons that can be learnt for the slums of Kisumu. A market-based approach in Kisumu would require commitment from service providers (like the municipality) to empty and replace filled-up toilets on a regular basis. In addition, factors such as maintenance, which would also determine costs, should not be overlooked, especially since most sanitation facilities are shared.

Shared sanitation facilities are often not considered ‘improved’ sanitation facilities since they do not hygienically separate human waste from human contact (UNICEF & WHO 2013). In slums, though, as it is the case in Kisumu, environmental conditions may not favour single household sanitation facilities, because of lack of space for construction. Such limitations call for a technology that is favourable to sharing, and which can be adequately and hygienically maintained by the users in order to ensure sustained use.

Aside from the users, other factors such as the significance of toilets, as well as tenure status and responsibility, are equally important in Kisumu’s slum settlements. It seems that the value attached to a toilet is beyond its importance in human waste disposal. Issues related to aesthetics, social and cultural norms, and human dignity are important considerations. In an urban slum where sanitation facilities are either lacking or not functional, a good looking and functional toilet is a source of pride to its owner/user, and to his/her standing in the community. This is why respondents talked of the EcoSan and flush toilets looking ‘nice/presentable’ and yet others described the emptying process of the EcoSan as being ‘embarrassing’. An appropriate sanitation technology, therefore, needs to be acceptable within the social and cultural contexts of the community and be aesthetically appealing to users.

These users include both landlords and tenants, both of whom have a role to play in sanitation. As has been discussed by Scott et al. (2013), tenants are more likely to share sanitation facilities, be responsible for maintenance, and at times pay for operation and maintenance costs of sanitation facilities. Landlords, on the other hand, are likely to be responsible for paying for initial sanitation investment costs. In this study, after participatory discussions, both landlords and tenants preferred pit latrines. This similarity in preference is due to a common reason – maintenance. Landlords were concerned about maintenance since they were investors, but tenants were concerned about maintenance because they were users. This finding not only highlights the difference in investment and maintenance responsibilities but further points to the need to involve both tenants and landlords in decision-making, as they have a crucial role to play in sanitation improvement in slum areas.

Finally, institutional settings also influence the type of sanitation technologies that can be implemented. The Kenya Environmental sanitation and Hygiene policy (Republic of Kenya 2016) recommends that county governments, like Kisumu County in this case, shall promote low cost and appropriate technologies. Apart from the cost and appropriateness, the policy also identifies factors such as environmental friendliness, affordable and manageable operation and maintenance requirements, and appropriateness for the needs of all population groups including those with disabilities. As Kisumu County domesticates this national policy, this study identifies some of the strengths and weaknesses of the available sanitation technologies. These strengths and weaknesses associated with sanitation technologies are pointers to characteristics of an appropriate sanitation technology, which can be summarised in seven thematic areas: technical, environmental, sociocultural, operation and maintenance, economical, public health and institutional settings as shown in Table 5.

Table 5. Characteristics of an appropriate sanitation technology for Kisumu’s slums.

Theme	Characteristics
Technical	It should be easy to use It should not be wholly water and sewer dependent, which would limit its functionality One that can be constructed using locally available materials Local masons can be trained to construct it, which will ensure that more households can acquire the technology It should safely contain human waste while awaiting transfer to waste recovery site
Environmental	It should be feasible in an area that experiences flooding and that has a high water table and loose soils
Sociocultural	It should be socially and culturally acceptable Responsibilities between landlords and tenants can be defined, hence encouraging healthy relationships It should be aesthetically presentable in design A possible history of success and replicability elsewhere presents lessons to learn from
Operation and maintenance	One that is easy to operate One that can be shared by a certain number of households who can then develop a maintenance plan, e.g. for cleaning
Economic	It should be affordable in terms of installation, operation and maintenance It should have additional economic benefits other than human waste disposal (e.g. waste recovery) It should have avenues for employment/job creation
Public health	Should limit contact with human waste thus minimising spread of disease
Institutional settings	It should fall within the policy recommended guidelines and recommendations

Conclusion

Four main sanitation technologies are used in the slums of Kisumu and were identified by residents: pit latrines, flush toilets, EcoSan toilets and the pour flush toilet. Residents identified strengths and shortfalls of each of these technologies that qualify them as ideal or less appropriate for the slums. The pit latrine was considered the most practical sanitation technology because, apart from providing a means for human waste disposal, it was affordable in terms of installation and maintenance, it was socially and culturally acceptable in the community, it was easy to use and construct using local resources and was ideal for sharing among households. Based on the strengths and shortfalls of the identified sanitation technologies, characteristics of an appropriate sanitation technology for the settlements were deduced. An appropriate technology for Kisumu’s slums should offer the advantages offered by pit latrines but should not be limited by environmental conditions. This study highlights the importance of integrating views from end users, including landlords and tenants, as they all are crucial in ensuring the sustainability of sanitation technologies in slum settlements. This study has also highlighted that participatory approaches allow slum residents to be deliberate and eventually arrive at a consensus since they are residents as well as end users of sanitation technologies. It is through such deliberations that sanitation technologies are identified and evaluated, facilitating decision-making. Notwithstanding, identifying an appropriate sanitation technology is just but one of the steps towards sustainable sanitation in Kisumu’s slums.

Implications for policy and areas for further research

As Kisumu County strives to domesticate the Environmental Sanitation and Hygiene Policy, this study provides further insight into factors that should be considered in the design of a sanitation technology that can be adopted for the city’s slum settlements. The involvement of various stakeholders, including the end users, in designing, testing and implementing this sanitation technology, is crucial.

The policy should also include components of training and/or awareness creation on proper use of and maintenance of sanitation technologies, especially in the light of the challenges in the slum settlements. Key persons in the community such as leaders and landlords should be involved as agents of awareness creation.

Further research should therefore look into designing an appropriate sanitation technology and testing its acceptability and functionality in the slums.

Limitations

The results of this study are based on the technologies that the respondents knew, which is also what was available in the slum settlements. The results may thus only be applicable to slums in Kisumu. Nonetheless, it is important to note that the characteristics of an appropriate sanitation technology identified in the study may be applicable to other slum settlements, but contextual differences need to be taken into consideration.

Acknowledgements

Appreciation to Beatrice Obondo, Esther Omunga, Petronilla Omollo, Wycliffe Opiyo, Francis Omungo, Molly Mariga, Zedekiah Tunya, Peter Owago, Jeniffer Akinyi and Kennedy Ochieng for their assistance during community mobilisation and data collection.

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Funding

This work was supported by the Sanitation and Hygiene Applied Research for Equity (SHARE) Consortium.

Notes on contributors

Sheillah Simiyu

Sheillah Simiyu completed her PhD from the University of Stellenbosch and is currently a Post-Doctoral fellow at the Great Lakes University of Kisumu. Her research interests are in Water, sanitation and Hygiene, Behaviour change, and Community health and development.

Notes

1. Persons appointed by landlords to oversee the rental premises, especially if the landlord is not resident in the compound. They act as point persons between tenants and the landlord, oftentimes being responsible for rent collection and any tenants’ complaints.