Anaerobic stabilisation of urine diverting dehydrating toilet faeces (UDDT-F) in urban poor settlements: biochemical energy recovery

As an innovative solution to enhance sanitary conditions in informal settlements in low income countries, urine diverting dehydrating toilets (UDDTs) have been adopted (Austin & Cloete 2008; Niwagaba et al. 2009a; Schouten & Mathenge 2010; Katukiza et al. 2012). Such is also the approach adopted by Sanergy, a social enterprise working on sanitation improvement within informal slum settlements, in Nairobi, Kenya. Sanergy fabricates and installs the Fresh Life^© toilets in collaboration with entrepreneurs in the slums who maintain them. Currently, approximately 7,000 kg of faeces is collected from the UDDTs, further referred to as UDDT-F, and delivered to a central treatment plant on a daily basis. Owing to the high pathogenic levels in human waste (Feachem et al. 1983), an extra pathogen inactivation step is required especially when the faecal matter will be valorised for agricultural purposes. A number of different treatment technologies were developed for source separated human faeces and include plain storage, composting, black soldier flies, chemical treatment, vermi-composting and anaerobic digestion (AD) (Vinnerås 2007; Niwagaba et al. 2009b; Rajagopal et al. 2013; Strande et al. 2014; Fagbohungbe et al. 2015). The main treatment technology applied by Sanergy for UDDT-F is composting, producing an end product that is sold as organic manure (Evergrow^@), available on the Kenyan market. Moreover, the increasing amount of collected UDDT-F on a daily basis sparked a need for diversification of the treatment options.

The potential for application of AD at any scale and almost any place (Van Lier et al. 2008; Pabón-Pereira et al. 2014), marked the decision to select AD as a faecal waste treatment option in informal slum settlements. In addition, by means of AD, the chemically stored bio-energy in the organic waste can be recovered as biogas, providing an alternative fuel for local use (Abbasi et al. 2012). AD is considered an efficient technology for the stabilisation of organic wastes, producing a digestate with a high fertiliser value (Berndes et al. 2003; Martín-González et al. 2010; Park et al. 2016). The key reported drawback in AD is inadequate pathogen inactivation (Kunte et al. 2000; Chaggu 2004; Horan et al. 2004; Massé et al. 2011; Chen et al. 2012; Fagbohungbe et al. 2015) and low methane production especially from human faecal matter (Rajagopal et al. 2013; Fagbohungbe et al. 2015). It must be noted that the microbiological safety of the digestate and treated sludge is essential as it has implications for human health and cycling of pathogens in a densely populated environment through the food chain (Avery et al. 2014). As such, this study is part of a wider research on the potentials for the anaerobic stabilisation of UDDT-F, enhancing biogas production and pathogen inactivation, with the present paper focusing on the production of another side-product next to hygienised sludge, i.e. biogas.

In our previous study, we evaluated the accumulation of volatile fatty acids and their effect on pathogen inactivation during the digestion of UDDT-F and mixtures of UDDT-F and organic market waste (OMW) in a one- and two-stage plug flow anaerobic digester (Riungu et al. 2018b). Results showed higher pathogen inactivation in the two-stage plug flow digester, with the digestate meeting WHO standards of 1,000 CFU/100 mL, applying a solids retention time (SRT) of 29 days. The used OMW, widely available and at close proximity to UDDT-F source, is characterised by a vast readily degradable organic fraction (Zhang et al. 2008; Riungu et al. 2018a). In addition to sludge hygienisation, the production of an alternative fuel (biogas) from the faecal matter will very likely accelerate the acceptance of the proposed technology. As such, the research described herein focused on the potential for biogas production during anaerobic stabilisation of UDDT-F using laboratory-scale biochemical methane potential (BMP) tests and pilot-scale plug flow one- and two-stage anaerobic digesters. Under pilot-scale experiments one-stage and two-stage plug flow digesters were researched, applying UDDT-F:OMW ratios of 4:1 and 1:0, at about 10 and 12% total solids (TS) slurry concentrations.

UDDT-F samples used for this study were obtained from the Fresh Life^® UDDT within Mukuru Kwa Njenga/Mukuru Kwa Reuben informal slum settlement, Kenya. Fresh Life^@ toilets are offered on a pay-and-use basis in the form of serviced shared facilities, charging between 0.05–0.1 euros per use. Within each toilet facility, a 30L container is used for waste collection, with approximately 10g sawdust added by the user after every toilet use. The toilets are emptied on a daily basis, where used containers are replaced by clean ones. Five containers with UDDT-F were randomly selected after which mixing of the contents was done in order to obtain a homogeneous mix.

OMW was collected from vegetable vendors, eating points and waste disposal points within Mukuru Kwa Njenga and Mukuru Kwa Reuben informal slum settlements. About 20kg of the waste was collected and contained food waste, vegetable waste and fruit waste, in equal proportions. Size reduction was achieved by manual chopping to about 1cm size for pilot-scale test substrates whereas samples for laboratory-scale tests were blended using Ramton^© domestic blender for 1 minute. Table 1 shows the characteristics of the UDDT-F and OMW that was used in the study. After collection the waste was refrigerated at 4 °C to minimise bioconversion of the samples prior to testing.

Inoculum for the AD experiments used in this study was obtained from an onsite fixed dome anaerobic digester within Kibera informal settlement, Kenya. The bio-centre was erected by Umande Trust, a non-governmental organisation (https://umande.org/) and managed in partnership with a community-based organisation, Kibera Kids Youth Organisation (KIDYOT). The inoculum upon collection was incubated for 1 week to methanise any organic matter before use.

BMP test experiments were performed to access methane production during anaerobic stabilisation of faecal waste. Three substrate ratios, based on our previous study (Riungu et al. 2018a), that investigated the effect of volatile fatty acids (VFAs) on pathogen inactivation were applied; UDDT-F:OMW ratios 1:0, 4:1 and 0:1. An inoculum to substrate ratio of 2:1 (Zeng et al. 2010) was used, maintaining approximately 1.5 g volatile solids (VS)/100 mL solution, based on initial VS concentration of inoculum and substrate.

Batch digestion experiments were conducted in triplicate using 100 mL glass serum vials (80 mL working volume). After adding the required amounts of substrate and inoculum in each serum vial, basic anaerobic medium (BAM) was added according to Angelidaki et al. (2009) (Table 2), in addition to 1 g/L sodium carbonate buffer. Hereafter, tap water was added to a volume of 80 mL. The vials were sealed with butyl rubber stoppers and flushed with argon gas for 30 seconds to purge out oxygen. The vials were incubated at 35(±1) °C for 30 days, with manual mixing. Triplicate blanks that contained inoculum and BAM were incubated in order to correct for gas production from the inoculum. Gas pressure in the digesters was measured regularly with a digital pressure meter model GMH 3150 (Greisinger, Germany) utilising a sensor model MSD 4 BAE with a resolution of 1 mbar.

Pilot-scale substrate selection was based on a series of laboratory-scale batch-tests derived from previous experimental data (Riungu et al. 2018a) applying UDDT-F:OMW ratios of 4:1 and 1:0. In addition, research aimed at treating the highest possible substrate's TS concentration that can freely flow through the plug flow digester without the necessity of using pumps. As such, 12% TS was chosen as the highest substrate TS concentration with additional experiments at 10% TS for assessing the impact of lower TS concentrations on biogas production.

Two sets of digesters were used, namely a one-stage digester (R_o) and a two-stage digester (R_am) comprising a hydrolysis/acidogenic digester (R_a) and a methanogenic digester (R_m).

The hydrolysis digesters (R_a) were fabricated from 30 L plastic containers, with a working volume of 20 L. These digesters were equipped with a cover, incorporated with two separate ports, i.e. a feeding port and a port fixed with a manual stirring mechanism, whereas the bottom of each digester was equipped with a discharge/effluent valve.

Six plug flow digesters (Figure 1) were constructed using 175 L tubular polyethylene bags, with polyethylene material thickness being 0.2 mm. The digesters had a liquid capacity of 145 L, with up to 30 L available for in-vessel biogas storage. The majority of biogas produced flowed by pressure to a 175 L biogas storage bag that was installed directly above each digester. In addition, three separate ports were incorporated onto each bag: inlet port (SP₁); sampling port (SP₂) at 0.7 m digester length; a gas discharge port at 1.4 m digester length; and effluent/discharge port (SP₃) at 2.1 m digester length. A total SRT of 29 days was maintained for the AD process.

Digesters were inoculated using the inoculum described above under ‘Inoculum’. The six plug flow digesters D₁, D₂, D₃, D₄, D₅ and D₆, were divided into two groups, D₁-D₃-D₅ and D₂-D₄-D₆, referring to one-stage digestion of UDDT-F:OMW ratio 1:0 at 12% TS (R_o-1:0,12%) and UDDT-F:OMW ratio 1:0 at 10% TS (R_o-1:0,10%), respectively. Every morning, 5 L/day of the substrate was fed to the respective digesters. Stabilisation of the digesters was achieved after 6 weeks, and sample collection and analysis commenced and continued for a further 8 weeks.

The impact of co-digestion on biogas production and organic matter stabilisation was assessed by applying both one- and two-stage digesters, utilising a UDDT-F:OMW ratio of 4:1 at 12% TS, i.e. R_o-4:1,12% and R_am-4:1,12%, respectively. In these experiments, the six plug flow digesters were also divided into two treatments groups, where digesters D₁-D₃-D₅ consisted of the two-stage R_am-4:1,12% digesters and digesters D₂-D₄-D₆ comprised the one-stage R_o-4:1,12% digesters. Every morning, 5 L of feed substrate was fed into the one- and two-stage digesters, with feed substrate being prepared as follows: (1) One-stage: freshly prepared UDDT-F:OMW ratio of 4:1 at 12% TS concentration, (2) Two-stage: Hydrolysis/acidogenic (R_a) digester effluent acted as influent to the methanogenic digesters (R_m). The pH of R_a effluent (4.9 ± 0.1) was adjusted by titration using two-stage (R_am) digester effluent to a range of 5.8–6.2 prior to feeding it to the R_m digesters. For all digesters, stabilisation of biogas production was achieved after two months when data collection commenced. Finally, the concentration of the feed into R_o-4:1,12% was reduced to 10% TS.

Samples from experiments were taken on a weekly basis for analysis of TS and volatile solids (VS), whereas biogas and methane analysis was carried out on a daily basis over the entire experimental period.

The net gas production for calculating the BMP values was obtained by subtracting the gas production of the blank samples.

Biogas flow measurements in the pilot-scale digesters were performed using American Meter Company gas flow meters (Model AC-250) with IMAC Systems pulse digital counters and a vacuum pump.

Determination of methane content in biogas in laboratory- and pilot-scale experiments was performed by liquid displacement method. Herein, a known amount of biogas was passed through a 5% sodium hydroxide solution to strip CO₂. Under laboratory-scale BMP test, methane measurement was carried out twice a week while in pilot-scale test, methane measurement was done once a day. In this approach the quantity of H₂S in the biogas is considered negligible.

TS and volatile solids (VS) analysis were conducted according to the gravimetric method (SM-2540D and SM-2540E), as outlined in Standard Methods for the Examination of Water and Wastewater (APHA 1995).

Bivariate Pearson's correlation test was used to assess trends in methane production from individual digesters within a given experiment. From each of the three trials, the data obtained was analysed by computing the averages, standard deviations and standard errors. Results obtained were presented either in table or figure form.

Figure 2 shows cumulative methane produced against time for UDDT-F:OMW ratios 1:0, 4:1 and 0:1. Highest methane production was recorded within the first 10 days of the experiment, with UDDT-F:OMW ratio 0:1 attaining 45.8 mL CH₄/g VS added/day (Figure 2) whereas UDDT-F:OMW ratios 1:0 and 4:1 depicted 27.3 and 17.1 mL CH₄/g VS added/day respectively. After the 10th day, a decline in methane production was observed up to the 30th day of the experiment.

Overall, 271 ± 13, 315 ± 26 and 521 ± 36 mL CH₄/g VS_added was recorded from UDDT-F:OMW ratios 1:0, 4:1 and 0:1 respectively (Figure 2). An average of about 0.26–0.30 L CH₄/g VS_added has been reported in batch-scale BMP assays of black water (Rajagopal et al. 2013), and about 250 mL CH₄/g VS_added in AD of human faecal material (faeces + urine) (Fagbohungbe et al. 2015). The findings showed an increasing trend in biogas production with the increase in OMW fraction within the feed substrate, which is congruent to the observed higher VFA build-up at increasing OMW fractions in our previous work (Riungu et al. 2018a). The produced VFA was subsequently converted to biogas. In practical situations where biogas generation is the main driver for implementing AD, the use of OMW as sole substrate may lead to excessive VFA build-up and subsequent system acidification (Angeriz-Campoy et al. 2015; Riungu et al. 2018a). In our previous work, pH levels declined to below 4 at UDDT-F:OMW ratios lower than 1:2, inhibiting methanogenic activity. In general, OMW is carbohydrate rich, has a high C/N ratio and is easily hydrolysable (Gómez et al. 2006; Lim et al. 2008), in addition to containing appreciable amounts of fats that are easily hydrolysable to long chain fatty acids (Silva et al. 2014; Angeriz-Campoy et al. 2015). As such, in co-digestion of OMW and UDDT-F, both substrates complement each other: UDDT-F is characterised by a low carbon to nitrogen ratio (Mata-Alvarez et al. 2011; Fonoll et al. 2015) and low methane production (Rajagopal et al. 2013), it provides adequate micro/macro nutrients, alkalinity and moisture content (Silvestre et al. 2015).

The experiments evaluated the impact of digester configuration, co-digestion and substrate concentration on the accumulating methane production during an 8-week time period. All digesters showed a linear increase in accumulating methane with time (Figure 3). Overall, the obtained trend in accumulated methane production per g VS added was in the order: R_am-4:1,12% > R_am-4:1,10% > R_o-1:0,12% > R_o-1:0,12% > R_o-1:0,10%, with minimal differences between the co-digesting experiments (Figure 4). Results from bivariate Pearson's correlation test performed on triplicate samples within each experiment showed high and significant correlation in methane production within a particular experiment, all being within the range of r = 0.474**–0.840** (**correlation is significant at the 0.01 level (2-tailed)), indicating good digester progress throughout the experimental period.

The effect of digester configuration was gauged by applying a UDDT:OMW ratio of 4:1 at 12% TS in a one-stage (R_o-4:1,12%) and two-stage digester (R_am-4:1,12%). Methane production in R_o-4:1,12% and R_am-4:1,12% digester was comparable with corresponding values being 314 ± 15 and 325 ± 12 mL CH₄/g VS added (Figure 4), respectively.

However, average biogas production in R_am-4:1,12% was 571 ± 25 mL CH₄/g VS added and was about 12% higher than in the R_o-4:1,12% system which is significantly more than the slight difference in methane production, i.e. 3–4%. The small difference in methane production may be attributed to enhanced waste solubilisation in the two-stage digester, as reported in our previous study (Riungu et al. 2018b) and in agreement with related studies (Zuo et al. 2014; De Gioannis et al. 2017; Gaby et al. 2017). On the other hand, the observed higher biogas production in the two-stage digester likely can be ascribed to increased CO₂ dissolution, resulting from mixing the bicarbonate-rich methanogenic effluent for neutralisation purposes with the low pH (4.9) influent coming from the pre-acidification stage. The latter also explains the lower CH₄ content in the produced biogas in the gas bags of the two-stage digester and the higher pH in the effluent (Table 3). In the two-stage set-up, part of the produced acidity is already lost as CO₂ in the pre-acidification step that was open to air, leading to a higher overall alkalinity of the methanogenic effluent compared to the one-stage process.

In the two-stage digestion set-up with digestate recycling, the chances for possible acidification in the front part of the methanogenic plug flow digester is reduced. An active methanogenic activity in the front part of R_am-4:1,12% digester as indicated by a decline in total volatile acids (TVFA) concentrations between SP₁ and SP₃, and their subsequent conversion to biogas (Table 3) was observed. As mentioned before, the stabilised methanogenic conditions in the early stages of the plug flow digester of the two-stage set-up were achieved by digestate or effluent recycling that re-introduces active methanogenic biomass upfront (Cavinato et al. 2011). However, the R_o-4:1,12% digester showed an increasing trend in total volatile acids (TVFA), and non-dissociated volatile acids (ND-VFA) build-up between SP₁ and SP₂, resulting in an acidic pH 5.4 thus indicating predominating acidogenesis in the first part of the plug flow digester (Table 3). In this digester, highest methanogenic activity was observed between SP₂ and SP₃, where high reduction in TVFA indicated their subsequent conversion to biogas.

The impact of co-digestion on methane production in the one-stage digestion process was assessed by applying a UDDT:OMW ratio of 4:1, at 12% TS (R_o-4:1,12%) and a UDDT:OMW ratio of 1:0, at 12% TS (R_o-1:0,12%). Methane production in R_o-4:1,12% and R_o-1:0,12% digester system was 314 ± 15 and 228 ± 19 l CH₄/g VS added (Figure 5) respectively, representing a 37% increase when OMW was added. The corresponding percentage methane content in biogas in R_o-4:1,12% and R_o-1:0,12% digester system was 62.8 ± 2 and 70.0 ± 4.5% respectively. The lower CH₄ content in the biogas of the co-digester comes from the OMW fraction of the feed substrate which is generally characterised by carbohydrate-rich organic matter with a somewhat higher oxidation state than the UDDT-F, which agrees with the lower COD/TOC ratio for OMW as presented in Table 1. Also, the higher methane production (in L CH₄/g VS added) in R_o-4:1,12% is attributable to the highly digestible OMW fraction, reflected by the high TVFA build-up attained during the digestion process (Table 2). Intrinsically, during co-digestion, hydrolysis of OMW enhances TVFA build-up in the digestion medium (Zhang et al. 2005; Zhang et al. 2008; Riungu et al. 2018a), which is subsequently converted to biogas but may lead to subsequent system acidification (Angeriz-Campoy et al. 2015) if used as sole substrate.

The potential impact of substrate concentration on methane production was investigated applying two substrate concentrations, i.e. 12 and 10% TS. These concentrations were applied in both the one-stage and two-stage plug flow digesters, at UDDT-F:OMW ratios of 1:0 and 4:1 respectively. Using both digester systems, slightly higher methane production was observed at 12% TS than 10% TS. The two-stage digesters R_am-4:1,12% produced 325 ± 12 mL CH₄/g VS_added whereas the corresponding value in R_am-4:1,10% was 313 ± 17 mL CH₄/g VS_added. Similarly, in the one-stage digester, R_o-1:0,12% and R_o-1:0,10% the observed methane production was 228 ± 19 mL CH₄/g VS and 204 ± 22 mL CH₄/g VS_added respectively. The percentage methane in the biogas of R_am-4:1,12% and R_am-4:1,10% digesters at 10 and 12% TS were 57 ± 8 and 59 ± 4%, respectively (Figure 3). Corresponding values in R_o-1:0,12% and R_o-1:0,10% digesters being 70 ± 2 and 71 ± 8% respectively. Furthermore, all digesters showed stable digestion performance with effluent pH in the one-stage digesters a bit lower compared to the two-stage digesters, i.e. 7.3–7.7 and 7.6–8.1, respectively. The average pH in the methanogenic stage is considered optimal for the methanogenic biomass, i.e. 7.5–8.1.

This study is part of a wide research seeking to enhance biogas production and pathogen inactivation from UDDT-F in high-density informal slum settlements. The key objective of the research was to maximise the amounts of UDDT-F that can be treated, while producing biogas and stabilised digestate that can be used for agricultural applications. The findings obtained in this research demonstrated the technical feasibility of AD technology in UDDT-F management. Moreover, in addition to efficient management of the waste, the produced biogas has a wide range of applications.

The possibilities of reactor failure are apparent during co-digestion, especially at increased OMW fraction due to reactor acidification. The UDDT-F:OMW ratio 4:1 adopted in this study was based on recommendations from our previous study (Riungu et al. 2018a). Increasing the OMW fraction in the feed substrate leads to rapid acidification thereby lowering the pH and increasing the ND-VFA concentration that has a toxic effect not only to pathogens but all anaerobic bacterial population, thus process failure. As such, precaution should be taken to ensure application of optimal UDDT-F:OMW ratios during co-digestion.

Results showed that co-digestion in the proposed plug-flow digester produced a low pathogen content–digestate, i.e. <1*10³ CFU/100 mL (Riungu et al. 2018b)) and a biogas stream that can be used as an alternative fuel source for slum residents, delivering about 6,500 MJ/month for a bio-centre with a user load of 500 persons/day. The system presents a cost-effective solution for the many slum areas in sub-Saharan Africa: the plug-flow reactor can be assembled with locally available materials and the high population density assures a constant supply of raw materials, whereas the prevailing high temperatures ensure the system's zero energy operating requirements.

Experiments were conducted to investigate the biochemical energy recovery during digestion and co-digestion of faecal matter collected from urine diverting dehydrating toilet faeces (UDDT-F) and mixed OMW under laboratory- and pilot-scale conditions. Laboratory-scale BMP tests showed a positive correlation between methane production and increasing OMW fraction in the feed substrate.

Under pilot-scale conditions, comparable methane production was observed in one-stage (R_o-4:1,12%) (314 ± 15 mL CH₄/g VS added) and two-stage (R_am-4:1,12%) (325 ± 12 mL CH₄/g VS added) digesters, when applying 12% TS slurry concentration. However, biogas production in R_am-4:1,12% digester (571 ± 25 mL CH₄/g VS added) was about 12% higher than in the R_o-4:1,12%, significantly more than the slight difference in methane production, i.e. 3–4%. The increased methane and biogas production was attributed to enhanced waste solubilisation and increased CO₂ dissolution, resulting from mixing the bicarbonate-rich methanogenic effluent for neutralisation purposes with the low pH (4.9) influent coming from the pre-acidification stage. Moreover, compared to the one-stage reactor, higher process stability was observed in the first parts of the two-stage plug flow digester, characterised by lower VFA concentrations. The observed high VFA concentrations and acidic pH (5.4) in the first parts of one-stage digester indicate low process stability, particularly with increased OMW fractions in the feed substrate.

Within the wide research, overall findings have shown the potential application of two-stage AD technology in addressing the human waste menace, especially in high density slum settlements. The proposed system can be applied at either small or large scale, depending on available space. The treatment system has almost zero energy requirements when implemented in warm areas where optimal mesophilic temperatures can be reached without heating.

This research is funded by the Bill & Melinda Gates Foundation under the framework of SaniUp project (Stimulating local Innovation on Sanitation for the Urban Poor in Sub-Saharan Africa and South-East Asia) (OPP1029019). The authors would like to thank Ani Vabharneni, Sanergy Kenya, and DVC-ARS, Meru University, Kenya for their valuable support during this study.