Relation of organic fractions in fresh and stored fecal sludge and foodwaste to biogas production

The sanitation needs of one-third of the world's population are met through non-sewered sanitation (NSS) (WHO 2018), and NSS provides the majority of sanitation in low- and middle-income countries. However, NSS, in general, lacks adequate management; for example, in South Asia, less than 6% of fecal sludge undergoes any form of treatment, and its improper disposal poses a significant risk to public health and safety (Maqbool et al. 2022). Fecal sludge is defined as what accumulates during storage in onsite containments with NSS. A range of constructions are utilized for onsite containment, including unlined, partially lined, fully lined, and with or without overflows, and are commonly referred to as pit latrines, septic tanks, or cesspits (Gold et al. 2018). Fecal sludge consists of everything going into the containment, including excreta, flush water, cleansing material, gray water, and solid waste (Ahmed et al. 2018). The characteristics of fecal sludge arriving at treatment facilities are highly variable with up to two orders of magnitude higher organic strength (as chemical oxygen demand (COD) or volatile solids (VS)) as compared to municipal sewer-based wastewater. This is due to the wide range of inputs, management practices, and batch-wise delivery to treatment (Strande et al. 2018). The water content in fecal sludge is more than 70–80%, which is linked to challenges in transport, effective dewatering, and subsequent treatment (Gold et al. 2018). The difficulty in transporting fecal sludge through congested urban areas to treatment plants, together with highly variable characteristics, makes decentralized treatment plants an attractive option (Semiyaga et al. 2022). However, currently established technologies for the treatment of fecal sludge rely on passive technologies (e.g. settling-thickening tanks, or drying beds) and have large footprints (Tayler 2018; WHO 2018), making them unsuitable for densely populated urban areas. There is an urgent need for tenable solutions with low footprints for dense urban areas.

Anaerobic digestion is an extensively applied treatment process for the stabilization of wastewater sludge with the added benefit of biogas production (Tchobanoglus et al. 2014); however, to date, there is a lack of controlled experimental data on fecal sludge. Part of the problem is that it is not understood how much degradation fecal sludge undergoes during storage in containment. Existing studies report a range of biogas and biomethane production, including no biogas production from pit latrine sludge (PLC) (Madikizela et al. 2017), 32 mL of biomethane from pit latrine sludge (Cui et al. 2023), and 57–811 mL biogas/gVS from fresh fecal sludge (Sam et al. 2022). Based on the literature on anaerobic digestion of wastewater sludge, biogas production, and stabilization of fecal sludge have been investigated in relation to COD (Colón et al. 2015; Le Phuong & Thai 2018; van Eekert et al. 2019) and VS (Kilucha et al. 2022). However, based on the low and unpredictable performance of anaerobic digestion, there is a need for further understanding of the level of degradation of fecal sludge during storage, and the fractions of bioavailable organic matter remaining for treatment (Levira et al. 2023). Due to the lack of controlled experimental results, biomethane potential (BMP) tests are needed to understand the biogas potential and anaerobic biodegradability of fecal sludge (Filer et al. 2019). In addition, there is limited knowledge of microbial communities present in fecal sludge, and the role they play is not clear (Sam et al. 2022; Ward et al. 2023).

The objective of this study was to evaluate the kinetics of anaerobic degradation with BMP tests for a range of fecal sludge, in order to gain an understanding of whether different organic fractions reflecting different levels of stabilization can predict biogas potential. The experimental setup included mono-digestion of fresh, stored, and dewatered fecal sludge, and co-digestion with fresh food waste. Dewatering performance and particle size distribution were evaluated during the BMP tests, and microbial communities were evaluated. The analysis included volatile solids to total solids (VS/TS), total organic carbon to total nitrogen (TOC/TN), BOD/COD as metrics of stabilization, and hydrolysis rate coefficients for COD (K_COD) soluble COD (K_sCOD), polysaccharides (K_PS), and degradability extent (f_d).

Eight fecal sludge samples were collected during January 2022 from household onsite containments located in Islamabad, Pakistan, four from pit latrines and four from septic tanks. The collected four pit latrine sludges (PL₁, PL₂, PL₃, and PL₄) were mixed in equal proportions to produce a composite pit latrine sludge (PL_C). Other than PL_C, pit latrine sludge 1 (PL₁), pit latrine sludge 2 (PL₂), and septic tank sludge 1 (ST₁) were used. The selection of fecal sludge samples for BMP tests was based on source (residential and commercial), time in containment (<1, <10, and >10 years), and TS concentration (<10, 10–50, and >50 g/L). One sample was collected from a restaurant septic tank sludge (ST_R) in Jalpaiguri, India in January 2022. One sample was collected from a household septic tank in Switzerland in April 2022, was dewatered as described in Shaw et al. (2022) and referred to as dewatered septic tank sludge (ST_D). The grab composite sampling method was used to collect samples (Velkushanova et al. 2021). Fresh fecal sludge (FS_F) was prepared by mixing feces and urine (1:2.5) collected in Eawag, Switzerland in May 2022 and diluted to 5% as described in Sam et al. (2022). In this study, fresh was defined as what would be coming directly from a toilet and not stored in containment. PL₁, PL₂, PL_C, ST₁, ST_D, ST_R, and FS_F were used as feed during BMP tests. All fecal sludge samples were stored at 4 °C and shipped (excluding FS_F, which was obtained from Eawag) to Eawag for characterization and subsequent use in BMP experiments. Inoculum was collected from an anaerobic digester at the Neugut wastewater treatment plant in Zurich, Switzerland. Readily available organic content in waste from green markets and restaurants makes it an ideal co-substrate for decentralized anaerobic digestion. Synthetic green market waste (GW_S) and synthetic restaurant waste (RW_S) were prepared according to Song et al. (2020) and Carmona-Cabello et al. (2020), respectively, and their recipes are included in the supplemental information (Tables S1 and S2). Restaurant food waste (RW_R) was collected from the restaurant at Eawag and comprised rice, pasta, vegetables, and meat. The RW_R was manually shredded and blended with a Phillips Avance Collection Blender (HR2096, The Netherlands). GW_S, RW_S, and RW_R were used as co-feed during BMP. Table 1 presents the characterization of feed, co-feed, and inoculum used in the study.

Table 1

Characterization of feeds pit latrine sludges (PL₁, PL₂), composite pit latrine sludge (PL_C), septic tank sludge (ST₁), restaurant septic tank sludge (ST_R), fresh fecal sludge (FS_F), dewatered septic tank sludge (ST_D), restaurant food waste (RW_R), synthetic green market waste (GW_S), synthetic restaurant food waste (RW_S), and anaerobic digester inoculum

Sample .	Time since the last emptying .	TS (g/L) .	VS (g/L) .	VS/TS .	(mg/L) .	sCOD (g/L) .	COD (g/L) .
PL1	7 years	4.53 ± 0.06	1.71 ± 0.07	0.38	547 ± 5.77	0.78 ± 0.009	79.67 ± 5.42
PL2	14 years	5.81 ± 0.08	2.43 ± 0.13	0.42	295 ± 5.13	0.75 ± 0.002	79.95 ± 6.58
PLC	–	19.07 ± 0.23	8.57 ± 0.06	0.45	430 ± 5	0.33 ± 0.001	14.78 ± 0.38
ST1	1 year	0.99 ± 0.11	0.37 ± 0.18	0.38	105 ± 2.89	0.71 ± 0.027	5.78 ± 0.23
STR	3 months	54.45 ± 4.61	40.06 ± 2.25	0.74	170 ± 2.08	5.66 ± 0.045	66.77 ± 0.76
FSF	0 years	64.98 ± 0.29	55.15 ± 0.25	0.85	2,130 ± 42	28.85 ± 0.3	118.75 ± 8.13
STD	2 years	139.31 ± 0.56	111.34 ± 0.08	0.80	164 ± 3	5.50 ± 0.022	252.75 ± 3.18
RWR	0 years	336.93 ± 8.80	323.19 ± 8.6	0.96	61 ± 1	72.50 ± 3.13	513.00 ± 23.81
GWS	–	0.98 ± 0.002	0.97 ± 0.05	0.96	–	–	–
RWS	–	0.99 ± 0.005	0.98 ± 0.03	0.98	–	–	–
Inoculum	–	21.94 ± 0.87	14.19 ± 0.41	0.65	938 ± 31.82	0.68 ± 0.01	18.43 ± 0.70

Sample .	Time since the last emptying .	TS (g/L) .	VS (g/L) .	VS/TS .	(mg/L) .	sCOD (g/L) .	COD (g/L) .
PL1	7 years	4.53 ± 0.06	1.71 ± 0.07	0.38	547 ± 5.77	0.78 ± 0.009	79.67 ± 5.42
PL2	14 years	5.81 ± 0.08	2.43 ± 0.13	0.42	295 ± 5.13	0.75 ± 0.002	79.95 ± 6.58
PLC	–	19.07 ± 0.23	8.57 ± 0.06	0.45	430 ± 5	0.33 ± 0.001	14.78 ± 0.38
ST1	1 year	0.99 ± 0.11	0.37 ± 0.18	0.38	105 ± 2.89	0.71 ± 0.027	5.78 ± 0.23
STR	3 months	54.45 ± 4.61	40.06 ± 2.25	0.74	170 ± 2.08	5.66 ± 0.045	66.77 ± 0.76
FSF	0 years	64.98 ± 0.29	55.15 ± 0.25	0.85	2,130 ± 42	28.85 ± 0.3	118.75 ± 8.13
STD	2 years	139.31 ± 0.56	111.34 ± 0.08	0.80	164 ± 3	5.50 ± 0.022	252.75 ± 3.18
RWR	0 years	336.93 ± 8.80	323.19 ± 8.6	0.96	61 ± 1	72.50 ± 3.13	513.00 ± 23.81
GWS	–	0.98 ± 0.002	0.97 ± 0.05	0.96	–	–	–
RWS	–	0.99 ± 0.005	0.98 ± 0.03	0.98	–	–	–
Inoculum	–	21.94 ± 0.87	14.19 ± 0.41	0.65	938 ± 31.82	0.68 ± 0.01	18.43 ± 0.70

View Large

BMP tests were carried out in 250 mL bottles with 70% working volume. Three sets of BMP tests were conducted: (i) mono-digestion of fresh (FS_F), stored (PL₁, PL₂, PL_C, ST₁, ST_R), and dewatered fecal sludge (ST_D), (ii) co-digestion of fresh (FS_F), dewatered (ST_D), and stored (PL_C) with RW_R individually and stored (PL_C) fecal sludge with GW_S and RW_S, respectively; and (iii) effect of micronutrients addition on mono-digestion (in 1× and 2× concentrations represented as PL_C-MN1 and PL_C-MN2, respectively) and co-digestion of fecal sludge (PL_C-MN1+GWS and PL_C-MN1+RWS). The inoculum-to-feed (I/F) ratio was maintained at two for all the experiments. BMP tests for positive and negative controls were run to ensure the reliability and accuracy of the results obtained from experimental samples. The positive control was inoculum and microcrystalline cellulose (control substrate) and the negative control was inoculum only with no substrate added. BMP bottles were purged with N₂ gas to maintain the anaerobic environment and placed at 37 °C and 100 rpm in a 5,000 L orbital shaking incubator (VWR, Avantor, USA). Triplicates of BMP bottles were used, and two sets of extra sacrificial bottles were used for weekly analysis during the BMP tests. Micronutrients (NH₄Cl, 1,000 mg/L; NaCl, 100 mg/L; MgCl₂·6H₂O, 100 mg/L; CaCl₂·2H₂O, 50 mg/L; K₂HPO₄·3H₂O, 400 mg/L; FeCl₂·4H₂O, 2 mg/L; H₃BO₃, 0.05 mg/L; CuCl₂·2H₂O, 0.038 mg/L; ZnCl₂, 0.05 mg/L; MnCl₂·4H₂O, 0.05 mg/L; (NH₄)₆Mo₇O₂₄·4H₂O, 0.05 mg/L; AlCl₃, 0.05 mg/L; CoCl₂·6H₂O, 0.05 mg/L; NiCl₂·6H₂O, 0.092 mg/L, ethylenediaminetetraacetate, 0.5 mg/L; Na₂SeO₃·5H₂O, 0.1 mg/L; HCl conc. 0.001 mL/L) were added (Angelidaki et al. 2009) to determine the effect of their addition on anaerobic digestion and biogas production. Biogas was measured daily with the water displacement method (Filer et al. 2019), and gas volume produced by negative control was subtracted and normalized to VS (g/L) in the BMP bottle. The BMP test was stopped when the daily biogas volume reduced to <1% of the cumulative biogas (CBG) for 3 consecutive days (Filer et al. 2019). CBG produced from the negative control and positive control was 28.67 ± 1 and 892 ± 17 mL/gVS, respectively.

pH and electrical conductivity (EC) were determined using a multi-parameter portable meter (WTW ProfiLine pH/Cond 3320, Germany). Total solids (TS) were determined volumetrically by drying the sample at 105 °C and VS by combusting the residue obtained from the TS at 550 °C until a constant weight was achieved (Velkushanova et al. 2021). COD, soluble chemical oxygen demand (sCOD), TN, and ammonium nitrogen were determined using Hach Lange test kits according to the manufacturer's instructions based on standard methods in American Public Health Association (APHA) standard methods 5220 D, 4500-N C, and 4500-NH3 F, respectively (APHA 2017). The 5-day biological oxygen demand (BOD₅) was determined using APHA standard 5210–B at 20 °C (APHA 2017). Alkalinity was determined by titration and volatile fatty acids (VFA) were analyzed using an ion chromatograph (Shimadzu 881 compact IC pro, Japan). TOC was analyzed using a TOC analyzer (TOC-LCPH Shimadzu, Japan). Polysaccharide concentrations were determined using the colorimetric method with the anthrone assay method as described by Loewus (1952). pH, EC, VFA, alkalinity, TN, and were used to evaluate process inhibition. TS was used as an indicator of total organic and inorganic matter, while VS, COD, polysaccharides, and TOC for pools of organic matter, BOD₅ for biologically degradable organic matter and sCOD for soluble organic matter only. pH, TS, VS, COD, sCOD, ⁠, and polysaccharides were measured weekly and TN, BOD₅, and TOC were measured at the start and end of experiments.

To evaluate the dewatering performance of fecal sludge, supernatant turbidity following centrifugation and capillary suction time (CST) were monitored on a weekly basis. Samples were centrifuged at 3,300 × g for 20 min and decanted and supernatant turbidity was quantified with a turbidity meter (Hach TL 2300, USA) (Ward et al. 2021). CST was measured in quadruplicate with a CST apparatus (Triton 319, Canada) as described in Velkushanova et al. (2021). Particle size distribution was analyzed with the laser light scattering method with a laser diffraction particle analyzer (Beckman Coulter LS 13 320, USA) using universal liquid module (ulm) to determine how particle size changes during anaerobic digestion. In this study, the common stabilization indicators VS/TS, TOC/TN, and BOD/COD were used to evaluate stabilization. Relative decreases in VS and BOD indicate that degradation is occurring, and TOC/TN indicates nitrogen availability for the growth of microbes with an ideal range of 20–30 to utilize carbon for energy and nitrogen for cellular structure (Tayler 2018).

To evaluate the microbial community in BMP tests, samples were taken from selected BMP bottles (FS_F, PL₁, PL₂, PL_C-MN1, ST_D, and ST_R) at time 0 days and anaerobic digester sludge inoculum. 2 mL of sample was centrifuged at 6,000 × g for 10 min. The supernatant was discarded and 1 mL of RNAlater was added to each sample. The samples were stored at −20 °C until DNA extraction. DNA extraction was performed according to the modified method by Sam et al. (2022). 16S rRNA gene amplicon sequencing was carried out by Novogene as described in Sam et al. (2022). The plots of dominant microorganisms at the phylum and genus level were plotted for each sample.

The observed differences in CBG in this study resulted from feeds and not the inoculum, as the same inoculum and I/F were used throughout the study. Based on the results of pH, EC, alkalinity, VFA, and (Table S3), no inhibition was observed. The pH was always in the acceptable range of 7.04–8.08. The alkalinity was adequate and between 1,500 and 3,500 mg/L for all BMP tests other than ST₁. The VFA to alkalinity ratio was always <0.4, including for ST₁, which indicates stable BMP performance (Filer et al. 2019). The VFAs were never high enough to be inhibitory and were even as low as zero, which indicates their production and consumption at the same time. Low VFA production with rapid by methanogens is also observed during anaerobic digestion of blackwater due to a low hydrolysis rate (Elmitwalli et al. 2011). Overall, there was an increase in ⁠, EC, and alkalinity from time 0 to time 21. As organic nitrogen in proteins and urea are broken down, is released, which subsequently increases alkalinity (Adou et al. 2020). However, the concentration of was never high enough to cause inhibition (Colón et al. 2015).

Metrics of stabilization for BMP tests at time 0 are reported in Table 2. The VS/TS in BMP tests at time 0 with an I/F ratio of two was relatively consistent (VS/TS of feeds are presented in Table 1). It was higher than 0.49 reported for fecal sludge alone (Andriessen et al. 2023), and lower than primary wastewater sludge (0.60–0.80), or waste-activated sludge (0.59–0.88) (Tchobanoglus et al. 2014; Odirile et al. 2021). The observed TOC/TN values of the BMP tests at time 0 were consistent with literature values for fecal sludge of 1.5–6 (Meher et al. 1994; Manga et al. 2022), and lower than primary (13.6) and waste-activated sludge (9.2) (Sakaveli et al. 2021). The fecal sludge that had been stored had relatively lower BOD/COD values (0.03–0.07) as compared to the fresh waste streams FS_F and RW_R (0.09–0.10). A wide range of BOD/COD values have been reported for fecal sludge, including 0.18–0.62 in Argentina, 0.08–0.44 in Ghana, 0.35 in Palestine, 0.16 in Burkina Faso, and 0.10 in the Philippines (Tayler 2018). A BOD/COD of 0.5 is typical for domestic wastewater, with 0.31 reported for primary sludge and 0.11 for waste-activated sludge (Metcalf et al. 2014).

Table 2

Metrics of stabilization (VS/TS, TOC/TN, BOD/COD) for BMP tests time 0, hydrolysis rate coefficients for COD (K_COD), sCOD (K_sCOD) PS (K_PS), and degradability extent (f_d) during mono- and co-digestion of fecal sludge

The color scale from darker to lighter shading illustrates higher to lower values.

View Large

The first-order hydrolysis rate coefficients were calculated for COD (K_COD), sCOD (K_sCOD), and polysaccharides (K_PS) with Equation (1) and are shown in Table 2. K_sCOD had the highest values, followed by K_PS and K_COD. All the observed rate coefficients reported in Table 2 were toward the lower end of the values reported in the literature for wastewater 0.09–0.12 day⁻¹ (Elmitwalli et al. 2011), wastewater sludge 0.077–0.15 day⁻¹ (Pavlostathis & Giraldo-Gomez 1991), cellulose 0.29–0.42 day⁻¹ (Jensen et al. 2011), and primary 0.13 day⁻¹ and waste activated sludge 0.11 day⁻¹ (Abubakar et al. 2017), indicating that hydrolysis was limiting fecal sludge degradation, and that fecal sludge is less digestible than wastewater sludges. The degradation extent (f_D) increased from stored fecal sludge (0.002–0.050), to fresh fecal sludge (0.082), to co-digestion of stored fecal sludge and food waste (0.052–0.084) and fresh fecal sludge and food waste (0.103) (Table 2). These values were all low in comparison to the theoretical degradation extent (yield) of 0.35 L_CH4/g_COD (Filer et al. 2019), and 0.28 L_CH4/g_COD that has been reported from a simulant feces (Colón et al. 2015).

The dewatering metrics over time 0–21 days are presented in Table S3 and Figure S1. After the first week of the BMP test, there was a discernible improvement in CST and turbidity with no significant changes afterward. A range of CST (1.88–11.68 sL/gTS) was observed at time 0 of the BMP test with a 42–79% reduction over 21 days. All samples had improved dewatering performance during the first 7 days (other than PL_C-MN1), whereas from time 7 to time 21, changes in dewatering performance did not follow consistent trends, from no overall change to continued improvement. In addition, there were no significant differences in overall improved dewatering for fresh (43–45%) and stored (32–63%) waste streams upon anaerobic digestion.

At time 0, supernatant turbidity values were 24-656 Nephelometric Turbidity Units (NTU) with a 28–79% reduction over 21 days. Supernatant turbidity also had a clear decrease during the first week, with no significant overall changes from time 7 to time 21. There were clear improvements in turbidity for fresh (49–60%) and stored (32–60%) fecal sludge during anaerobic digestion. In comparison, Sam et al. (2023) observed that stored fecal sludge had lower turbidity and CST than fresh fecal sludge (not stored), which could be likely due to partial stabilization during storage in containments. In comparison, anaerobic digestion of activated sludge typically worsens dewaterability (Houghton et al. 2000; Christensen et al. 2015) and anaerobic digestion of primary wastewater sludge shows inconsistent results.

Particle size distribution (PSD) over time 0–21 days is presented in Figure S2. The peaks were mainly unimodal, and at time 0, the peak of the curve was between 84 and 110 μm, comprising 3.9–5.1% of the sample volume. Based on a comparison of the peaks from time 0 to time 21, no major differences were observed (other than ST_D); therefore, no strong relation was seen between PSD and dewatering metrics. In comparison, Sam et al. (2022) reported an increase in supracolloidal particles (1–100 μm) and a decrease in larger particles during anaerobic storage of fecal sludge; however, no relationship between supracolloidal particles and turbidity or CST was observed.

In this study at the genus level, Methanosaeta (Halobacterota phylum) had the highest relative abundance of 36% in PL_C-MN1 and 21% in ST_R. The highest abundance of Methanosaeta (Halobacterota phylum) was observed in AD inoculum (12%), PL₁ (14%), and PL₂ (13%), while their lowest abundance was found in ST_D (6%). Longilinea (Chloroflexi phylum) had the highest relative abundance of 36% in FS_F.

Based on the results of this study, it appears that fecal sludge is, in general, less biodegradable in anaerobic conditions than wastewater sludge. An average reduction of total VS over 21 days of 26% was observed for fresh fecal sludge, 15% for co-digested fecal sludge with food waste, and 11% for mono-digestion of fecal sludge that had been stored in containments (Table S3). Similar trends with fresh fecal sludge have been reported by Sam et al. (2022) (20% VS removal) and Ward et al. (2023) (20% VSS removal). In contrast, a 30–50% reduction in VS (due to limiting hydrolysis) is considered low for wastewater sludge (Appels et al. 2008). In contrast to fecal sludge, primary wastewater sludge is mainly composed of settled solids from the transport of fresh excreta, and waste-activated sludge is biomass that has been growing in aerobic conditions. As demonstrated with PLc, the limited degradation does not appear to be due to a limitation of micronutrients, but rather the overall degradability; however, this should be further verified by comparing the effect of different inoculums (Koch et al. 2020). This is also supported by the degradation extent (f_D), which was quite low in comparison to the theoretical yield.

In tests with greater CBG, a biphasic biogas production was observed, which can be attributed to the sequential consumption of readily and then slowly biodegradable organic components, as supported by sCOD, BOD, and polysaccharides and their K values (Table 2). A 66% sCOD removal was observed for both FS_F and RW_R, which had initial sCOD concentration three to six times higher than fecal sludge that had been stored in containments. The fresh waste streams with higher sCOD and polysaccharides had higher K values, whereas mono-digestion of stored fecal sludge had the lowest K values. The lower K values indicate that hydrolysis is limiting the biogas production after the readily degradable components are used up (Pavlostathis & Giraldo-Gomez 1991). The results do not fit well into the ADM1 anaerobic digestion model due to the high variability of fecal sludge and varying hydrolysis rates of slowly and readily biodegradable fractions (Batstone et al. 2002). The ADM1 model uses fixed fractions of proteins (30%), lipids (30%), carbohydrates (30%), and particulate inert material (10%), which are not established for highly heterogeneous fecal sludge samples. The development of models for the anaerobic digestion of fecal sludge will first require a more detailed understanding of the readily and slowly biodegradable particulate and dissolved fractions of organic matter including lipids, proteins and carbohydrates in these pools.

The greatest degradation of sCOD was observed with co-digestion of fecal sludge and foodwaste, with a 74% sCOD removal with co-digestion of FS_F + RW_R. The synergistic benefit of increased CBG was not explained by the concentration of the sCOD, BOD, and polysaccharides (Table S3) in the two feeds, and CBG was higher than either when digested alone (Koch et al. 2020), indicating a priming effect due to addition of readily degradable organic matter (Insam & Markt 2016). The concept of priming is not commonly reported in anaerobic digestion; however, it has been reported with slowly biodegradable organic matter becoming readily bioavailable upon co-digestion (Vivekanand et al. 2018). The priming effect in anaerobic reactors enhances microbial enzyme production with the consumption of readily degradable organic matter, allowing for the decomposition of slowly biodegradable compounds and an increase in the total production of biogas (graphical abstract & Figure 1(c)) (Insam & Markt 2016). This has also been observed with BMP tests of septic tank sludge and food waste at ambient temperatures (Le Phuong & Thai 2018), and wastewater sludge and municipal solid waste (Nielfa et al. (2015). Further research is needed to identify the enzymes involved and the underlying mechanism in order to quantify the effect.

It is commonly perceived that fecal sludge steadily undergoes anaerobic digestion and stabilization with storage time (or time since last emptied) in onsite containments (Cofie et al. 2006). However, the effect of longer-term storage on stabilization of fecal sludge remains unclear, following the first week of storage or digestion where readily degradable organic matter is used up (Ward et al. 2023). Rose (2015) had similar results, with a CBG (276 ± 151 mL/gVS) from fecal sludge that had been stored less than 1 week (i.e. 4 days), in comparison to samples that had been stored in containment with time since last emptied of 2–10 years (50.6 ± 19.4 mL/gVS), with the majority of methane production ceasing within 10 days. Fresh food waste is also known to undergo rapid degradation during storage at ambient conditions, and is therefore not recommended prior to AD (Påledal et al. 2018). In our study, storage time in containment after more than 1 week also did not show any clear pattern with CBG, for example with 3 months since last emptied (ST_R, 37 ± 4 mL/gVS), compared to 14 years (PL₂, 74 ± 4 mL/gVS). This has also been observed with controlled anaerobic storage of fecal sludge in laboratory reactors, revealing no clear trend for stabilization based on storage time in reactors after the first week (Sam et al. 2022; Ward et al. 2023). In the field, time since last emptied does not indicate the actual storage time or age in containment of the sludge, as there are continual fresh inputs, variable emptying patterns, and storage conditions that are not analogous to controlled anaerobic digestion (Shaw & Dorea 2021). Further confirming this, there is an analysis of fecal sludge from 450 onsite containments in three countries, which found no relation to time since emptied and level of stabilization (Ward et al. 2021; Ward et al. 2023).

As time since last emptied is not a reliable predictor of stabilization, we need more dependable metrics for process control to predict the treatment performance of AD. Conventional wastewater parameters such as TS, VS, TOC, and COD are used to predict the stabilization and biogas potential of wastewater sludge (Odirile et al. 2021), but are not adequate to predict the biogas production from fecal sludge. As illustrated in the graphical abstract, these indicators measure overlapping pools of organic matter that have a direct impact on the production of biogas. In this study, the VS/TS values in the BMP tests at time 0, which were relatively consistent, were not predictors of biogas production, as shown by the lower f_d values for fecal sludge in comparison to fresh waste streams and co-digestion (Table 2). The feeds alone had differing VS/TS that were more reflective of biogas potential, however only at the level of fresh waste streams versus stored fecal sludge, and VS/TS of the stored fecal sludge feed alone did not have clear patterns in relation to f_d (Table 1). TOC/TN was also not a predictor of biogas production from fecal sludge. Comparative metrics like VS/TS and TOC/TN are more useful for AD of municipal wastewater, as both primary (settled solids) and waste-activated sludge (biomass) generated during wastewater treatment are more similar to each other than fecal sludge. Fecal sludge in contrast to municipal wastewater treatment processes has a wide range of inputs and storage conditions, resulting in fractions of readily and slowly biodegradable organic matter making up pools of VS, COD, and TOC also being highly variable. In addition, TS of fecal sludge contains varying amounts of inert material from soil or non-biodegradable wastes (Ahmed et al. 2018), whereas wastewater commonly undergoes grit removal. Because TOC measures both readily and slowly degradable organic carbon, the TOC/TN of wastewater also does not always correlate with stabilization or biogas production (Manga et al. 2022).

In contrast, metrics that are based on the potential for biological activity will be more reliable indicators of stabilization and biogas potential. BOD measures biodegradable organic matter, sCOD measures the soluble organic fraction, and polysaccharides include more readily degradable organic content (Płuciennik-Koropczuk & Myszograj 2019) (Table S3). In this study, the BOD/COD ratio of BMP tests at time 0 was a better indicator of f_D values (Table 2) (Adou et al. 2020), and the higher the ratio, the greater the biodegradability (Tchobanoglus et al. 2014). Similarly, Levira et al. (2023) also observed that the treatment performance of two similar mesophilic anaerobic fecal sludge digesters in Tanzania was significantly different and could not be predicted based only on the chemical parameters COD, total suspended solids, and free ammonium nitrogen. One anomaly in this study was ST_R, which had a high sCOD but a low f_d (0.006), the reasons for the low CBG with this sample are not known. Since the sample was from a restaurant, it is possible that the fractions of sCOD were comprised of less biodegradable fractions such as high contents of oil and grease, or inhibition could be due to cleaning products (Krueger et al. 2021). It is important to keep in mind that non-household sources of fecal sludge can account for half of the fecal sludge in a city and can have different characteristics than household sludge (Strande et al. 2018).

The dewatering metrics followed the same pattern as stabilization, with dewatering performance initially improving the first week as readily biodegradable organic matter was used up and then plateauing, with any further hydrolysis of particulate organic matter having no significant relation to dewaterability. This is similar to the findings of Shahid et al. (2022), who observed improved CST of simulant fecal sludge under anaerobic conditions. Semiyaga et al. (2017) also observed that during the early stages of anaerobic digestion, fine colloidal particles were degraded which slightly improved dewatering.

Regardless of the high presence of methanogens Halobacterota at the phylum level and Methanosaeta at the genus level in PL_C-MN1 and ST_R, the biogas production was still low. The presence of Halobacterota being a slow-growing organism (Lyu et al. 2018) also suggests that the limiting step in biogas production during anaerobic digestion of fecal sludge that had been stored in containment is not methanogenesis but rather the breakdown of organic material through hydrolysis.

Based on the results of this study, anaerobic digestion can be a viable treatment option for fresh fecal sludge with little or no retention time in onsite containments, such as container-based sanitation, or public toilets if the is not too high. Although in this study the stored fecal sludge still had significant organic matter, the readily biodegradable fraction was low and biogas production was limited by hydrolysis, indicating that anaerobic mono-digestion is not an optimal technology for biogas production or the stabilization of stored fecal sludge. However, the co-digestion of fecal sludge with other waste streams that have more bioavailable organic matter such as fresh food waste could result in a priming effect with increased biogas production (Insam & Markt 2016). Co-digestion also has the potential to buffer the variability of fecal sludge coming into the treatment (Levira et al. 2023). Alternatively, thermal pretreatment, or pretreatment with strong oxidants and alkalis resulting in membrane disruption, could potentially enhance the hydrolysis of fecal sludge by converting the slowly biodegradable organic matter into readily bioavailable fractions (Cui et al. 2023). Further knowledge of the pools of organic fractions arriving at fecal sludge treatment facilities, including aggregates and bulk solutions, will lead to an understanding of how it performs with treatment technologies. This includes identifying reliable metrics in order to predict stabilization, biogas production and dewatering performance instead of relying on VS, COD and TOC as metrics of total organic matter, versus what is available for biological degradation (BOD and sCOD). With implementation, it is important to consider that fecal sludge is highly variable, and anomalies observed as ST_R in this study are to be expected.

Key conclusions of this study include:

• Anaerobic digestion of fecal sludge for biogas production is not a recommended treatment option for fecal sludge that has undergone storage in containment due to hydrolysis limitation. However, it could be applied with short retention times to improve dewatering, or in co-digestion with fresh foodwaste.

• Anaerobic digestion for biogas production could be viable for feces, excreta, or blackwater that has not been stored or has been stored for less than 1 week.

• Following 1–2 weeks of storage in containment, the time since last emptied is not a predictor of the level of stabilization of fecal sludge.

• Physical–chemical metrics of stabilization (i.e. VS/TS, TOC/TN) are not reliable predictors of biogas production from fecal sludge, and instead, metrics based on the potential for biological degradation should be used (e.g. BOD, sCOD).

• Dewatering performance (filtration and turbidity) improves following 1 week of anaerobic digestion, but then levels off.

The authors appreciate Anik Dutta, Helena Verloo, and Michael Vogel for providing ST_R and ST_D samples, Sylvia Richter and Kelsey Shaw for laboratory assistance, and Barbara Jeanne Ward for critical feedback.

This work received support from Swiss National Science Foundation (SNF), Eawag Partnership Program (EPP), ETH for Development (ETH4D) and International Research Support Initiative Program (IRSIP), Higher Education Commission (HEC), Pakistan.

N. M. conceptualized the research, supported in funding acquisition, developed the methodology, conducted the research, analyzed data, and wrote the original draft. S. S. developed the methodology, contributed to the research, reviewed, and edited the article. S. J. K. reviewed and edited the article. L. S. conceptualized the research, supervised the work, supported in funding acquisition, validated the data, and wrote the original draft.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.