Assessment of On-farm Anaerobic Digestion Performance under Variable Feedstock Mixes in Smallholder Farms

Thesis Abstract

Smallholder farms increasingly rely on on-farm anaerobic digestion (AD) to manage agricultural residues, reduce greenhouse gas emissions, and improve rural livelihoods; however, the performance of on-farm AD systems under variable feedstock mixes remains inadequately understood, limiting optimization and adoption. This study aims to evaluate how heterogeneous feedstock compositions influence biogas yield, methane content, digestate quality, and process stability in smallholder AD facilities integrated with mixed livestock and crop residues. The specific objectives are (i) to quantify the effect of feedstock mix ratios on hydraulic retention time, volatile solids reduction, and measured biogas yield; (ii) to characterize methane content, hydrogen sulfide levels, and digester stability indicators (pH, alkalinity) across feedstock scenarios; (iii) to assess digestate nutrient profiles and potential environmental risks linked to different mixes; (iv) to identify operational practices that mitigate inhibition and maximize gas production; and (v) to develop a practical decision-support framework for feedstock selection tailored to smallholder constraints. A mixed-methods approach is employed in a multi-site field study across 40 smallholder AD installations in a mid-latitude tropical agroecosystem. The population comprises smallholder farmers operating continuous-feed anaerobic digesters with variable livestock and crop residue inputs. A stratified sampling design yields 120 daily digester performance observations over a 12-month period, supplemented by 20 in-depth interviews with digester operators and 6 focus group discussions with farm households. Data collection instruments include calibrated biogas meters for volume and methane concentration, portable gas analyzers, online pH and redox sensors, and standardized digestate sampling protocols for nutrient analysis (N, P, K, total Kjeldahl nitrogen) and heavy metal screening. Feedstock characterisation involves proximate and ultimate analyses, moisture content, and biodegradability assessments (BOD/COD) to derive mix-ability indices. Validity and reliability are ensured through cross-validation of gas measurements with a secondary volumetric flowmeter, instrument calibration before each sampling session, and pilot testing of survey instruments. Data analysis employs a combination of regression modeling (multivariate linear and non-linear models) to quantify relationships between feedstock mix ratios and performance indicators, ANOVA to test differences among predefined mix categories (e.g., high-lignocellulosic vs. high-carbohydrate blends), and time-series analyses to capture seasonal effects. The analysis framework is anchored in Resource-Based View and Technology Acceptance Theory to interpret how feedstock diversity and user capabilities influence adoption and sustained operation. A conceptual model linking feedstock composition, biochemical methane potential, and digester stability is developed and tested. Expected findings indicate that increasing agricultural residue diversity enhances methane yield up to a threshold beyond which inhibition due to volatile fatty acid accumulation may occur, with optimum mix ranges dependent on substrate C/N balance and moisture content. Digestate will show improved nitrogen availability but varying phosphorus and potassium profiles, requiring context-specific land application guidelines to mitigate leaching risk. Process stability metrics (pH, alkalinity, ammonium concentration) are predicted to decline under high-lignocellulosic dominance unless buffer management and inoculum strategies are employed. The study anticipates considerable heterogeneity in performance across sites, underscoring the role of operator training, feedstock preprocessing, and day-to-day management practices in achieving reliable gas production. The contribution to knowledge encompasses (i) an empirical, field-based quantification of feedstock mix effects on on-farm AD performance in smallholder contexts, (ii) a validated feedstock-mix decision-support framework, and (iii) practical guidelines for optimal substrate blending, digester operation, and digestate management that integrate technical, economic, and environmental considerations. The main conclusion is that modestly varied feedstock mixes, when coupled with targeted operational controls, can significantly improve biogas outputs and process stability without compromising digestate quality. Recommendations include implementing routine feedstock characterization, developing farmer training modules on mix optimization and inhibitor minimization, and promoting policy support for affordable inoculants and monitoring equipment to enhance the scalability and resilience of on-farm AD in smallholder settings.