Comprehensive analysis and prediction of biogas-diesel dual-fuel combustion in direct injection diesel engines: Experimental and artificial neural network approach

• Biogas-diesel dual-fuel system enhances engine performance, with up to a 7.9% increase in brake power and up to 17.8% improvement in brake specific fuel consumption (BSFC). • Experimental results show improved fuel efficiency and reduced emissions, particularly at part-load conditions. • A significant reduction in power output and thermal efficiency is observed at higher biogas substitution ratios (up to 37.7% decrease in brake power at full load). • Artificial Neural Network (ANN) model predicts engine performance with high accuracy (R² > 0.96), providing a reliable tool for real-time optimization of biogas-diesel dual-fuel engines. • Future work should address power loss at full load and explore the long-term effects of dual-fuel operation on engine components and wear. • These highlights summarize the key findings and future research directions of the study, emphasizing the benefits and challenges of biogas-diesel dual-fuel systems. Biogas–diesel dual-fuel systems provide a sustainable route to reducing fossil fuel dependence and mitigating greenhouse gas emissions in compression ignition engines. Their role is particularly significant in enhancing fuel efficiency and enabling renewable resource utilization, though power trade-offs remain a critical challenge. This study focuses on optimizing the performance and emission characteristics of a direct injection diesel engine (DIDE) operating under biogas–diesel dual-fuel conditions by integrating experimental testing with Artificial Neural Network (ANN) modeling. Experiments were conducted at substitution ratios ranging from 5% to 35% biogas under part-load (10 Nm) and full-load (20 Nm) conditions, with performance metrics including brake power, brake thermal efficiency (BTE), brake specific fuel consumption (BSFC), and exhaust emissions. Results revealed that biogas substitution significantly improved part-load performance: brake power increased by 3.18% at 15% biogas and 7.9% at 35% biogas, while BSFC decreased by 4.8% and 17.8% at 1500 rpm, indicating superior fuel utilization. Conversely, at full load, power output declined sharply, with brake power reductions of up to 37.7% at 35% substitution. To address these variations, an ANN model was developed and validated (R² > 0.96), achieving high predictive accuracy for brake power (MAE < 0.07 kW), BTE (MAE < 0.62%), and emissions. The dual contribution of this work lies in delivering a comprehensive experimental characterization of dual-fuel combustion across load regimes while providing a robust ANN-based predictive tool. Findings highlight the promise of biogas substitution for efficiency gains and emission reduction in part-load operations while acknowledging the trade-offs at higher substitution ratios.