Enhancing CI engine performance and reducing emissions via fuel blend homogenization with organic co-solvents: a novel biofuel strategy

The imbalance between diesel fuel supply and demand, coupled with the escalating air pollution crisis, necessitates the development of alternative fuels. This study addresses the phase instability of ternary diesel–biodiesel–bioethanol blends by integrating organic co-solvents and employing a hybrid experimental-data-driven framework. While the optimized base ternary blend (D67B25E8) exhibited satisfactory combustion potential, it suffered from rapid phase separation. To resolve this, three oxygenated co-solvents—diethyl ether (DEE), 1-butanol, and tetrahydrofuran (THF)—were evaluated. Among these, a formulation comprising 90% (v/v) optimized base blend and 10% (v/v) DEE demonstrated superior physicochemical stability, maintaining complete phase homogeneity until the onset of phase separation at approximately 390 h, with complete phase separation occurring after approximately 428 h. Crucially, to isolate the intrinsic chemical effects of the fuel blends from engine control variables, all experiments were conducted under fixed injection timing (22° BTDC) and constant injection pressure (19.6 MPa). This methodological constraint allowed for a direct comparison of combustion kinetics influenced solely by fuel properties. A Support Vector Machine (SVM) model, trained on data across a wide BMEP range (0.52–5.85), demonstrated high predictive accuracy (R 2 > 0.95). Experimental results demonstrated that the optimized DEE-containing blend enhanced engine performance while simultaneously reducing exhaust emissions under the investigated steady-state laboratory operating conditions. The optimum formulation increased brake thermal efficiency (BTE) by 12.45% and reduced brake-specific fuel consumption (BSFC) by 4.23% relative to neat diesel at full load, while the maximum improvement in brake power reached 12.34% at medium load. In addition, the optimized fuel achieved simultaneous reductions in NO x (33%), smoke opacity (43%), CO (43.2%), and measured CO 2 emissions (4.79%), thereby substantially mitigating the conventional NO x –soot trade-off without requiring engine hardware modifications. Multi-objective optimization using NSGA-III confirmed DEE as the superior co-solvent. This study provides an experimental framework for developing stable biofuel blends and demonstrates that significant improvements in engine performance and measured exhaust emissions can be achieved under controlled steady-state laboratory operating conditions through fuel chemistry optimization. Verification under certified engine emission test cycles would be required before drawing conclusions regarding regulatory compliance.