The optimization of pellet fuels composed of rice straw, mustard straw, and sawdust was investigated in the present study to improve their properties and utility. Response surface methodology (RSM) and an artificial neural network (ANN) integrated with a multi-objective genetic algorithm (MOGA) were applied to optimize pellet composition for enhanced heating value and minimized ash, nitrogen, and sulfur content. An optimal blend of 74.40% rice straw, 15.60% mustard straw, and 10% sawdust was identified by RSM. These proportions were closely approximated by the MOGA-ANN model within ±1%, and the results were confirmed through experimental validation. Combustion ion chromatography was also used, to analyze the biomasses and the optimized blend, revealing reduced chloride (4189 mg/kg) and sulfur (2716 mg/kg) levels. These results were validated subsequently through experimental tests, confirming the accuracy of the proposed models. A techno-economic analysis indicated that a generation cost of Rs. 10.71 per unit would be associated with a fully agro-residue-based power plant, while less than Rs. 5.28–Rs. 5.31 would be the cost of generation per unit of electricity observed with 5% biomass co-firing in thermal plants. This study demonstrates that improved fuel quality and economic feasibility for biomass power generation can be achieved through strategic biomass blending and co-firing. These findings demonstrated that the blending of various biomass can be a viable strategy for enhancing the characteristics of pellet fuels on an industrial scale.

Optimization of mixed agro residue pellets for enhanced fuel quality and economic feasibility
    Abstract
        Graphical abstract
    1 Introduction
    2 Materials and methods
        2.1 Collection and pretreatment of agro-residues
        2.2 Experimental planning and design
            2.2.1 Optimal mixture design
            2.2.2 Screening of variables and constraints
        2.3 Experimental method
            2.3.1 Selection of compaction techniques
            2.3.2 Characterization of pure and mixed biomass pellets
        2.4 Analysis of experimental data and optimization
            2.4.1 RSM-based statistical analysis
            2.4.2 RSM-based optimization
            2.4.3 ANN-based data analysis
            2.4.4 Optimization using ANN-MOGA
            2.4.5 TOPSIS to determine the optimum solution
    3 Results and discussion
        3.1 Characterization of the pure sample
        3.2 Relationship between factors and responses
        3.3 RSM-based statistical analysis for biomass mixtures
        3.4 Optimization of process parameters using desirability function in RSM
        3.5 Optimization of process parameters using ANN-MOGA
        3.6 Validation of optimized results from RSM and MOGA-ANN with experimental responses
        3.7 Combustion ion chromatography
    4 A generalized view on techno-economics and combustion efficiency in co-fired and 100% agro-residue based thermal power plants
    5 Conclusion
    Acknowledgements 
    References

• Arghya Datta(Chemical and Electrochemical Conversion Division, Sardar Swaran Singh National Institute of Bio Energy (An Autonomous Institute of Ministry of New and Renewable Energy, Government of India), Punjab, India)
• Anil Kumar Sarma(Chemical and Electrochemical Conversion Division, Sardar Swaran Singh National Institute of Bio Energy (An Autonomous Institute of Ministry of New and Renewable Energy, Government of India), Punjab, India)
• Bhautik Gajera(Chemical and Electrochemical Conversion Division, Sardar Swaran Singh National Institute of Bio Energy (An Autonomous Institute of Ministry of New and Renewable Energy, Government of India), Punjab, India, Centre for Energy and Environment, Dr B R Ambedkar National Institute of Technology, Jalandhar, Punjab, India)
• Mithilesh Kumar Jha(Centre for Energy and Environment, Dr B R Ambedkar National Institute of Technology, Jalandhar, Punjab, India)