Geothermal plant operation and control under demand uncertainties

Geothermal energy plant operations are significantly influenced by uncertainties in key parameters and processes, including the variability heating demand in the built environment compared to the more stable demand in horticulture, necessitating a robust framework for real-time decision-making. This paper introduces a novel robust optimization framework to enhance geothermal plant performance under uncertain heating demand. The proposed method integrates a genetic algorithm with a geothermal plant simulator, optimizing dual objectives: emission reduction and profit maximization. Operational constraints are incorporated via penalties in the objective function. The approach identifies distinct control strategies for each objective, effectively capturing varying operational behaviors and demonstrating adaptability to different performance goals. Results from the numerical case study indicate that, under the considered modelling assumptions, robust optimization delivers more resilient and effective control strategies across all considered realizations of uncertain heat demand compared to deterministic optimization. For a fixed daily heat demand, the robust approach achieved a 5.9% reduction in emissions and a 1.4% increase in profit compared to the best deterministic scenario. These findings underscore the potential of robust optimization in addressing uncertainties and improving the operational efficiency of geothermal energy plants.