Project start: 1.10.2025.

Based on the measured operating parameters from five different energy power plants, an analysis and optimization of each plant will be performed. The plants that will be considered are two thermal power plants (TPP Plomin 2 and TPP Rijeka), one combined-cycle power plant (CCPP Jertovec) and two marine steam turbine propulsion plants (older and new generation). Analysis and optimization of each plant will be performed in several different operating regimes. The necessary operating parameters for the new generation marine steam turbine propulsion plant are currently being collected, for the Plomin 2 thermal power plant, in addition to the existing ones, new operating parameters are planned to be collected, and for other plants, the necessary data in different operating regimes have already been collected at the moment. During optimization, conventional numerical methods will be used (examining the possibility of renewable energy sources implementation, examining the possibility of a waste heat recovery system application, examining the possibility of energy storage installation, defining optimal operating regimes of crucial components, etc.), as well as machine learning methods (multilayer perceptron, genetic programming, gradient boosted decision trees, etc.), and combined methods (combination of conventional numerical methods and machine learning methods). Special attention will be placed to combined optimization methods due to their complexity, but also they enable the best optimization solutions. The final goal of each plant optimization in all observed operating regimes will be to reduce fuel consumption and harmful gas emissions while maintaining the existing power output. Finally, for each observed energy plant, the possibilities of existing fuel replacement with alternative variants will be examined in order to further reduce harmful gas emissions.