Stochastic modeling and performance optimization of geothermal power plants using nature-inspired algorithms

The main purpose of this study is to develop an efficient stochastic model for the performance evaluation of geothermal power plants. The geothermal power plant is a very complex structure involving five subsystems in a series configuration, having component-wise redundancy. Here, a transition state diagram is developed using a Markovian approach, and Chapman-Kolmogorov differential equations are derived. All the random variables associated with geothermal power plants are considered to be exponentially distributed and statistically independent. Later, various nature-inspired algorithms at different population sizes and iterations are used to optimize the availability of geothermal power plants. It is observed from the findings that the increase in the failure rate of the subsystem results in a decrease in the overall availability of the geothermal power plant. The importance of computational techniques in optimizing geothermal power plant availability is also highlighted in this study. The results highlighted that the Artificial Bee Colony Algorithm (ABC) outperforms all the other algorithms for optimizing the performance of geothermal power plants with 0.9981613 as the highest optimal predicted availability. Additionally, statistical estimation for all the proposed algorithms was also calculated in which ABC attains the highest values of mean, median, and mode.