The primary purpose of this study is to develop a mathematical model for performance optimization of power generating unit of large-scale sewage treatment plants through the computational intelligence techniques cuckoo search algorithm (CS), dragonfly algorithm (DA), and grasshopper optimization algorithm (GOA). For this purpose, a state transition diagram is developed, and Chapman-Kolmogorov differential difference equations derived by the Markovian birth–death process. The nature of failure and repair rates are exponentially distributed, with flawless fixes. The numerical results for availability and parameters related with each optimization strategy are calculated for various population sizes and iterations. After 500 iterations and population size of 50, the system achieves its maximum availability of 0.9801022 by GOA. Finally, the detailed numerical results are presented. The findings of this study are intended to set the framework for future breakthroughs in sewage treatment plant design and operation.