Integrating Artificial Intelligence and Optimization Models for Predictive Compressor Washing in Industrial Gas Turbines: A Systematic Review and Future Framework

Compressor washing is a proven method for mitigating fouling and maintaining the performance of industrial gas turbines. However, traditional wash scheduling based on operator experience often results in suboptimal outcomes and increased operational costs. Recent developments in artificial intelligence (AI), machine learning (ML), and optimization models now enable condition-based and predictive maintenance strategies. This study proposes the first comprehensive framework that integrates traditional compressor washing practices with modern artificial intelligence (AI) techniques and optimization models, enabling predictive, cost-effective, and condition-based maintenance strategies. The review’s main contribution lies in proposing a unified framework that connects fouling diagnostics, AI-based prediction, and cost-optimized washing scheduling, thereby identifying research gaps and providing a roadmap for intelligent compressor maintenance in industrial gas turbines. This review traces the evolution of compressor washing from fixed-interval approaches to data-driven, optimization-based frameworks. A systematic review of academic and industrial publications was conducted to evaluate both conventional practices and emerging intelligent techniques. Research employing continuous-time Markov chains, Mixed Integer Linear Programming (MILP), and AI-driven diagnostics demonstrated significant economic and operational gains, including cost reductions of up to 66%, annual profit improvements exceeding $1 million, and measurable efficiency recoveries. Technologies such as digital twins, neural networks, and deep learning models further demonstrated to have provided enhanced real-time monitoring and predictive scheduling. Overall, this review underscores the transformative potential of AI and optimization in transitioning compressor washing toward predictive, sustainable, and cost-effective gas turbine operations.