There is long-term interest in evaluating the life-cycle cost of various gas turbine air filters. The work described in this paper is designed to help form the basis for a life-cycle cost evaluation.

This article reviews the basics of inlet air filtration and offers a guide for choosing the preventive strategy to optimize the performance and health of your engine.

This paper provides an integrated, quantitative, and transparent approach to life cycle cost analysis for gas turbine inlet filtration. This work also serves as a technical summary of the underlying physics models that lead to the development of EPRI’s Air Filter Life Life-Cycle Optimizer (AFLCO) software.

This extensive course provides an overview of common analytics types and machine learning, a standard process for model creation, and examples of use cases.

When applied properly, AI has a multitude of powerful uses. This presentation introduces application of AI to gas turbine industry and provides a summary of machine learning archetypes.

This paper describes how variations in fuel composition influence gas turbine emissions, operability, and operational range (turndown), and further explains resulting fuel composition sensitivities as well as approaches for identifying and mitigating operational risk.

This article focuses on operability issues that Turbine Logic routinely observes when performing root cause analyses. They pose risks to gas turbines and are closely tied to operating conditions and fuel composition.

Demonstrates a physics-based algorithm to detect DC faults at utility-scale PV plants, offering a scalable solution for improving operational efficiency. It uses real-time plant data and has been validated at numerous sites with varying weather.

Benchmarks a physics-based anomaly detection model against industry-standard tools, showing superior performance in detecting subtle PV faults. The model identifies nearly twice as many faults as traditional methods while maintaining low false-positive rates.

Presents a technique to estimate PV plant site architecture using operational data, speeding up configuration and improving fault detection. This approach minimizes reliance on site drawings and enhances model accuracy for real-time anomaly detection.

Describes a method for detecting hardware faults in PV plants by analyzing amperage data. The approach improves maintenance cycles and helps operators identify faulty combiner boxes more efficiently, boosting overall plant performance.

Details a software-based method for identifying subtle PV underperformance issues using existing plant instrumentation. Proven across multiple sites, the method enables timely detection of string outages and tracker faults, validated through aerial infrared scans.

Explains AI/ML methods to automate fault detection and diagnosis in large-scale PV plants, with a focus on real-time data analysis, removing seasonal and soiling effects, and improving fault identification accuracy.