Predicting Gas Turbine Power Output in Single Shaft Combined Cycle Power Plants: A Physics-Based and Data Analytics Approach

The measurement of gas turbine power output is essential for original equipment manufacturers (OEMs) and plant operators to evaluate the performance of their gas turbines. Accurate knowledge of the power output allows for better gas turbine control based on customer requirements. By measuring the GT power, OEMs can predict critical internal performance parameters such as turbine inlet temperature, compressor inlet flow, compressor efficiency, and turbine efficiency. Furthermore, accurate prediction of steam turbine efficiency can also be achieved using a mass and energy balance approach. In a multi-shaft combined cycle power plant, gas and steam turbine power are measured separately, enabling analysis of the heat and mass balance approach within individual turbine boundaries. However, in a single-shaft combined cycle power plant, only the total cycle power is measured, making it difficult to predict gas turbine and steam turbine power output separately. Accurate gas turbine power output prediction is critical to prevent improper gas turbine operation, such as over-firing or under-firing. Over-firing the gas turbine can damage the hot gas path life of the turbine while under-firing can result in a loss of gas turbine power. Additionally, accurate power output prediction enables OEMs to provide precise performance guarantees, reducing the risk of paying liquidity damages. To validate the single-shaft approach for predicting gas turbine power output, data from a multi-shaft combined cycle power plant was used, and accurate gas turbine power output was predicted and compared with the power observed using multiple-shaft gas turbine power. The first approach of the paper is solely founded on the principles of the first law of thermodynamics. The objective is to validate this first law approach using only a limited data set from a multishift combined cycle power plant, such as compressor inlet temperature, fuel flow, fuel lower heating value, combined cycle power output, and gas turbine power output. The other methods are Hybrid modeling approach and Data driven approach. In the hybrid method a model has been obtained using mass and energy balance equation and the real gas turbine data is used to obtain the values of unknown parameters in the hybrid model. In the Data driven approach, real gas turbine’s data is used to obtain a black box model. Hybrid models are generally regarded as superior to data-driven models in terms of accuracy because they combine the use of physics-based principles and data. Data-driven approaches, on the other hand, offer a simpler way to model plants or processes as they do not rely on prior knowledge of the system and can be developed solely using process data.