Relevant Points for Thermodynamic Diagnosis Application Using the Reconciliation Method

After some hours of operation, every power plant’s equipment begins to show degraded performance. The amount of equipment degradation and the gain that can be achieved in plant heat rate and power generation by the replacement or repairing of each component is important information for proper plant maintenance management. This paper aims to show how diagnosis and prognosis of a power plant can be performed based on thermodynamic data measured by plant instrumentation on site. Some important points such as how to increase the accuracy of the prognostic by developing the complete Taylor series and how it affects the thermodynamic state are also explained. Practical considerations for implementing a thermodynamic diagnosis/prognosis system in a real plant are also discussed. The system receives thermodynamic data from the plant information system (PI) and after a filtering step, the performance factors (PFs) of each component are calculated based on the component’s performance curves. Thus, the current state can be modeled using these PFs. By replacing the calculated performance factors, component by component, with factors that represents no degradation, future cycle performance can be estimated. This work is a partial result of an ongoing research program that implements a thermodynamic diagnosis/prognosis system on a 130 MW combined cycle co-generation power plant.