MODEL BASED PERFORMANCE CORRECTION METHODOLOGY - A CASE STUDY

摘要

Correction curves are of great importance in the performance evaluation of heavy duty gas turbines (HDGT). They provide the means by which to translate performance test results from test conditions to the rated conditions. The correction factors are usually calculated using the original equipment manufacturer (OEM) gas turbine thermal model (a.k.a. cycle deck), varying one parameter at a time throughout a given range of interest. For some parameters bi-variate effects are considered when the associated secondary performance effect of another variable is significant. Although this traditional approach has been widely accepted by the industry, has offered a simple and transparent means of correcting test results, and has provided a reasonably accurate correction methodology for gas turbines with conventional control systems, it neglects the associated interdependence of each correction parameter from the remaining parameters. Also, its inherently static nature is not well suited for today's modern gas turbine control systems employing integral gas turbine aero-thermal models in the control system that continuously adapt the turbine's operating parameters to the "as running" aero-thermal component performance characteristics. Accordingly, the most accurate means by which to correct the measured performance from test conditions to the guarantee conditions is by use of Model-Based Performance Corrections, in agreement with the current PTC-22 and ISO 2314, although not commonly used or accepted within the industry. The implementation of Model-based Corrections is presented for the Case Study of a GE 9FA gas turbine upgrade project, with an advanced model-based control system that accommodated a multitude of operating boundaries. Unique plant operating restrictions, coupled with its focus on partial load heat rate, presented a perfect scenario to employ Model-Based Performance Corrections.