CCGT performance simulation and diagnostics for operations optimisation and risk management

This thesis presents a techno-economic performance simulation and diagnostics computational system for the operations optimisation and risk management of a CCGT power station. The project objective was to provide a technological solution to a business problem originated at the Manx Electricity Authority (MEA). The CCGT performance simulation program was created from the integration of existing and new performance simulation codes of the main components of a CCGT power station using Visual Basic for Applications (VBA) in Excel ®. The specifications of the real gas turbine (GT) engines at MEA demanded the modification of Turbomatch, a GT performance simulation code developed at Cranfield University. The new capabilities were successfully validated against previous work in the public domain. In the case of the steam cycle, the model for a double pressure once-through steam generator (OTSG) was produced. A novel approach using theoretical thermohydraulic models for heat exchangers and empiric correlations delivered positive results. Steamomatch, another code developed at the university, was used for the steam turbine performance simulation. An economic module based on the practitioners’ definition for spark spread was developed. The economic module makes use of the technical results, which are permanently accessible through the user interface of the system. The assessment of an existing gas turbine engine performance diagnostics system, Pythia, was made. The study tested the capabilities of the program under different ambient and operating conditions, signal noise levels and sensor faults. A set of guidelines aimed to increase the success rate of the diagnostic under the data and sensor restricted scenario presented by at MEA was generated. Once the development phase was concluded, technical and economic studies on the particular generation schedule for a cold day of winter 2007 were conducted. Variable ambient and operating conditions for each of the 48 time block forming the schedule were considered. The results showed error values below the 2% band for key technical parameters such as fuel flow, thermal efficiency and power output. On the economic side, the study quantified the loss making operation strategy of the plant during the offpeak market period of the day. But it also demonstrated the profit made during the peak hours lead to an overall positive cash flow for the day. A number of optimisation strategies to increase the profitability of the plant were proposed highlighting the economic benefit of them. These scenarios were based on the technical performance simulation of the plant under these specific conditions, increasing the reliability of the study. Finally, a number of risk management strategies aimed to protect the operations of a power generator from the main technical and economic risk variables were outlined. It was concluded that the use of techno-economic advanced tools such as eCCGT and Pythia can positively affect the way an operator manages a power generation asset through the implementation of virtually proven optimisation and risk management strategies.