Study On Gas Path Diagnostics For Marine Gas Turbine

Marine gas turbine engine runs under marine environment,suffering poor working conditions with high temperature,high pressure and high mechanical and thermal stress as well as salt fog in the air,and thus the performance of its gas-path components degrades gradually,easily leading to various serious faults.So the current health status of the engine gas-path components is essential information for users and operators.However the performance degradation is not easily foreseeable when the level of degradation is small.Based on the review of related literature,study on gas path diagnostics for marine gas turbine was carried out.The technique of gas-path analysis helps users to make a more accurate and detailed maintenance schedule to minimize shut down time,save maintenance costs,increase reliability and availability.From the perspective of nonlinear gas-path diagnostics,the paper mainly focuses on the study of key problems lying in the existing thermodynamic model based gas-path diagnostic methods,including accurate gas turbine thermodynamic modeling for the purpose of engine performance analysis and gas-path diagnostics,diagnostic precision highly affected by the deviated ambient(e.g.,pressure,temperature and relative humidity)and operating control conditions,diagnostic reliability highly affected by the number of components within engine and the measurement noise level,diagnostic accuracy highly depending on the reliability of gas-path sensors.Aiming at these problems,related new gas-path diagnostic approaches were proposed for marine three-shaft gas turbine,and the specific contents are as follows:Firstly,set up gas turbine thermodynamic performance model with high accuracy.The working fluid thermophysics of air and gas were established based on NIST website database;nonlinear thermodynamic modeling of each component suitable for characteristic calculation under different working fluid components was implemented,which provides basis for development of simple and effective component characteristic map modification method and gas-path diagnostic method;a gas turbine component characteristic map adaptation method based on particle swarm optimization algorithm has been developed in order to improve the thermodynamic calculating accuracy,which is suitable for the modification of component characteristic maps from the same fleet or other types,and by which the modified component characteristic maps match with the actual ones very well.Secondly,solve the problem of diagnostic precision highly affected by the deviated ambient(e.g.,pressure,temperature and relative humidity)and operating control conditions.The health parameters for gas-path components were redefined using component similar corrected parameters in order to eliminate the side effect on diagnostic result produced by the deviated ambient(e.g.,pressure,temperature and relative humidity),and were regarded as independent parameters directly with gas-path measurements as objective parameters,and thus an improved nonlinear gas-path diagnostic approach has been proposed as main framework for the study of marine gas turbine gas-path diagnostics.Next,solve the problem of diagnostic reliability highly affected by the number of components within engine and the measurement noise level.In order to more effectively isolate the degraded components and accurately quantify the fault degradation rate,a nonlinear steady-state model based gas turbine health status estimation approach with improved particle swarm optimization algorithm has been proposed further,from the perspective of global optimization to improve the diagnostic results.At the meantime,in order to improve the diagnostic accuracy as well as real-time performance,in the view of the pattern recognition techniques,a hybrid gas-path diagnostic approach integrating gray relation theory into nonlinear gas path analysis method has been proposed from the perspective of reduction of the dimension of fault coefficient matrix to improve the diagnostic results.The hybrid gas-path diagnostic approach includes two steps.Firstly the faulty components are recognized and isolated by gray relation theory,which deeply reduces the dimension of fault coefficient matrix(i.e.,the subset of actual health parameters to be diagnosed is determined),and secondly the magnitudes of detected component degradations are quantified by the improved nonlinear gas-path diagnostic method.At last,solve the problem of diagnostic accuracy highly depending on the reliability of gas-path sensors.Usually,it is essential to use correct measurement information to obtain correct fault signature for producing accurate gas-path diagnostic results.However gas-path components as well as sensors may degrade or even fail during gas turbine operations which may produce significant measurement biases,and misleading diagnostic results may be obtained.Aiming at this problem,a methodology to improve the robustness of gas-path fault diagnosis against sensor faults was proposed for the typical nonlinear gas-path diagnostic method.The proposed method involves two steps:Firstly carry out process data reconciliation based on Gaussian correction principle to detect degraded or fault sensors for all the gas-path measurements from one steady-state operating condition.And secondly detect,isolate and quantify the degradation of major gas-path components(compressors,combustor and turbines)using the nonlinear gas-path diagnostic method based on gas-path measurements from multiple operating points.