Methodology to assess impact of operational parameters on operation of turbine feed pump

Background. Turbine feed pumps (TFP) are one of the main consumers of thermal energy for in-house needs of thermal power plants and nuclear power plants. Currently, much attention is paid to improve the efficiency of operation of turbine feed pumps. Thus, it is necessary to assess the impact of operational parameters. The problem is that it is necessary to process a large array of data. The solution to the problem is the development of a methodology that allows quantitative and qualitative assessment of operational parameters. Materials and methods. As initial data, the authors have used the data of 20 thermodynamic and hydraulic parameters recorded during the industrial operation of the TFP equipment and the equations of thermal balances to find the key performance indicators of the TFP. A complex of multidimensional statistical methods is used for data processing, i.e cluster and factor analysis in the Statistica software package. Results. A methodology to assess the impact of parameters on the operation efficiency of the TFP has been developed. The cluster analysis method is used to diagnose and classify the parameters under study. By the method of factor analysis, several parameters have been identified: feed water flow in the pipeline behind the TFP, differential pressure of water of the TFP, condensate flow in the pipeline behind the TFP condensate pumps, pressure drop of circulating water before and after the TFP condenser, condensate temperature of the drive turbine. The analysis of the results has showed that the identified parameters are the main components of the useful power of the turbine feed pump and the criteria for the efficiency of the TFP. Conclusions. The proposed methodology makes it possible to carry out a quantitative and qualitative assessment of operational parameters, solve the problem of processing large amounts of information, reduce the time and complexity of calculations. The results obtained are based on the developed methodology and can be used to design a statistical model using neural network technologies. This methodology can be chosen to evaluate the operation of technological systems and subsystems at thermal power plants and nuclear power plants.