Application of mathematical models in cycle chemistry monitoring systems for optimization of cycle chemistry at thermal power plants

One of the tasks of water chemistry control and monitoring at fossil power plants is prevention of water chemistry failure. This aim may be achieved by prediction of impurities concentration in different parts of the cycle and analysis of corrosion products behavior. To describe those processes it is necessary to understand the corrosion and scaling formation mechanisms. This understanding gives possibility to develop mathematical models that can provide real-time calculations based on regular cycle chemistry measurements. Application of those mathematical models is possible if cycle chemistry monitoring system (CCMS) is applied in thermal power plant. Integration of mathematical models in CCMS can gives opportunity to improve chemistry controlwith calculation of chemical parameters that can not be measured directly, to predict impurities behavior in water steam cycle, and also gives opportunity to analyze scaling and corrosion processes. To improve possibilities of mathematical models is possible by computational mathematics methods application. Artificial neural network (ANN) is the one of those methods. ANN is algorithms that can provide generalization of impute data and apply those results for current measurement data analyzes. This paper gives brief information about mathematical modeling with ANN application for single-phase flow accelerated corrosion (FAC) analyzing. For this propose mathematical model with two group ANN was developed. Those ANNs can identify possible place of single-phase ANN and possibility of it’s appear.