Expert System for Diagnostics and Status Monitoring of NPP Water Chemistry Condition

Abstract

Water chemistry condition (WCC) has been the subject of constant study and improvement up to the present day. It is connected with the presence of a direct relationship between the violation of water chemistry regulation on the one hand and components reliability of the circuit’s equipment and cost-effectiveness of their operation on the other. It dictates the necessity to apply different optimization methods in the field of monitoring and use of information — analytical and diagnostic systems to assess WCC quality, control and support. By now NPP experts have broad experience in revealing and removing the causes of WCC disturbances. However this knowledge is often of an intuitive, non-classified nature, scattered among various working documents, which makes their transfer difficult. Based on what has been mentioned above, special attention is currently being paid to the problem of creating expert diagnostic systems for supporting the optimum WCC. The existing developments in this field (DIWA, Smart chem Works, the water quality control system at the Onagava NPP etc. [1,3,4,5] are based on wide use of experts’ knowledge. Such expert diagnostic systems for supporting WCC refer to the new generation of intellectual control methods, which allow the incorporation of the latest achievements both in the field of water chemistry simulation and in the field of artificial intelligence and computer technologies. LI “VNIPIET” employees have, for several years, been developing an expert diagnostic system for supporting WCC and status monitoring of RBMK - reactor NPPs [2]. This system has not only conveniently organized the traditional functions of information acquisition and storage, a complete presentation of information in the form of tables, graphs of a dynamical changes of parameters and formation regular reports, diagnostic functions and issuing recommendations on WCC correction, but it also allows the assessment of confidence in the diagnosis made, relying on a wide range of numerical estimates, which were calculated by the use of expert data, and to make a credible prediction of an existing situation development. The integrated use of analytical methods and artificial intelligence methods is one of the system’s advantages. This combination allows the successful implementation of one of the main purposes of the system: the early detection of deviations from specified process conditions and the taking into account of even minor changes in parameters to provide an advanced WCC control and to prevent non-regular situations.