Experience in Use of Remote Access and Predictive Analytics for Power Equipment’s Condition

• Published in: Thermal engineering 2018, Vol.65 (4), p.189-199
• Main Authors: Naumov, S. A., Krymskii, A. V., Lipatov, M. A., Skrabatun, D. N.
• Format: Article
• Language: English
• Subjects: Analytics; Approximation; Automation and Heat Control in Energy; Change detection; Condition monitoring; Empirical analysis; Engineering; Engineering Thermodynamics; Gas turbine engines; Heat and Mass Transfer; Information technology; Inspection; Internet; Maintenance; Mathematical models; Mathematical problems; Predictions; Predictive analytics; Remote monitoring; Statistical analysis; Statistical methods; Statistical models; Thermal power plants
• ISSN: 0040-6015; 1555-6301
• DOI: 10.1134/S0040601518040055

Digital technologies, software of predictive analytics, and advanced equipment will make it possible to improve economy, reliability, and safety of electricity generation. The industrial Internet begins with the introduction of systems based on mutual penetration of information technologies and automation devices of manufacturing equipment, such as the systems of remote monitoring and diagnostics. One of the inspection methods of the equipment’s condition is its continuous monitoring, which is a necessary condition for the transition to a service system on the operating condition. Using traditional modeling methods, it is possible to obtain only approximate data about the behavior of industrial systems and objects even in the cases when all factors influencing their work and operating condition are known, owing to the necessity to solve complex mathematical problems to carry out this modeling. For this reason, to monitor the operating condition of industrial systems, the statistical modeling of such systems based on empirical regulations defined by the samples of values of technological parameters recorded in the object operation period, which is considered by reference, found application in recent decades. The statistical methods of monitoring makes it possible to detect the changes in the operating condition of the system at early stages as well as to reveal the most important factors influencing them. The work presents a review of Russian systems of predictive analytics and mathematical methods on which they are based and also the PRANA system of prediction and remote monitoring that is implemented at the gas-turbine plant of V 94.2 Siemens type installed in the Perm TPP-9 (thermal power plant), the Vladimir TPP-2, the Izhevsk TPP-1, and the Kirov TPP-3, which are branches of PAO T Plyus. The efficiency of PRANA to detect the negative change of operating conditions before actual fault events was shown, which makes it possible to determine the residual life of a product and its components, schedule the optimal terms, the duration of equipment stop and preparation for its repair, and evaluate the quality of fulfilled repairs. The condition of the industrial Internet in Russian power engineering and the problems delaying its development are considered.