T Plus Group is one of the largest Russian private companies in the field of electricity and heat supply.
T Plus Group owns 6% of the installed capacity of power plants in Russia.
T Plus Group is the leader in the country's heat supply market with a share of 8%.
The Company provides stable and uninterrupted heat and electricity supply in 16 regions of Russia.
The T Plus Group of Companies unites 53 power plants, including 51 thermal power plants and 2 hydroelectric power plants.
PJSC T Plus PJSC T Plus is the backbone of the Group’s Generation assets. It produces, transmits and sells heat and electricity.
“Prognostics” is a system for predicting equipment defects, based on the analysis of historical big data using modeling methods, statistical analysis, machine learning and a number of other technologies.
Russian Federation, regions of the company's presence
Electricity and heat supply
G
July 2022 – present
-
Minimizing forced outages and downtime is critical to the operation of an energy utility. A timely “signal” is needed to indicate developing defects and allow one to avoid serious consequences.
Project goal: to achieve a reduction in accident rates at generating equipment facilities through automated monitoring and forecasting of technical condition using import substitution of foreign software.
Project objectives:
Analysis of system performance indicators in real time by equipment units.
Identification of developing defects.
Creation of hybrid models for diagnostic objects.
Introduction of machine learning and formation of a defect classifier for the units and equipment used.
Consumers and local communities of the Company, the heat and power generation industry of the Russian Federation.
Prognostics helps to identify a possible defect using modeling methods, statistical and parametric analysis, as well as machine learning, neural networks and a number of other modern models and technologies.
When the data is collected and analyzed, two dynamic models are built on its basis: physical and mathematical, which make it possible to predict how the station equipment will behave in a range of up to 7 days. The models are constantly compared with data that comes from the station equipment in real time. For this purpose, thermal power plants have sensors installed on the main equipment (boilers, turbines) and auxiliary equipment (for example, pumps). Sensors analyze the condition and operation of each element of the station. The collected data goes directly to the Prognostics system, and it, in turn, signals operational personnel about possible deviations from optimal parameters.
With the help of Prognostics, a reference mathematical model of equipment is built, taking into account its technological features in various operating modes: load, unloading, start and stop (that is, how ideally it should work under any external and internal factors). When a deviation from this model occurs, the system notifies that a failure has occurred.
As part of Prognostics, a defect classifier has been developed, which is regularly updated with new data. The information comes from the system itself through integrated machine learning. Additionally, the relationship between the parameters is analyzed by specialists from the expert data processing center, which makes it possible to develop competencies in industry solutions and provide diagnostic support.
The Prognostics module has been submitted and received a certificate of state registration of the computer program.
As a result of software integration, the following effects were identified:
Early detection of developing defects;
Reducing accidents at facilities;
Reducing the number and timing of eliminating unplanned shutdowns.
Thanks to Prognostics, three incidents at the Akademicheskaya TPP have already been prevented, which could have put the station out of action for a total of 24 hours. As a result, the company would not have supplied 5.5 million kWh of electricity to the network and would not have supplied 3,36
National goal "Digital Transformation"
National project "Digital economy"
National project "Labour Productivity"
