Digital transformation of the power system of Kemerovo "Nitrogen": from manual accounting to predictive analytics

2025: Results of Digital Grid Transformation

On October 13, 2025, Mavismart shared with TAdviser information on what the digitalization of the energy economy means and how it is integrated into the general digital transformation of an industrial enterprise using the example of one of the largest chemical plants in Russia - KAO Azot, Kemerovo.

The power supply system of the chemical plant consists of 110 kV and 220 kV built by the CHPP plant, external supply lines of PJSC Rosseti and an extensive distribution network of the enterprise (180 PS/KTP) on an area of ​ ​ 300 hectares. Consumption is now just over 100 mW. During the construction of the plant, a reserve of 200-220 mW was laid.

KAO Azot has an energy saving program, introduced the foundations of lean production and building its own business system aimed at continuous development and improvement.

In 2018, KAO Azot decided to comprehensively modernize the power supply control system, and the company simultaneously switched from obsolete methods to a modern centralized electrical management system. Large-scale, costly, but as efficient as possible.

A single platform as the core of digital power

The essence of a centralized system is to collect visualization and all data about to power in a single information space, provide this data to users and transfer it to adjacent enterprise systems. The technological core of the system is a platform SEDMAX that, in addition to direct purpose, acts as an power enterprise data bus.

The project started in March 2018 and lasted several years in two main functional areas:

• dispatching (telemechanics) to monitor and analyze the operation of the power system
• energy accounting for monitoring and analysis of consumption.

Accounting and Analytics

Initially, the task of the accounting system (ASTUE) was simple: correctly distribute energy resources between structural units. Accounting was organized in accordance with the approved methodology (description of the belonging of accounting points to a particular workshop).

As of October 2025, accounting in Nitrogen is carried out in line with two tasks:

• operational control of shop load
• analysis of statistics for the period

Online control of current consumption makes it possible to quickly intervene in the event of an obvious overspend.

The overspend is not estimated based on the results of, for example, a ten-day period, but in real time - we raise the data history to understand exactly when the skew occurred, - explained Egorov Anatoly Vasilievich, Head of Electrical Department of UGE KAO "Azot"

𝓲

Фото: Mavismart

Statistical data (reports for the month) allow you to find implicit overspending, bottlenecks, analyze the operation of each large engine in time. Based on this data, every month the energy management group develops targeted measures to change technological processes and improve energy efficiency. The so-called data-driven planning, based on statistics accumulated since 2018, has become an important tool for energy efficiency: PTW analyzes actual consumption and draws up an annual consumption plan, which technologists then adjust each quarter taking into account seasonality, repairs, gas shutdowns and other factors. This, in turn, allows you to compare actual unit rates with planned ones and analyze any deviation in a mode close to real time.

Another important area is the control of the declared capacity during the maximum hours, the excess of which leads to a significant increase in the cost of electricity (capacity). This indicator is controlled in the commercial electricity accounting system (AIIS KUE), implemented on the basis of the Energosphere PC, which records the volumes "at the entrance" from the 220 kV PS.

dispatching: visualization, reaction, warning

The impetus for the automation of dispatch control was not so much the high level of accidents in the power supply system of KAO Azot (as in many large enterprises), but the lack of monitoring and control tools in principle. As in the Soviet joke, I had to "fly on a pack of Belomor." Therefore, the initiator and the main driver of this project were the power engineers of the plant themselves, and not the top management responsible for the development of the "digitalization of the plant" in general, shared in Mavismart.

The scale of the Nitrogen power utilities demanded to bring mnemonic diagrams not only to the dispatchers' AWS, but also to organize a multimedia video wall 4.1 m x 2.4 m in the control room, consisting of 24 LCD panels 55.

𝓲

Фото: Mavismart

With its implementation, the dispatchers received the most visible single interface for managing their power system. It turns a huge amount of raw data (over 30 thousand tags) from hundreds of devices into a simple and understandable form, allowing the dispatcher to quickly make the right decisions.

For operational operation, the system allows, for example, when one of the three lines fails, to quickly assess whether it is possible to leave the load on the two remaining cables for the period of replacement.

A striking example of the effectiveness of the new system at the business level is the emergency situation in 2021, when on one GPP when replacing a lightning protection cable, it blocked the reserve line. 4 transformers were disconnected, many workshops were de-energized. Another example is the descent of snow on the trump card, which under pressure fell on the line from the GLP and turned off the 110 kV transformer and followed by 10 PS.

The system cannot prevent an accident caused by such reasons, but it allows you to determine in minutes instead of hours the place "where to run" for the brigade and correctly localize it. The faster the power engineers can restore the normal scheme, the faster the technologists can start the workshop.

In this case, it may take 5-6 hours to start the workshop (the time when the workshop enters the mode). And, if an emergency shutdown is a loss, then the time of entering the mode is underutilized products. Naturally, energy resources are additionally consumed during this period of time. For example, burning gas to a candle per day costs 7-8 million rubles. Therefore, one simple plant such as Azot can cost from 1 million to 40-50 million.

Intersystem Integration and Data Value

End-to-end digitalization is an important requirement of the Nitrogen digitalization process: data from the commercial accounting system (AIIS KUE) is transmitted to a single SEDMAX system, and SEDMAX transmits both the accounting data and "own" data to the MES-system of the ASOU (automated operational control system). In ACEI, this data is used to:

• output to MES mnemonic diagrams
• distribution of electricity consumption by workshops;
• real-time control of efficiency parameters;
• report generation.

𝓲

Фото: Mavismart

The only "manual" intervention is associated with the need to compare accounting points with workshops. This problem is characteristic of many enterprises implementing digital solutions, and is a consequence of the dynamism of this kind of systems, in which something changes regularly (new industries appear, new accounting points are added, old ones are removed, etc.).

How Energy Data Affects Technological Processes

ASEI has introduced "speedometers," which in real time compare three indicators: current consumption, consumption for the last year and the planned value. If in similar conditions there was less consumption a year ago, technologists, together with the energy management group, are looking for the root causes and developing corrective actions.

On electronic boards in workshops, the same indicators are visible to operators - this (and, of course, bonuses) motivates them to conduct a more efficient production regime. For example, the chemical reaction can proceed at a temperature of 15 ° C to 30 ° C. At 15 ° C, power consumption is one, at 30 ° C - another, but product quality is the same. The operator can work as efficiently as possible, but produce products, but can work as efficiently as possible, choosing the optimal mode and releasing the same products.

At the end of the month, the ASOU also prepares reports on the distribution of electricity consumption by workshops, which are used by economists to calculate the actual energy consumption of products.

"Compressors" subsystem based on ML

Compressors are the most energy-intensive consumer at the plant (up to 30% of total consumption). To optimize them, Nitrogen specialists developed an intelligent compressor balancing subsystem using machine learning using the open Grafana platform.

The system analyzes historical data on the consumption and capacity of compressors, takes into account temperature, humidity, air pressure and other parameters. Based on this data, she predicts which compressor will be most efficient in the next hour or month. Calculations are immediately displayed in monetary terms.

𝓲

Фото: Mavismart

For example, if a technologist loads 4 compressors in the amount of 1 million rubles/hour, the system can offer another set of compressors for 800 thousand rubles/hour.

As of October 2025, the subsystem operates in test mode and is advisory in nature.

Energy Management System

The company has an energy saving program with KPI to reduce consumption by 3% per year. In the "Energy management" system, measures, justifications of deviations are recorded, analytics is built in the form of dependency graphs, all documentation is accumulated. While the data is entered manually by the workshops, in the future it is planned to automate this process.

Conclusion

The created system of end-to-end automation of the chemical plant allows Nitrogen not only to reduce energy consumption, increase the reliability of energy supply, but also to create conditions for the further development and optimization of many related processes.

The digital transformation of Nitrogen power is one example when the modern dispatching system becomes part of the industrial infrastructure, influencing technological processes and increasing their efficiency and reliability.

SEDMAX allows you to organize continuous monitoring of power consumption parameters for individual workshops, units and equipment. This makes it possible to quickly detect deviations from the regulatory operating modes associated with changes in technological parameters (pressure, temperature, capacity, etc.), as well as to identify the potential for energy saving, - noted Shkuts Nikita Vadimovich - head of the energy management group of the Technical Directorate of KAO "Azot"