A digital twin is being introduced at Azot to optimize production

2023: Implementing a Digital Twin to Optimize Production

On September 25, 2023, the MCD company announced that a digital twin was being introduced at KAO Azot to optimize production.

KAO "Azot"

As reported, the development of modern large-tonnage technological production is a difficult technical task. The return on investment and further business development depends on how accurately the company can identify the most restraining areas and choose measures to modernize them. The management of KAO Azot understands that the search for technological limitations that reduce production efficiency and the analysis of the consequences of replacing or reconstructing equipment cannot be based only on the expert knowledge of production personnel, production experiments or technical proposals of third parties. The company needs to have own objective source of information on technological processes "as is" and "as will be." At the same time, it must be accurate in order to assess the impact of any change on the operation of the system and minimize the risk of making an incorrect decision. This source is a detailed mathematical model of enterprise processes. It reproduces the transformations of all material flows in the workshop, shows restrictions and allows virtual tests of technical solutions.

In order to create such a model for the workshop for the production of ammonium nitrate, the specialists of KAO "Azot" turned to the experts of the company "Modeling and Digital Twins" (JSC "MCD").

Computer-Aided Engineering (CAE) is a well-known technology that allows you to efficiently develop products. The ability to perform comprehensive virtual tests prior to physical prototyping reduces development time and optimizes product quality.

Thanks to the development of mathematical modeling programs for September 2023, they are used much more widely, for example, to optimize the work of existing enterprises. The main feature of such tasks is that the model must accurately correspond to the operating production. If there is no match, its calculations cannot be trusted. This requirement creates certain difficulties for the developer - you need to take into account a large variety of processes, while accurate data on how they proceed in the equipment is often not enough. Therefore, it is impossible to create a high-quality model without the participation of the production itself, as well as the people who are responsible for it at the enterprise.

It is also important that the final solution should be alienated from the developer. That is, the model should, working in the contour of KAO Azot, help the specialists of the enterprise to quickly solve current production problems, maintaining relevance and efficiency for a long time.

Experts of MCD JSC needed to develop a model of the large-tonnage saltpeter production workshop, commissioned in the 80s of the last century and underwent many changes. The workshop operates at the maximum design capacity, exceeding the production plan. But the production agenda requires further systematic growth in production. At the same time, a large number of equipment, being outdated, can restrain productivity growth and limit production. Given the workshop's long history and its many changes, it is not easy to predict how equipment upgrades will affect production, whether the stability of process units will remain. It is also important how quickly the cost of upgrading will pay off, although it is not the only task of the digital twin. It has a long-term effect, helping to solve problems and choose effective rationalization proposals for staff.

Various adverse events can occur during the operation of the process equipment, which have a poor effect on production. However, problems appear irregularly, and there are many influencing factors. Therefore, it can be very difficult to understand and eliminate their causes. The problem can be solved by technological modeling. It makes it possible to avoid unnecessary risk and losses for production. told Andrey Krylov, Director of the Center for Digital Technologies of MCD JSC

Specialists of KAO "Azot" have done a difficult job in preparation for the creation of a digital model. They collected a large array of initial data regarding the operation of all elements of the technological complex for the production of ammonium nitrate. Among them are retrospective data on analytical control, initial data on design characteristics, technological regulations of the workshop and technological diagrams of the workshop.

{{quote 'author
= Andrey Krylov told' Initial technological data play an important role for mathematical modeling. At the same time, there is relatively little objective and accurate data at existing industries to perform analytical tasks. Over the years of its work, the specialists of MCD JSC have developed certain approaches to solve this problem, including reverse design, machine learning methods and much more. However it cannot be completely solved alone, so the active position of production in the search and collection of source data is a key factor in success.}}

The work of the experts of MCD JSC began with a technical examination - checking the actual configuration of the production for compliance with the collected data and documentation. Technological data on the operation of the saltpeter production workshop included several hundred key parameters of the integrated process. The finished digital model consisted of a total of more than one and a half thousand individual elements - heat exchangers, pipelines, valves, mass exchange dynamic equipment. In real operating conditions, the operating modes change, the equipment works in different ways. Over time, a gradual change in its characteristics occurs due to wear, corrosion, and external factors. It was impossible to guarantee the correctness and consistency of the entire data set without site work and technical inspection.

The trip of the specialists of MCD JSC to the facility coincided with a technological shutdown - one of the workshop's installations was stopped for planned repairs. They managed not only to check the dimensions of the pipelines for compliance with the technical documentation, inspect the operating units and units, remove the temperature and other technological indicators on the spot, but also to study the internal structure of the stopped units - the pellet tower, scrubbers. This provided useful information for the correct construction of a mathematical model.

Nevertheless, in the process of building and verifying the model, experts from MCD JSC found a number of discrepancies in technological data. They inevitably arise in most digital twin projects: certain inconsistencies in technological data may not interfere with the work of the workshop, but interfere with accurate modeling of processes. Therefore, the greatest time in the project was taken by work with data, which included the identification of obvious discrepancies and their settlement, taking into account production experience, compliance with the material balance, the laws of physics and thermodynamics. At this stage, the specialists of KAO Azot proposed a number of changes that made it possible to guarantee compliance with material and thermal balances without prejudice to the main goal. Among them - less detail, an extension of the tolerance range for those areas whose data were ambiguous, and the impact for the main production indicators insignificant. Thus, the experts of JSC "MCD" managed to ensure the general convergence of the material and heat balance, model results and readings of measuring instruments of real production.

After the digital model was ready and verified, the project entered the next phase of results analysis and development on the production task model.

First of all, it was necessary to determine using the model which parts of the technological scheme could limit the production of the product. These parts - the so-called "bottlenecks" - were laid down in the original design and did not limit the productivity of the workshop during normal work. But if you increase the supply of raw materials and optimize product production, they will not cope with the increased load: pipelines and valves will give optimization hydraulic resistance, heat exchangers will not provide the necessary heat transfer, fans and pumps will not be able to give the necessary flow supply. The load on the units differs in different technological modes, and the search for "bottlenecks" requires a series of calculations and compare their results.

When "bottlenecks" are defined, it is necessary to reconstruct the equipment in such a way as to reliably remove the restriction. If, for example, you simply increase the diameter of the pipeline or replace the heat exchanger, this will most likely not solve the problem, but will transfer the "bottleneck" to another element of the circuit. The model helped the experts of MCD JSC to select a set of necessary changes, removing restrictions in the entire range of operating modes of the units. Further, specialists needed to determine how to modernize production with minimal costs in order to achieve greater productivity while reducing the cost of production.

The technological model is not able to answer this question by itself, so the experts of MCD JSC have additionally developed an economic production model. It made it possible to calculate the payback period of changes and arrange all events in the order of its increase. Since KAO Azot is considering modernization projects with a payback period of no more than three years, the associated technological and economic model, as well as the experience of experts, made it possible to exclude ineffective or too expensive technical re-equipment, and draw up a short list.

Large-tonnage production of KAO Azot has been operating and developing for more than a decade. Sustainable development is a long-term process, so a one-time study of production tasks is not sufficient. That is why the main result of the joint work of KAO Azot and JSC MCD is the introduction of a digital twin in production enterprise processes.

The management of KAO Azot actively involves production personnel in the process of generating ideas, stimulates an increase in technological competencies. Proposals from the personnel of ammonium nitrate production should be promptly analyzed on the digital double of the workshop. Thanks to the digital twin, experts of KAO Azot can analyze dozens of proposals per year, checking the effects, finding the optimal solutions, identifying potential risks.

{{quote 'author
= Nikolay Denisov, Director of Business System Development at NAO Azot|Creation of a digital twin of one of the large-capacity units of Nitrogen is the result of teamwork of specialists of the process modeling group of the Directorate for the Development of the Business System "Nitrogen" and experts of JSC "MCD." For the largest chemical enterprise behind the Urals is a confident step forward to production achievements. We could not do without an effective search and qualitative elimination of losses - the key tasks of lean production. Digitalization allows you to solve them with minimal costs and risks. It is logical that a number of projects involving a cybernetic approach to the systematic analysis of the operation of aggregates and the progress of complex chemical processes were included in the Transformation of the Nitrogen business system. The next step is to optimize them. The strategic goal set by the company's management is to create an efficient, dynamically developing and safe production. Digital solutions can help us do this.}}