The modeling technologies applied at RUSAL production facilities are based on creating digital twins of process equipment. They help to define optimal parameters of production processes.
For the first time in the world, full-featured mathematical models of parallel Bayer sintering (Ural Aluminium Smelter and Bogoslovsk Aluminium Smelter) and sintering (Achinsk Alumina Refinery) were developed and implemented at RUSAL alumina refineries. For this purpose, Company engineers used SysCAD software.
The digital models completely replicate the logic of plant management, are integrated with CHP models, and contain detailed heat-and-mass balances, the chemical reactions and compositions of all flows. Each of the models includes the digital twins of more than 2000 units of equipment, more than 10,000 parameters of production processes. Mathematical modeling and forecasting of production processes in SysCAD allows defining the optimal production parameters for each area, site and unit of equipment, identifying the causes of process disruptions, evaluating the efficiency of capital expenditures, analysing the performance for certain periods, and carrying out current and long-term planning.
At the same time, RUSAL alumina refineries actively apply computational fluid dynamics methods to model all types of production equipment. This approach allows choosing design solutions that contribute to production process intensification and reduce equipment renovation and upgrading costs.
In the aluminium business, reduction process modeling is provided by the Virtual Pot dynamic digital model developed by the specialists at RUSAL's ETC and scientists at the Siberian Federal University using Big Data algorithms. As a result, almost all the reduction cells have now got a 'digital twin' in the Company. The Virtual Pot model is constantly being improved, and the current 3.0 version has got an optimisation and predictive capability based on virtual sensors (Soft Sensors).