Digital twins and their value for energy industry
No need for real-life crash testing any more. Just break a mathematical model of a car in virtual reality! A digital twin is this very model to manipulate and obtain valuable information without affecting the object itself.

Dec 12, 2019
To this end, an object or facility model is built, combining information from the facility management system and data about its components and materials. We build a 3D model to forecast the object behavior, using algorithms developed during machine learning. For example, we can take into account plant running hours, weather, and its position against other campus facilities. In addition, the model includes technical maintenance and repair history to show malfunctions and how they were eliminated. The twin shows the object past and current states in detail to more accurately forecast its future and thus improve performance.
Digital twin is a promising technology for energy enterprises, and here are two our success stories to prove it.
Heat power plant digital twin
Energy companies are embracing digital twin technology, like one large heat power plant that engaged us to optimize its operations. The project was to reduce fuel costs and improve efficiency at both day-ahead and balancing wholesale energy markets. For this purpose, we decided to build a decision-making support system covering both production and financial functions. We developed a simulation model to rather accurately calculate all key technological parameters (pressure, temperature, fuel consumption, etc.) at each point of a technological process.

Our developers gradually checked the digital twin parameters against those of a real-life equipment and finally achieved the required accuracy (of max 2% deviation). After that, the customer started to use the twin for "what-if" analysis, i.e. to run different modes and evaluate their effect. The digital twin was then followed by an optimization model that allowed the customer to calculate the best-fit per-hour load for the entire plant with regard to day-ahead market operations — all with due consideration of weather and heat consumption forecasts, as well as regulatory restrictions. In addition, the model was configured to minimize losses caused by operations in the balancing market and reduce fuel consumption rate.
    Data center digital twin
    During the project, we made a digital twin to monitor computer room microclimate in real time and thus optimize the operation of air conditioning and ventilation units in the data center.

    Servers consume a lot of electric power and produce the equal amount of heat, removal of which also requires electric power. In terms of overall energy consumed by computing and air conditioning equipment (about 8 MW in this case), a large data center can be compared to a medium-size plant. We developed and installed wireless temperature transducers across the computer room. These transducers helped us simulate heat flows, detect hot and cold aisles, and build a dynamic heat map. The project payback period was less than a year.

    © NGN International 2019. All rights reserved.
    NGN Group