Artificial intelligence in the energy industry

The industrial site of the future is not only a consumer - but also a producer of electricity and heat. If it produces more than just its own needs, the companies involved can even provide grid-serving services in the future.

One of the specialties of the Bavarian construction company Max Bögl is hybrid towers made of steel and concrete for wind turbines. But where does the electricity for energy-intensive production actually come from? Especially for larger industrial sites, electricity costs are an important factor. Currently, the most cost-effective way to generate electricity is photovoltaics (PV) . Here, the cost of electricity for ground-mounted systems has fallen from around 30 cents/kWh to around four cents/kWh in recent years. Accordingly, PV systems have long been a sensible investment for businesses.

However, a photovoltaic system produces electricity only during the day and primarily at midday. Weather and seasonal fluctuations in generation make their use a challenge for industrial companies. Combining it with a battery storage system brings the necessary security of supply, but again drives up costs. The same applies to wind turbines .

Just in time thanks to AI

Machine learning can detect typical and atypical production, generation and consumption patterns and control automatable processes to make the best use of fluctuating electricity. With the goal of supplying the Sengenthal site with 100 percent self-generated electricity from renewable sources, Max Bögl has digitized its in-house grid and is in the process of transforming it into an intelligent energy cell. The energy mix of 9.6 MW of wind power, 2.5 MW of photovoltaics and a 1.5 MW battery storage system is intended to cover the annual consumption of 26 GWh - equivalent to the demand of a city with around 30,000 inhabitants.

The interaction is "intelligently" controlled on the basis of consumption and load profiles as well as forecast data. Intelligent adaptive systems are needed for the simultaneous balancing and optimization of a wide variety of generators, consumers, storage units and sector couplers such as power-to-gas and power-to-heat. Especially if operations management wants to optimize such energy cells, also with regard to volatile electricity and heat prices.

"Industrial grids can be the nucleus and stability anchors for our future decentralized energy supply," Josef Bayer, who heads the "Electrical Engineering and Wind" department at Max Bögl, is convinced. The industrial site of the future is therefore no longer just a consumer - it is also a producer of electricity and heat. If it produces more than its own requirements, it will also be able to participate in the primary control market in the future and provide grid-serving services. In this way, industrial sites can contribute to the energy transition and stabilize grids regionally. Max Bögl's plant in Sengenthal should be able to do this in the future.

Artificial intelligence is now a component of all modern energy concepts and is therefore the focus of the energy technology cluster . One example is the KOSiNeK project, which maps a holistic energy system analysis for Bavaria, Germany and Europe.