“The energy transition will fundamentally change the future energy landscape,” said Peter Weckesser, chief digital officer at Schneider Electric. That’s because every prosumer, regardless of size or complexity, becomes a microgrid, managing the flow of power within a building and how it accesses the main power grid.
This increasingly decentralized and dynamic energy system makes energy management even more complex. Both professional consumers and power transmission companies will need to predict how much power each microgrid will draw from the network, how much power it will deliver, and when it will deliver it. If the network operator cannot manage all these control points, the grid will become unstable.
This new complexity poses significant optimization challenges, both within each microgrid and in the broader energy system, Weckesser says. It argues that optimization relies on high-performance artificial intelligence (AI), which it says is “the most powerful tool for demand-side management of energy consumption.”
Even as electrification and the broader energy transition create complexities, they also bring new solutions and opportunities. As Weckesser points out, “EV infrastructure will be the greatest energy storage capacity we have.” However, it can only play this role if it can be managed to meet the needs of the grid.
Tackling control systems’ biggest challenges
Weckesser explains that, taken together, these complexities represent the biggest control system challenge grid operators have ever faced. But AI can help all parties, from homeowners to system operators, optimize their assets and provide answers to questions such as when consumers should consume, store, and distribute electricity.
Until recently, the industry took a more classical approach to creating demand models. Even with today’s level of complexity, this approach cannot keep up. Weckesser warns that the problem will become significantly more difficult as the transition accelerates. “New prosumers and new nodes are added to the grid every day. Manually maintaining these grid models becomes a very difficult task.”
