Hospital optimizes energy use with digital twin

The energy transition has connected many renewable energy sources to the power grid in recent years. As a result, energy flows less predictably and prices fluctuate based on supply and demand. In 2025, 57% of the energy consumed in the Netherlands came from renewable sources. This share is leveling off because supply and demand are not sufficiently aligned. We're addressing this with Energy Flexibility Services. "We start by working with customers to identify how they unlock of the flexibility already in their energy system," explains Jeroen Bakker, who helps customers make their energy systems more flexible. "Then we help them increase and optimize that flexibility."

MST's energy plant houses two Combined Heat and Power (CHP) units that generate heat from gas. They ran year-round, without any insight into the actual heat demand. As a result, a lot of heat was lost because it wasn't fully used. In 2023, we built a digital twin of the physical energy system to explore how we could optimize it. We fed the twin with real-world field data. "The simulation showed that it wasn't necessary at all to run both CHP units continuously," says Kuno Kamstra, sustainability and energy advisor at MST. "Now one CHP unit is switched off, while the second only runs when it's needed."

In early 2025, we installed the digital twin model – called BEOS (Building Energy Optimization System) – on a server at MST so it could learn from real-time energy meter data. Since August 2025, it has been controlling the gas engines in real time. CHP efficiency has risen from 67% to 89%, as heat overproduction no longer occurs. CO2 emissions have already dropped significantly and energy costs have decreased considerably. "The model responds automatically to changes in the energy market," says Kamstra. "The energy tax on CHP gas has been rising every year since 2025. We used to manually calculate whether it was still profitable to run the engines. The model now does that entirely for us, saving time and cutting costs. In time, we'll connect to the district heating network. By 2035, we want to be completely off gas, but until then we want our gas engines to run as efficiently as possible."

The simulation showed that autonomously controlling MST's energy system can reduce emissions by 35% and save 15% on energy. Even greater savings are possible if MST switches from a fixed to a dynamic energy contract. "We're locked into our current contract until 2028. After that, we're considering switching to a variable electricity contract," says Kuno. "For gas, we don't want that, as gas prices can spike suddenly. The beauty of the digital twin is that you can run various scenarios. We'll also use the model to simulate how an additional heat pump could perform." Jeroen adds: "The scenario that saves the most money doesn't necessarily reduce the most CO2. That's why it's essential to clearly define your goal upfront: do you primarily want to cut costs, reduce emissions or both?"

After heating, MST wants to autonomously control its cooling systems in a second phase. A welcome side effect, says Kuno, is that hospital staff increasingly come up with their own initiatives to improve sustainability. "There used to be a sticker next to every door saying 'lights off.' People often didn't bother, assuming the cleaner would take care of it – but they only come once a day. Now people come up with energy-saving tips themselves and spontaneously switch off equipment when it's not needed. Especially equipment that can be quickly restarted. It's great to see how this initiative has taken on a life of its own."

July 2026