Can Waste Heat Cool a Data Center? What Resalta’s Simulation Reveals
Cooling is one of the biggest levers data center operators have for cutting electricity use, and one of the hardest to get right when power and cooling are designed as a single system. In our new case study, European energy services company Resalta used Modelon Impact to test a bold idea: run a data center in island mode on onsite gas turbine power, and use the turbine’s waste heat to drive the cooling. The headline result is an estimated 81% reduction in cooling-related electricity. Here’s why that matters, and what engineering teams can take from it.
Why use waste heat to cool a data center?
Every onsite generator produces heat. In a conventional design, that heat is rejected to the atmosphere while compressor chillers draw electricity to make cooling. An absorption chiller flips that logic: it runs on thermal energy, in this case gas turbine exhaust, so electricity is needed only for pumps, fans, and other auxiliary equipment. For operators facing grid constraints or building off-grid, pairing power and cooling this way means more of the generated electricity goes to IT load instead of chillers.
Resalta’s results at a glance
- An estimated 81% reduction in cooling-related electricity compared with conventional compressor chillers
- About 12.7 GWh of electricity saved per year
- A modeled Power Usage Effectiveness (PUE) of about 1.1
- 75% less time spent on model preparation and scenario evaluation, from roughly one month to one week
The full case study covers the system architecture, operating assumptions, and energy flow analysis behind these numbers.
Three takeaways for data center engineering teams
- Evaluate the system, not the components. A chiller’s datasheet efficiency says little about how it behaves when its heat source, heat rejection, and load all shift at once. Resalta modeled the gas turbine, absorption chiller, dry cooler, and backup heat pump as one connected system, which is where the real performance picture emerged.
- Simulate a full year, including the transitions. Seasonal temperature swings, changing IT loads, startups, shutdowns, and maintenance windows are where integrated designs succeed or fail. Dynamic simulation captures this transient behavior; steady-state spreadsheets can’t.
- Build models you can reuse. Resalta started in Excel, but interconnected formulas became hard to maintain as the concept grew. With reusable, physics-based components in Modelon Impact, the team can swap equipment performance maps or control strategies for each new customer project instead of starting over.
See the full story
Read the full Resalta case study for the system schematic, energy flow analysis, and complete results. Exploring cooling concepts for your own facility? Learn more about Modelon’s data center solutions or contact our team.
