Autonomous cooling, costs near zero.
When a DHC network is not there or contracting one does not make sense, thermal storage makes you independent. Capture the heat, store it, manage it in the convenient hours. And in a blackout the cooling does not stop.
Reference: 10 MW IT data center, 150 kW per rack.
You are far from district heating. Or you want to stay free of external contracts.
Energy for cooling remains the data center's first operating cost, and with AI densities it grows non-linearly. Thermal loads are not uniform: daytime peaks, underused night-time hours.
Without a buffer you cannot shift demand in time, so you pay everything at peak price. And on cooling, blackout risk stays exposed: no UPS protects the data center's most critical asset.
You need autonomy. Technological, contractual, operational.
Three economic levers, one machine.
RHP coupled with a thermal buffer (tanks or PCM): captures excess heat, stores it, releases it during the convenient hours. Thermal peak shaving without touching the IT loads. In parallel, Thermal UPS: continuous cooling even in a blackout. The platform is DC-native, compatible with the 800V bus of high-density AI racks: you eliminate the double AC/DC conversion, reduce Joule losses, simplify the electrical distribution.
Electricity savings
PUE from 1.12 to 1.01, with auxiliaries self-powered via ORC.
Avoided CAPEX
No traditional cooling batteries: Thermal UPS already built in.
Reduced conversion losses
DC-native architecture, native integration with the 800V bus.
Future-ready
If tomorrow a DHC network reaches you, you are already ready to connect.
What changes, in figures.
Reference: 10 MW IT data center, 150 kW per rack.
| Electricity savings (PUE 1.12 → 1.01) | ~€1M/year |
|---|---|
| Cooling-battery CAPEX avoided | ~€900K |
| ORC electricity self-consumed | 2,356 MWh/year |
| DC bus 800V | Native compatibility |
| Water consumption | <2 m³/year |
| CO₂ avoided | 18,200 tCO₂/year |
| Payback | 3.1 years |
| Customer IRR | ~30% |
Waste heat remains available for a future DHC connection, with no additional structural work.
When this solution is not the right choice.
Storage mode does not make sense in these cases. We tell you upfront, before you book the call.
- You do not have enough external footprint for the storage tanks (significant ground space)
- Your CAPEX budget is capped at zero (consider EaaS mode)
- You already have a DHC network within 2 km (DHC mode adds direct revenue from heat sale)
- The 800V DC bus advantage only pays off in full with compatible power infrastructure
In those cases: consider RHP for DHC for direct revenue, or Energy as a Service for zero CAPEX.
Validated technology. Ready to scale.
Pilot plant operational since June 2026 on the base technology (small-scale Carnot battery). Measured performance exceeds expectations on COP and heat recovery.
Reversible ORC patent filed October 2025. Co-inventors: Marco Margotti (CEO, 20 years B2B + 6 years on ORC Kaymacor) and Giuseppe Toniato (CTO, 30 years of thermodynamic systems).
The technology is validated. We are ready to bring it to commercial plants up to 10 MW IT and beyond.
Let's size the footprint and the business case for your site.
30 minutes of technical analysis: available footprint, storage sizing, preliminary payback.
See also: RHP for DHC· Energy as a Service