Home / Innovation / Case study 01
OUR INNOVATION · CASE STUDY 01

Fuel cell + absorption chiller. 85% efficiency.

A trigeneration system that turns waste exhaust heat into useful cooling — pairing a solid-oxide fuel cell with a lithium-bromide absorption chiller to deliver power and chilled water from one fuel input.

85% Combined efficiency
2.4× vs grid + e-chiller
−62% CO₂ per kWh
3.2yr Payback period
01 — THE PROBLEM

Traditional power generation throws away two-thirds of its fuel.

A typical combustion generator converts only ~35% of its fuel energy into electricity. The remaining ~65% leaves the engine as waste heat — radiator coolant, exhaust gases, jacket water — and is vented to atmosphere. For sites that also need cooling, that fuel gets burned a second time at the utility scale to drive electric chillers, compounding the inefficiency.

For our data-center and hospital customers, this means a 1 MW IT load typically requires ~1.4 MW of generation capacity plus ~300 kW of electric chiller load — burning nearly 3× the fuel needed for the actual useful work.

02 — OUR APPROACH

Recover the heat. Drive cooling with it. Don't burn fuel twice.

We pair a solid-oxide fuel cell (SOFC) with a lithium-bromide absorption chiller. The fuel cell electrochemically converts natural gas or hydrogen into electricity at ~60% efficiency — already much higher than combustion. Its 500°C exhaust stream then feeds a single-effect absorption chiller, which uses thermal energy (not electricity) to drive a refrigeration cycle.

  • SOFC delivers clean, low-NOₓ electricity at ~60% LHV efficiency
  • High-temperature exhaust (~500°C) routed through a recovery heat exchanger
  • Single-effect LiBr/H₂O absorption chiller produces 7°C chilled water
  • No mechanical compressor — chiller runs silently with minimal moving parts
  • Combined system efficiency reaches 80–87% (HHV basis) under design load
// KEY INSIGHT
A grid-powered electric chiller has a COP of ~5, but only after the grid has thrown away 65% of fuel as heat. Our absorption chiller has a COP of only 0.7 — but it runs on heat we'd otherwise vent. Net result: we get ~3.5× more useful cooling per unit of fuel.
03 — THE NUMBERS

Side-by-side: conventional vs SPM trigeneration.

Based on a 500 kW electrical / 300 ton cooling load — typical of a small data hall or hospital wing.

Metric Grid + e-chiller SPM trigen system
Primary fuel input 2,310 kW 965 kW
Electrical efficiency 35% 60%
Heat recovery 0% 25%
Combined efficiency 35% 85%
CO₂ emissions (kg/MWh) 485 185
Annual fuel cost $1.42M $594K
Cooling redundancy Grid-dependent Islandable
04 — PILOT DEPLOYMENT

500 kW trigeneration plant, regional data center, Texas — 2024.

We installed our first commercial trigeneration package at a Tier III edge data center near San Antonio. The site previously ran on grid power with three 150-ton air-cooled chillers — combined PUE of 1.62 and a recurring concern about grid stability during summer load peaks.

The SPM system replaced both power and cooling with one integrated package: a Bloom Energy ES5-300kW SOFC array (scaled to 600 kW), a Yazaki WFC-SC50 absorption chiller (175 tons primary + 125 tons backup electric), and our proprietary heat-recovery and controls package.

  • Commissioning: 14 weeks, including civil work and gas service upgrade
  • PUE post-install: 1.18 (vs 1.62 baseline)
  • Grid imports: reduced 92%; site is islandable for 96+ hours on stored fuel
  • Measured combined efficiency: 83.4% (vs 85% design target)
  • Payback: 3.2 years against avoided grid + chiller capex/opex
// OPERATOR QUOTE
"We expected to swap one set of headaches for another. Instead we ended up with a plant that's quieter, cheaper to run, and resilient through two ERCOT alerts in its first summer."

— FACILITIES DIRECTOR, PILOT SITE
05 — WHAT'S NEXT

Scaling to 5 MW packages and hydrogen-ready stacks.

The pilot validated the thermodynamics and the controls package. Our engineering team is now productizing two standard configurations: a containerized 1 MW / 600-ton unit for edge and healthcare sites, and a 5 MW / 3,000-ton skidded plant for hyperscale customers.

Both will ship with hydrogen-capable SOFC stacks — letting customers transition to zero-carbon operation as green hydrogen pricing falls, without replacing the heat-recovery side of the system.

  • Q3 2026: 1 MW containerized package — production release
  • Q1 2027: 5 MW skidded plant — engineering samples
  • Q3 2027: H₂-blend testing at pilot site (up to 40% by volume)
  • 2028: Full hydrogen capability on production units

Could trigeneration work at your site?

Our engineering team will run a free feasibility study based on your electrical and cooling load profile, fuel supply, and capex constraints.

Request a feasibility study