Skip to content
KomorebiEnergy + AI

The case

The application that gets picked.

Buy the interconnect. Permit the cooling. Host the IT.

The interconnect approval queue decides who wins compute capacity — not GPUs. Undifferentiated 50 MW interconnect filings wait. You are not buying solar. You are buying an existing grid connection and filing a load the operator can say yes to. Texas (Pecos) is the worked example. The same two clocks exist on every grid: power out, and permission to pull power in after dark.

CFO
Economics

About $150M, then $370M, then $405M a year.

CTO
Cooling & generation

50.7 MWth. A 250 MWh cold storage tank.

Counsel
Permitting

Same fence. Phase 3 is full off-grid.

Firm offtake, flex hours, and what the overlay costs.

The purchase

Buy the interconnect. Phase 1 is 58.5 MW. The fence then grows.

Phase 1 rejects 50.7 MWth and hosts 50 MW. The same fence then hosts 153.72 MW, still on the grid, and 175.04 MW full off-grid.

Acquire
100 MWdc, commercial operation 2021 or earlier. The generation connection is already built.
Overlay
Cassette rows, a buried pipe, a 250 MWh cold storage tank, and an air-cooled chiller. No dry coolers. No towers.
Host
Phase 1 is 58.5 MW. Phase 2 is 153.72 MW, still on the grid. Phase 3 is 175.04 MW, full off-grid.
Earn
Planning value: about $150M, then $370M, then $405M a year. Phase 2 and Phase 3 are after operating cost.
Time
A new interconnect is years. This plant already exports. Phase 3 does not import.

CFO

  • Phase 1 uses $3 million per MW-year. Phase 1 is 58.5 MW, about $175.5 million. Phase 2 and Phase 3 are not $3 million times every megawatt.
  • Phase 2 is about $369.6 million a year, about 1.87 years versus Phase 1. Phase 3 is about $405.2 million, 2.22 years.
  • Phase 2 is 153.72 MW, still on the grid. Phase 3 is 175.04 MW, full off-grid. Phase 3 is not an 18 MW add-on.
  • Shared reject is about 35 acres, then 76, then 94. Jacket water sets the later acres.
  • The farm bid is a separate energy-asset purchase. The overlay is the cooling invoice. GPUs stay on the customer’s books.

CTO

  • 17 June at 15:00: cassette rows 10.5 MWth, buried pipe 3.1, tank 35, chiller 2.1. That is 50.7. The chiller nameplate is 8.45 MWe at COP 6, on surplus PV.
  • The cold store is a 250 MWh cold storage tank. The buried pipe is not that store.
  • No dry coolers. No cooling towers. Shared reject is 35 acres, then 76, then 94.
  • Phase 2 heat engine is about 5.8 MW. Phase 3 heat engine is about 13.2 MW. Firm IT is 153.72 MW, then 175.04 MW.

MWdc is the farm you buy and pipe. MW AC is what the interconnect carries. MW IT is the compute you contract.

First commercial plant

Scale from a 10 MW Phase 1 plant.

The slider is only the Phase 1 trial. 153.72 MW and 175.04 MW are the full fence. They do not grow with the slider. Phase 3 is full off-grid. The dollar line is the $3 million Phase 1 rate, not the Phase 2 or Phase 3 net.

Phase 1 trial
10 MW
Phase 2 adds, full fence
103.7 MW
Phase 2 · still on the grid
153.72 MW
Phase 3 · full off-grid
175.04 MW
Repowered PV
89.8 MWdc
Shared reject on the full fence
35 → 76 → 94 ac

Illustrative enabled-compute value, at $3M per incremental MW-year: $30.0M / year

Qualify this size

Permitting and economics per additional MW of AI

Capex, water, noise, air, and time are the same decision. Cooling-first reuses a permitted interconnect and keeps a smaller chiller, not a chiller for the whole duty. Time is to additional IT, not to a greenfield campus.

Path$M / AI MWFuel exposureGen. waterCooling waterTimeEmissions / noise
Komorebi cooling-first$0.72–0.78MLowNone addedLow6–12 moLow
Reciprocating gas engines$1.5–2.2MMediumMedium–highMechanical cooling12–24 moMedium
Komorebi full platform$3.2MLowNone addedLowStagedLow
Aeroderivative simple-cycle GT$3.2–4.6MHighHighMechanical cooling24–36 mo+High
Fuel cells$4.9–8.0MHighLowSite cooling still required18–30 moLow stack, fuel upstream

Reciprocating engines are typically more efficient than simple-cycle aeroderivative turbines. Combined-cycle GTs are not in this table. Fuel-cell generation-side water is Low in normal operation.

Indicative cost of energy that reaches the halls — $ / MWh

Cooling-first cuts overhead at the meter, so more of each purchased or generated MWh reaches IT. That is not a generation LCOE. Generator rows below are conventional generation LCOE.

Path$ / MWhBasis
Komorebi cooling-first$40–60Delivered to AI compute — overhead reduction
Reciprocating gas engines$55–85Generation LCOE
Komorebi full platform$35–55Delivered to IT — duration and ORC hours
Aeroderivative simple-cycle GT$70–110Generation LCOE
Fuel cells$95–145Generation LCOE

Illustrative Pecos, Texas planning days (17 Jun, 22 Jun, 6 Dec 2024; NSRDB GOES v4, 5-min, LST UTC−6). 50 MW firm IT on 100 MWdc. The availability spec is 99.9%. These three days show zero unmet firm load. They are not the full-year result. Phase 2 stays on the grid and reaches 153.72 MW. Phase 3 is full off-grid at 175.04 MW of firm IT. The published bars do not yet include engine exhaust. Not a contract.

Worked example · 100 MWdc existing PV · Pecos, Texas

Phase 0 is 50 MW, Phase 1 is 58.5 MW, then 153.72 MW, then 175.04 MW on one fence.

Phase 1 is the cooling plant: fanless cassette rows, a buried pipe, and a 250 MWh cold storage tank, sized to reject 50.7 MWth. That is 50 MW of firm IT. Phase 2 stays on the grid, adds engines and the heat train, and reaches 153.72 MW. Phase 3 is full off-grid at 175.04 MW of firm IT.

Phase 1 — cooling
50 MW AI

50 MW of firm IT. The plant rejects 50.7 MWth. The chiller is 8.45 MWe at COP 6 and runs on surplus PV.

Phase 2 — grid + engines
153.7 MW AI

Grid plus engines. 153.72 MW. About 1.87 years versus Phase 1.

Phase 3 — off the grid
175 MW AI

Full off-grid. 175.04 MW. Same fence. 2.22 years versus Phase 1.

Starting plant is 100 MWdc on the same land. Phase 1 reject is about 35 acres. Phase 2 shares about 76. Phase 3 shares about 94 and is full off-grid. Jacket water sets the later acres. No mirror acreage is drawn.

$150M / year — Phase 1 firm

50 MW of design-target 99.9% firm AI × $3M per MW-year planning value. A planning year, not a live meter.

About $370M / year

Phase 2 is 153.72 MW. About $369.6 million a year. About 1.87 years versus Phase 1.

175.04 MW — full off-grid

Phase 3 is full off-grid at 175.04 MW. About $405.2 million a year. 2.22 years versus Phase 1.

Phase 3 is full off-grid at 175.04 MW. About $405.2 million a year after operating cost. 2.22 years versus Phase 1. Same fence.

The three published days are 17 June, 22 June, and 6 December. The hall is 50 MW on each of them.

Compared with

The metric is firm IT.

Phase 1 hosts 50 MW of firm IT. On 17 June at 15:00 the chiller supplies 2.1 MWth of the 50.7. Nameplate is 8.45 MWe at COP 6, on surplus PV.

  • A new interconnect or a new plant

    A site with no interconnect and no heat rejection this year hosts nothing. That is a 5–7 year clock. This farm already exports. Overnight import is a separate filing for the 50 MW hall.

  • Chiller on surplus PV, not the whole load

    Same 50 MW hall. On 17 June at 15:00 the chiller supplies 2.1 MWth of the 50.7. Nameplate is 8.45 MWe at COP 6 and runs on surplus PV.

  • Grid, then full off-grid

    Phase 2 stays on the grid and reaches 153.72 MW. Phase 3 is full off-grid at 175.04 MW of firm IT.

  • More batteries as the cooling plant

    A pack can run compressors. It cannot dump server heat. Every MWh spent on chillers is a MWh that never reaches GPUs.

Why the overlay beats a new plant

  • Reuses the interconnect and land you already have
  • About 35 acres of radiator rows, plus a smaller chiller, reject 50.7 MWth for the 50 MW hall
  • The chiller nameplate is 8.45 MWe at COP 6. On the hottest hour it supplies 2.1 MWth of the 50.7.
Request the briefing pack