Sub-system
When the interconnection queue is longer than the business case can wait, operators generate power themselves. That has moved from backup diesel to serious primary generation — gas turbines, fuel cells, and long-term contracts with existing nuclear plants.
In one sentence
On-site generation is electricity produced at or adjacent to a data-centre site to serve its load directly, rather than being drawn from the grid through a standard interconnection.
The options differ in speed, emissions and permitting difficulty. Reciprocating engines and gas turbines can be procured and installed relatively quickly and are dispatchable, but require a gas connection and an air permit. Fuel cells are quieter and cleaner at the point of use and are being deployed at meaningful scale. Contracting output from an existing nuclear plant provides firm carbon-free power without building anything, which is why those agreements attracted so much attention.
None of these fully substitutes for a grid connection. Most sites want both — generation to get started or to cover the gap, and a grid connection for reliability and for the hours when generation is more expensive than buying.
The most common pattern is temporary: install generation to energise a site while the interconnection completes, then keep it for backup or peak shaving once the grid connection arrives. That reframes generation as a schedule purchase rather than an energy strategy, which is a different investment case.
Combustion generation needs an air permit, and permits for large sources in areas already close to limits are difficult and slow. This is a real constraint on the approach and one reason fuel cells and contracted existing generation are attractive despite costing more per unit of energy.
Buying the output of an existing plant — most visibly, nuclear units — provides firm, low-carbon supply without a construction schedule. The catch is that it does not add generation to the system; it reallocates it, which is why these agreements attract scrutiny from regulators and other customers.
4 of these are marked as a chokepoint: a handful of qualified suppliers, a multi-year lead time, or a single geography.
The most efficient option at this scale, and the one whose order books extend years beyond the schedule a site is trying to meet.
Gas turbinesModular engine packages are what most bridge power is actually built from, because they can be delivered and commissioned in months.
Reciprocating enginesThe non-combustion option, chosen where a permit for a combustion source would not be granted rather than where it is cheapest.
Fuel cellsGas-fired generation needs a firm fuel connection, which has its own queue and its own constraints.
Permitting a large combustion source is frequently the binding constraint, not the equipment.
Buying the output of an existing nuclear, geothermal or hydro unit is the only route to firm carbon-free supply with no construction schedule, and the number of units available to contract is small.
Firm low-carbon powerPlant running alongside the grid still needs an agreement covering synchronisation, protection and any export, so generating at the fence does not remove the utility from the schedule.
Interconnection queuesSeveral operators have contracted for reactors that do not yet exist, on delivery dates late this decade. Until one is licensed and built, the firm low-carbon power those agreements promise cannot be drawn on.
Small modular reactorsOther pages in this map that name On-site generation as something they cannot do without.
What each company supplies at this step, and — where a public figure exists — its share of this specific market — with what that share measures, the period it covers and who published it. Some rows also show the company’s own reported revenue for the segment covering this step, which is a different thing: it says how much this business matters to that company, not how much of the market it holds. Not a ranking and not a recommendation.
Supplies gas turbines and grid equipment for on-site and utility generation.
Supplies solid-oxide fuel cell systems used for on-site data-centre power.
Supplies reciprocating engine generation and power systems.
Contracts firm output from its nuclear fleet to large data-centre loads.
Supplies the turbines behind the largest behind-the-meter plants now being permitted.
Supplies the competing large-frame turbines, and the delivery slot is the scarce good rather than the machine.
Supplies fuel-cell plant as a second route to firm on-site power without a combustion permit.
Whether gas turbine order books ease enough to make on-site generation a fast option again.
Whether fuel-cell deployment scales beyond current pilot-sized installations.
How regulators treat contracts that reallocate existing generation rather than adding it.
Each page explains one technology in plain language, states what it depends on, and names companies by what they supply at that step. Company roles are described qualitatively and deliberately carry no market shares, revenue figures or rankings — those change faster than an explainer can, and a stale number is worse than none. Ticker links point at company pages on this site and are provided for reference only.
Nothing here is investment advice, a recommendation, or a forecast. A company named on a page about a technology is not thereby a good investment, and the chokepoints described are structural facts about supply chains rather than predictions about prices. Technology moves; where a page describes something as unresolved or in development, that was true when it was written.