Underwriting energization risk: who carries a slipped date

The instinct is to treat energization as the last box on a construction schedule — a date the project arrives at. Capital treats it as an input it underwrites, and the distinction is the whole subject of this page.

A completed building is not a cash-flowing asset. It becomes one only when it can draw firm power at the capacity its contracts assume, and the interval between physical completion and that moment is a period in which capital is fully deployed and earning nothing. So the party financing the project does not underwrite the developer's forecast energization date; it underwrites the confidence interval around it — how wide the range of plausible dates is, and what the structure costs at the far end of that range.

That reframing changes what diligence is for. The question is not "when does the developer say power arrives." It is "what is this date actually resting on, how firm is the least firm thing under it, and what happens to the structure if it lands six or twelve months later than planned." A date presented without that interval attached is a forecast of a forecast, and it is priced as one.

The width of the interval is set by two things the developer does not control: a place in an interconnection queue, and a place in a manufacturing queue. Both now run in years, and both are the binding input on the date.

The grid side is the first. A grid interconnection is not a service request filled on demand; it is a studied position in a queue, and the queues in the major markets are long and getting longer. The typical time from request to commercial operation now runs past five years. (Lawrence Berkeley National Laboratory, as of August 11, 2026) The demand surge behind that is not evenly spread — in Texas the large-load queue, most of it data centers, quadrupled in roughly a year. (Utility Dive / ERCOT, as of August 11, 2026) The stage the position has reached inside that queue, not the developer's schedule, is what sets the honest date: a request that has not cleared its studies is a forecast, an executed interconnection agreement is a position. The document ladder behind that distinction is the subject of verifying a power claim, and the mechanism itself is set out at interconnection queue. The short version a lender applies: a will-serve letter is not firm capacity, and what a financeable date rests on is a contract for transmission capacity by a date certain.

The equipment side is the second, and it is often the one that actually binds. The transformers, switchgear and turbines a site needs are ordered years ahead and paid for long before there is an asset to secure them against. Large power transformers have carried lead times measured in years rather than months. (Wood Mackenzie / Power Magazine, as of August 11, 2026) Where a project turns to on-site generation to get ahead of the grid queue, it inherits the turbine queue instead — and the dominant heavy-duty turbine makers have been effectively sold out into the late 2020s. (Utility Dive, as of August 11, 2026) A date that assumes equipment arriving faster than the order book allows is not a schedule; it is a wish with a drawdown attached.

A slip in the energization date is expensive in a specific, structural way, and naming the mechanism is more useful than any single dollar figure for it.

Before energization, every layer of the project is spending and none is earning — land carry, construction interest, equipment progress payments, standing charges on capacity contracted but not yet usable. A delay extends that period. It does three things to the structure at once, and they compound:

Whether a slip is manageable or fatal is decided by who carries it, and that is set in the commencement and abatement mechanics long before the delay happens.

The pivotal choice is how revenue commences. Fixed-date commencement — rent or offtake payments that begin on a stated calendar date whether or not the facility is energised — pushes the timing risk onto the developer and its equity, because the obligations start on schedule while the cash flow does not. Energization-linked commencement — payments that begin only when the facility can take power — pushes the risk the other way, onto whoever holds the debt, because the cash flow the structure was sized against simply arrives late. Neither is wrong; which one a structure uses is the allocation of energization risk, stated in the one clause that decides it.

Around that sit the tools that make the risk bankable rather than open-ended. The structure a rated financing looks for treats a power interruption as excused performance with capped, proportional abatement and no termination right — the tenant keeps paying, debt service is paid first, and any rent credit for lost capacity comes from residual cash before equity, so coverage never has to rely on reduced cash flow. A termination right triggered by delay is the dangerous term, because it can defeat the very support the structure was built on: a lease backstop that only activates once the lease commences is worth nothing if a long enough delay lets the tenant walk before commencement. The generic allocation tools — target and longstop dates, delay liquidated damages, phasing — are defined at energization; this page's point is narrower, that the choice among them is an underwriting decision made at structuring and priced accordingly, not a formality discovered later.

The feature of energization risk that most changes how it is underwritten is that it is correlated across the layers of the stack rather than independent of them.

A lender holding a diversified book relies on its exposures not all going wrong at once. Energization risk breaks that assumption inside a single project. The same transformer that is late to energise the facility is late to the lease that commences on energisation, late to the offtake that begins when the tenant can take capacity, and late to the compute that cannot be installed and contracted until there is power to run it. One delay does not move one obligation; it moves all of them, in the same direction, at the same time. That is why a rating agency testing the date tests it hard, and why the risk cannot be netted against the rest of the structure — there is nothing inside the project uncorrelated with it.

This is also the point at which energization risk becomes a tenor problem. Capital is arriving in these projects earlier in the development cycle than it used to, before the asset is stabilised and leased — which means more of the stack is exposed during exactly the pre-energization window where the correlation bites. Separating the layers so each is financed against its own date and its own dependency, rather than one facility against a single date every layer shares, is the structural response, and it is the argument that runs through the data-center capital stack.