Large electricity users can no longer treat a utility connection as a simple site prerequisite. Current US reforms point toward a negotiated operating package with explicit study, cost, flexibility and evidence terms.
The regulatory file is moving beyond a connection request
The observable change is not merely higher electricity demand. On 18 June 2026, the US Federal Energy Regulatory Commission issued show-cause orders covering the six regional grid operators under its jurisdiction. The operators and their transmission owners were directed to justify why existing tariffs remain just and reasonable without clear, consistent provisions for large-load customers, or to propose changes. FERC identified five reform areas: faster application and study processes, cost-shift prevention and cost transparency, co-location rules, transmission services for flexible large loads, and study processes for generation serving nearby or co-located loads [1].
Those categories convert what once looked like a utility-procurement task into a multi-party contract-design problem. The factual point is that regulators are asking grid operators to make the allocation of study rights, upgrade costs and operating responsibilities more explicit [1]. The rules will remain regional, but the direction is clear: a large customer seeking speed to power may need to demonstrate not just where it wants service, but how its demand behaves, what generation sits nearby, which upgrades it triggers and who carries the risk if the expected load changes.
Forecast uncertainty is now part of infrastructure planning
The US Department of Energy's July 2026 draft National Transmission Needs Study says long-term projections are evolving rapidly because of expected growth from large loads and data centres. It also warns that nationwide modelling available when the study was written may not fully reflect the current landscape of future energy markets [2]. This is important because a transmission plan depends on assumptions about when load appears, how quickly it ramps and whether associated generation or network upgrades arrive on the same timetable.
PJM's 2026 materials illustrate the timing mismatch. Its reserve-target shortfall analysis describes data-centre developments as materialising faster than transmission or generation can be permitted and built, and treats large-load growth as a system-wide planning issue rather than only a Northern Virginia constraint. The same analysis says its forecast has shifted toward faster winter peak growth than summer peak growth [3]. These are PJM's planning findings, not a universal forecast for every grid, but they show why a headline megawatt request is an incomplete description of infrastructure need.
Oakhampton inference: deliverable power needs its own transaction schedule
Oakhampton's inference from these sources is that serious large-load projects increasingly require a power schedule that is as disciplined as the land, construction and financing schedules. That schedule would distinguish requested capacity from studied capacity, interim service from firm service, customer-funded works from socialised network works, and ordinary operation from agreed curtailment or flexibility. It would also identify the dependencies that can invalidate a target energisation date: study completion, tariff approval, equipment procurement, network construction, generation availability and operational testing.
This is an inference about transaction control, not a claim that every project will receive the same tariff or contract. FERC expressly leaves room for regional variation [1], while DOE notes that the underlying demand outlook is changing quickly [2]. The practical consequence is that parties comparing sites or acquisition opportunities should avoid treating quoted megawatts as interchangeable. Two sites with the same nominal capacity can carry different study status, upgrade exposure, firmness, curtailment rights, ramp assumptions and delivery evidence.
A bankable evidence pack separates capacity, cost and behaviour
A bounded diligence file can separate three questions. First, capacity: what amount has been requested, studied, approved and physically deliverable, and on what dates? Second, cost: which deposits, facilities, network upgrades and security obligations belong to the customer, and what happens if the load is delayed or reduced? Third, behaviour: what ramp profile, power factor, redundancy, on-site generation, storage, co-location or interruptibility assumptions were used in the study? FERC's reform categories connect all three questions because study design, cost transparency, co-location and flexible service affect one another [1].
Evidence quality matters more than a single representation. Useful records include the executed service agreement, accepted study scope, milestone notices, upgrade estimates, tariff references, metering requirements and correspondence confirming the applicable queue or service status. PJM's description of a shorter load-development cycle than the build cycle for supply and transmission [3] reinforces the need to date-stamp each dependency. A stale capacity letter may describe an earlier network model, an earlier ramp date or a project configuration that has since changed.
Remaining uncertainty belongs in the commercial model
Several uncertainties remain open. The FERC proceedings begin a region-specific process; they do not establish one final national large-load tariff [1]. DOE's needs study was published as a draft for consultation, and its modelling caveat means later analysis may revise the scale or location of need [2]. PJM's findings are specific to its footprint and planning assumptions [3]. Project-level outcomes will also depend on local utility rules, the credibility of the load forecast, equipment availability, permitting, generation retirements and the customer's ability to meet deposits and operating conditions.
The defensible conclusion is therefore narrower than a demand forecast. Grid access for large loads is becoming an evidence-heavy commercial package in which capacity, timing, cost allocation and operating flexibility must reconcile. Where those elements do not yet reconcile, the gap should remain visible as a condition, contingency or scenario rather than being converted into an unsupported energisation promise. That approach preserves the distinction between current facts, Oakhampton inference and matters that only the relevant grid operator, utility or executed agreement can resolve.
Sources
- Fact Sheet: FERC Takes Action to Supercharge America's Grid for Efficiency, Reliability, and a Bold Energy FutureFederal Energy Regulatory Commission · 18 June 2026
- National Transmission Needs Study Draft for Consultation and Public Comment, July 2026US Department of Energy · 9 July 2026
- 2027/2028 Base Residual Auction Reserve Target Shortfall ReportPJM Interconnection · 19 February 2026