When the answer is not yet
In 2022, EirGrid, the operator of Ireland's electricity transmission system, indicated that new data center connections in the Dublin region were unlikely before 2028. It was not a formal ban. It was an operator explaining that the local grid could not safely absorb more large loads until new transmission and generation were built. The effect, for developers, was much the same as a moratorium. Dublin had become one of Europe's largest data center markets in part because of tax policy, connectivity and climate, and within a few years the binding constraint had become the wires.
Ireland's regulator, the Commission for Regulation of Utilities, had already set conditions on new data center connections in a 2021 decision, asking the operators to consider whether a site was in a constrained area and whether it could bring its own generation or flexibility. In December 2025, it published a new Large Energy Users connection policy. New data centers will be expected to provide on-site or nearby generation or storage matched to the demand they request, and to meet at least 80 percent of their annual demand from additional renewable projects in Ireland, with a six-year glide path to get there. EirGrid and ESB Networks were asked to publish the detailed application processes in 2026.
The Irish case teaches a pattern that recurs. Grid constraints tend to arrive first as operator statements, then as regulatory conditions, and only occasionally as outright prohibitions. Something similar happened in Northern Virginia. In 2022, Dominion Energy Virginia told data center customers in parts of Loudoun County that new connections could be delayed for several years while it built transmission capacity. No law changed. The effective supply of power in one of the most sought-after data center corridors in the world simply moved further into the future.
This is the sense in which location has become an advantage. For most of the industry's history, sites were chosen on land cost, fiber, taxes and proximity to customers, and power was assumed to follow. When power no longer follows on a predictable schedule, a site that already has it, or has a credible and dated path to it, is worth more than an otherwise identical site that does not. The advantage is not in the land. It is in the queue position, the signed service agreement and the transmission already planned.
How durable is that advantage? It depends on why the constraint exists. A constraint caused by a single missing transmission line may disappear when the line is energized, and the premium attached to nearby sites may shrink. A constraint caused by a regulator's policy choice, such as a requirement to bring matching generation, may persist and reshape the whole market. Distinguishing the two requires reading the operator's transmission plan and the regulator's decisions, not the marketing brochures.
There is also a question of where development goes when a market closes. When Dublin tightened, operators looked to other European markets; when parts of Loudoun slowed, developers moved further south and west in Virginia and into other states. A constraint in one place is often an opportunity somewhere else, although the new place may simply be earlier in the same cycle.
Faraday's read is that the most valuable signal about a market is not its current capacity but the date on its next transmission upgrade and the honesty of the operator describing it. EirGrid's early warning, unwelcome as it was, gave the market years to adjust. An operator that does not publish its constraints leaves developers to discover them the expensive way.
The newcomer's exercise is to ask of any promising market three dated questions. When was the last major transmission addition energized? When is the next one scheduled, and has it been approved? Has the operator or regulator said anything, formally or informally, about limits on new large loads? The answers are usually public, if unglamorous, and they often explain prices that otherwise look puzzling.