Constraint, Moratorium, and the Spillover Ring

A market can be the most desirable address in Europe and still be closed for construction. That paradox defines the current FLAP-D core, and understanding it is the first practical skill of international siting.

Begin with Amsterdam. In 2019 the municipality, alarmed by the pace of hyperscale construction, its land use, and its grid draw, imposed a temporary halt on new data-center permits. When the region reopened, it did so with tighter rules on where facilities could go, how power-efficient they had to be, and how much land they could consume. The Dutch national grid operator has since warned of congestion across large parts of the country, meaning that even where a permit is available, a timely grid connection may not be. The lesson is that a moratorium is only the visible part of the constraint. Underneath it sits a physical limit: the transmission and distribution network cannot absorb unlimited new load without reinforcement, and reinforcement takes years.

Dublin is the sharper case. Ireland's grid operator, facing the reality that data centers already consumed roughly a fifth of the country's electricity, effectively paused new grid connections for data centers in the Dublin region through the middle of the decade. The Irish situation is a clean illustration of the course thesis: the buildings were welcome, the tax posture was welcome, the jobs were welcome, but the electrons ran out. Ireland is a small system with limited interconnection to the rest of Europe, so it could not simply import its way around the shortfall. Demand that would have gone to Dublin did not disappear; it went looking elsewhere.

Frankfurt and London are less absolute but still constrained. Frankfurt has introduced zoning that channels data centers into designated areas and away from residential districts, and its grid, while robust, faces the same queue pressure as everywhere else. London's constraint is famously concentrated in its western corridor, where a cluster of data centers around the Slough area drew so much power that the local network operator warned new housing developments might have to wait for grid capacity. When data centers and homes compete for the same substation, politics enters the siting decision in a way it never did before.

Out of these constraints comes the spillover ring: the set of secondary markets that absorb demand the core can no longer take. In Germany, that means Berlin and Munich picking up what Frankfurt cannot. In the Nordic direction, it means Oslo, Stockholm, and the Finnish and Norwegian interior, where cheap hydro and cool air make large campuses cheap to run. Around London, it means regional English markets and, increasingly, the north. The spillover is not random. It follows the same logic that built the core: it goes where power is available, where fiber can reach back to the interconnection hubs with acceptable latency, and where land can actually be entitled and energized.

Here is where a practitioner earns their keep. The naive move is to chase the cheapest land in the spillover ring. The disciplined move is to ask three questions in order. First, can this parcel get a firm grid connection, and on what timeline? A signed connection agreement with a real energization date is worth more than any amount of cheap acreage. Second, is the land entitled for this use, or does it require a rezoning fight that could add years? Third, does the fiber path back to the interconnection core meet the latency budget of the intended workload? Training clusters tolerate more latency than real-time inference or financial trading, so the answer depends on what will run there.

Notice that all three questions converge on the same scarce thing: entitled, powered land. Cheap land that cannot be energized for seven years is not cheap; it is a stranded option. Expensive land with a firm connection agreement and existing zoning is, in the only accounting that matters, the bargain. This inversion of intuition, where the constraint rather than the price sets the value, is the mental model the rest of this course builds on. The spillover ring rewards operators who can read grid queues and entitlement status the way a previous generation read cap rates. The core taught the industry where demand lives. The constraints are now teaching it where supply can actually be built.