Summary

  • TenneT’s first Dutch Flexibility Needs Assessment puts North Holland’s modeled upward transmission-network need at 307 GWh in 2030 and 401 GWh in 2035, a 30% increase.
  • The report says the remaining demand-related overloads are concentrated in one area and stem from assumptions about data-centre development; it says those assumptions could overstate actual growth.
  • ACM approved the revised assessment after requiring a clearer account of uncertainty. The next report should use multiple scenarios and a realistic forecast of flexibility expected to be available.

The grid’s new load arrives first in a scenario

The Dutch electricity forecast has a neat trick: it can identify a shortage before the data centres behind part of it have arrived. TenneT’s first national Flexibility Needs Assessment calculates that North Holland will need 401 gigawatt-hours of upward transmission-network flexibility in 2035, against 307 GWh in 2030. That is a 30% climb in a modeled requirement, not a count of batteries or a meter reading from data centres already operating.

The pinch point is Haarlemmermeer. The report groups projected data-centre growth with household demand there. In North Holland, the remaining demand overloads are concentrated in one area and increase in the 2035 case. TenneT says the assumed data-centre development could be higher than actual growth, which would also make the modeled 2035 network need too high. The report gives no data-centre capacity, connection total or project-by-project schedule behind that assumption.

That caveat matters because the same report’s headline system result is easy to misread. Under its VVT scenario, the national system has no structural uncovered flexibility need in 2030 or 2035 after assumed resources are counted. The network calculation asks a different question: can power get through particular constrained lines and substations? On that map, large needs remain, especially on TenneT’s high-voltage grid. A national balance can look comfortable while one province still needs more upward flexibility at the right place and hour.

The distinction is commercial, not academic. A location-specific flexibility need can shape grid-service procurement, storage and demand-response business cases, connection plans and data-centre financing. But a forecast built on unmeasured growth can send those signals to the wrong scale. Developers may read a modeled need as a reason to reserve equipment or sites; network planners may treat a scenario as a pipeline; investors may price a service market that has not yet been demonstrated.

ACM’s approval decision makes the uncertainty part of the news. The regulator approved a revised report on 12 October, after TenneT resubmitted it on 11 September. ACM said the July version did not make clear enough that assumed flexibility growth was uncertain. For the next assessment, the regulator wants multiple scenarios, including one that meets the 2030 climate goal, and a realistic projection of the flexibility expected to be available. It also wants consistent weather inputs, better coverage of networks below 25 kV as metering improves, and a model of how system-wide and local network needs interact.

The first assessment therefore produces two useful outputs: a forecast of where flexibility could be valuable, and a list of reasons to test the forecast harder. For data-centre investors in North Holland, the watchpoint is not only whether new load arrives. It is whether the next grid assessment separates projects with connection evidence from growth that exists only in the scenario.

Sources