A company in the Netherlands applied for a grid connection in 2021.

It paid TenneT nearly half a million euros against written confirmations that the capacity was reserved, and it wanted power by 2029. Now it won't get connected until somewhere in the mid-2030s. They took TenneT to court, and lost their case.

The company is Goodman, a datacenter near Schiphol, asking for 70 megawatts. Roughly equivalent to the electricity consumption of 70,000 households. But it's worth being precise about why they lost, cause it's not the obvious reason.

The court didn't rule that the grid is full so tough luck. It ruled there was no signed connection contract. Under the Netcode, you only get a right to connection once a transport agreement is signed. A quote, a portfolio confirmation, two payments, years of contact.

None of it counted. Commitments confer no rights. And on top of that, grid stability outranks any single customer.

However, the headline reads like a grid problem. Underneath, it's a timescale problem.

1. Electrification moves on a five to ten year curve. Solar, EVs, heat pumps, industry switching off gas, all loading at once.
2. High-voltage infrastructure moves on a fifteen to twenty year curve. Stations and lines that take a decade-plus to plan, permit and build.

When the fast one outruns the slow one, the gap doesn't open cause someone planned badly. It opens cause the two clocks were never going to match.

And everything we reach for to manage it inside that gap buys time without making capacity. Some examples are:

• Flexibility markets.
• Batteries.
• Congestion management.
• Off-peak shifting.

All of them help run the grid smarter, as they are mitigating a full stop. But none of them build a transformer that actually solves it.

The visible problem is the connection stop. From July, large parts of Utrecht can't get new connections. The wider Flevoland-Gelderland-Utrecht region is close behind. More than 1,100 businesses are already in the queue, and now smaller companies and households join them.

This is the part people feel. A business that can't expand. A housing project that stalls. A waiting list with no clear date on it.

But the connection stop is the symptom, not the cause. The real problem sits one layer down, and it stayed hidden longer than it should have.

The congestion studies existed for years. They kept getting revised, each time worse than before. In February, TenneT recalculated the FGU region and found it worse again. Bad enough that a full-region connection stop went from something to avoid to something to expect.

The grid didn't suddenly break. The numbers finally caught up to a reality that had been building for years.

That's the distinction I keep landing on. This isn't a crisis in the sense of a sudden failure. It's drift. A system moving away from the conditions it needs to hold itself up, while the dashboard reads acceptable, until a recalculation forces everyone to see it at once.

None of this means nobody's acting. There's a national programme, action plans and accelerated permitting.

But the uncomfortable part is that the binding constraint is build time, and effort spent on the fast variables can't shorten the slow one. The only thing it can do is manage the symptoms while the gap runs its course.

Which raises the question I don't have a clean answer to. If the scarcity is locked in for a decade, is it more honest to govern the queue openly for ten years of it, than to keep framing each mitigation as a bridge to relief that the build timeline can't deliver?

Source: NOS