The real asset isn’t the building
The building takes one to two years. In Europe’s biggest hubs, the grid connection it needs takes seven to ten.
Those aren’t my numbers. They come from the International Energy Agency, in a commentary on Europe’s data centre goals published in November 2025. Put the two figures side by side and you have most of the story of where energy-intensive infrastructure gets built in Europe right now. The building is the easy part. The power decides.
I’m using data centres as the example because that’s where the public data is best. The logic holds for anything that draws power around the clock. I learned it in Bitcoin mining, where the lesson is immediate: a mine is basically an electricity contract with machines plugged into it. If the power isn’t there, or isn’t there at a price that works, nothing else on the site matters.
The queue is the bottleneck
The IEA’s numbers are worth reading slowly. Across the EU, securing a grid connection takes anywhere from two to ten years depending on the country. In the five historic hubs (Frankfurt, London, Amsterdam, Paris and Dublin) developers face connection queues that average seven to ten years. Dublin and Amsterdam have both had to pause new projects, citing a lack of grid availability.
It isn’t only slow, it’s expensive. The IEA quotes ACER, the EU agency that coordinates national energy regulators, which put direct grid congestion costs at EUR 4.3 billion in 2024. That figure doesn’t include what delayed projects cost the wider economy. The IEA’s own conclusion is plain: expensive delays in core markets are prompting investors to look toward regions with more available grid capacity.
The market has already moved
You can see it in the delivery figures. CBRE’s Europe Data Centres Figures for Q2 2026 count 685 MW of new colocation and hyperscale self-build supply delivered across Europe in the quarter. 63% of it landed outside those five hubs. Paris delivered nothing at all that quarter. Demand moved the same way: CBRE puts markets outside the big five at 55% of Continental European take-up in Q2.
CBRE’s explanation fits in one line, and I think it’s the right one. The growth reflects a continued shift toward parts of the continent “that have an abundance of available power.”
Nobody picked those secondary markets for their buildings. You can put up a shell almost anywhere in a year or two. You can’t print a grid connection, and you can’t buy your way up a queue that can run close to a decade.
What actually carries the value
So when I look at a project, I read it from the power side first. My opening questions rarely change:
- How many megawatts are contracted, and from what date?
- Is the connection firm, or can the grid operator curtail it?
- What does the power cost, and who carries the price risk?
- Who else is waiting on the same substation?
Flexibility belongs in that conversation, and it isn’t a weakness. The IEA itself points to non-firm connections, where the user accepts potential curtailment under specific conditions, as a way to connect loads faster on congested grids. A buyer who can live with those terms may skip years of waiting.
That’s also why I read a site from the connection out. A live connection next to industrial land is the scarce thing. So is a queue position, even if it often gets treated like paperwork. Whatever building eventually goes on top is the commodity.
The hubs aren’t finished
None of this makes the big hubs irrelevant. Most of Europe’s installed capacity still sits there, and the IEA notes that most announced capacity is still planned for existing hubs, which only adds pressure on grids that are already full. I don’t see that concentration unwinding quickly.
But new capacity follows the power, and the Q2 delivery numbers show that’s already happening. It isn’t a forecast anymore.
If you’re building anything energy-intensive in Europe, read the grid connection agreement before you read the floor plan. One of them can be redrawn next month. The other can take a decade to get.