Data centers
Does your data center really need a new grid connection for its next phase of growth?
Trend piece
August 03, 2026
5 min read
“We need another 40 MW” is becoming one of the most common conversations in the data center industry.
Artificial intelligence is accelerating demand for computing power. Existing facilities are expanding. New campuses are being planned. Customers want more compute capacity. And that means one thing: More power.
The challenge? The grid isn’t expanding at the same pace.
Across Europe, transmission system operators (TSOs) are receiving connection requests that far exceed the capacity currently available. As a result, connection queues are growing, while project prioritization is changing, and waiting years for additional grid capacity is becoming an increasingly common reality for large electricity users, including data centers.
In other words, access to the grid has become one of the biggest constraints to data center growth.
In this article, we look at the engineering principles behind assessing available grid capacity and explain why understanding the existing connection should come before deciding on the next investment.
Why a new data center grid connection isn’t always the first answer
Imagine a data center’s next AI expansion requires another 40 MW. The immediate response is often straightforward: “We need a new grid connection.” After all, if additional capacity is not available today, what other option is there?
For many projects, that’s where the conversation ends. But that conclusion is usually based on a small number of assumptions that do not match how the electricity system is actually planned and operated.
1.
The first assumption is that a grid connection is a fixed number, treating, e.g., 100 MW as an automatic, universal limit in all circumstances. In reality, planning standards define different loading limits for transformers and circuits under normal, post‑outage, and short‑term conditions, so practical capacity depends on time, operating scenario, and load profile.
2.
The second is that the grid connection ends at the customer’s fence. From a network planning standpoint, the connection is a chain that runs from the customer’s own switchgear, through the dedicated connection assets, back to upstream station busbars and the wider network. Any element along that chain can be the first to hit a limit.
3.
Third, and perhaps the biggest misconception, is assuming that the wider electricity network is automatically the limiting factor. In practice, the first constraint can appear inside the customer installation, in the connection assets, or in the upstream network. Locating that constraint is one of the first steps in a proper engineering assessment.
When organizations assess an existing grid connection, they naturally start with the infrastructure they own: transformers, substations, and other on-site electrical assets. But a grid connection is more than what sits inside the customer’s fence. Its practical capacity depends on how customer-owned assets, connection assets, and the wider electricity network interact under real operating conditions.
Identifying where the first constraint lies is therefore one of the first steps in an engineering assessment.
Practical example
At NovoGrid, part of Opoura, we recently worked with a data center owner who believed its next phase of growth would require a completely new grid connection.
Rather than assessing only the customer’s installation, we evaluated the entire connection – from the customer’s own infrastructure to the upstream network – to understand what was actually limiting additional capacity.
That broader engineering assessment identified the potential for approximately 40 MW of additional capacity at this British data center. For some projects, identifying additional capacity from an existing connection can help avoid years of delay, reduce infrastructure costs, and accelerate expansion plans.
Every project is different, and every opportunity must be validated with the network operator. But the project highlighted an important engineering principle: before deciding on the solution, identify the first constraint. Depending on where it lies, the next step could be upgrading customer-owned infrastructure, modifying the existing connection, assessing broader network constraints, or determining that a new grid connection is required.
The question every data center owner should ask before requesting more grid connection capacity
As competition for grid capacity continues to intensify, organizations need to rethink the questions they ask.
Instead of asking, “How do we get more MW?”, the better question is: “Have we fully understood the connection we already have?”
That’s an important shift, because before investing in a new connection, joining another connection queue, or committing to major infrastructure upgrades, it is worth understanding what is actually constraining your existing connection.
Sometimes the constraint lies within the customer’s own installation – for example, transformer capacity or other on-site electrical assets that no longer match the capability of the upstream network. Other times, it is influenced by upstream network configuration, circuit ratings, operational constraints, or the security standards used to operate the network.
These are fundamentally different engineering challenges, leading to fundamentally different solutions. In some cases, upgrading customer-owned infrastructure may be enough. In others, modifications to the existing connection, wider network assessments, or ultimately a new grid connection may be required.
The key is that the engineering solution should follow an understanding of the constraint – not the other way around.
Better data center grid connection decisions start with a better understanding
As grid constraints continue to grow across Europe, the companies that make the best expansion decisions won’t necessarily be the ones with the largest grid connections. They’ll be the ones who understand in detail how their existing connections perform under real-world operating and security‑of‑supply conditions. Sometimes, that understanding will confirm that a new grid connection is required. Other times, it may reveal standards-compliant opportunities that weren’t visible at first.
What matters is that the engineering decision follows the assessment of the connection and the network it sits on, not an assumption based solely on nameplate ratings or apparent headroom.
