Summary
- Colliers’ H1 2026 EMEA data centre report says power availability, grid connection timelines, and planning are now primary delivery constraints.
- The report highlights pressure in core FLAPD markets and acute grid constraints in Ireland, Germany, and the UK.
- Developers are increasingly weighing self-generation, microgrids, and energy-linked development strategies as capacity demand outpaces deliverable supply.
Colliers says power availability, grid connection timelines, and planning constraints are now setting the pace of data centre development across EMEA.
The firm’s H1 2026 EMEA Data Centres Market Report describes a market still expanding quickly, driven by hyperscale cloud, AI workloads, and sovereign digital infrastructure requirements. Growth is becoming harder to convert into live capacity, with secured power, planning confidence, and credible execution routes now separating buildable projects from paper pipelines.
The report points to delivery constraints in the core Frankfurt, London, Amsterdam, Paris, and Dublin markets. It identifies acute grid constraints in Ireland, Germany, and the UK, while noting that operators are increasingly looking at self-generation, microgrids, and energy-linked development strategies.
Demand remains strong. The harder question is whether capacity can be physically powered, connected, approved, built, commissioned, and handed over inside customer timelines.
Demand is not the scarce commodity
AI has not replaced conventional cloud demand. It has layered high-density requirements on top of enterprise cloud, sovereign infrastructure, digital public services, and data localisation. That extra demand is arriving in markets where land and grid access were already tightening.
Developers may announce multi-hundred-megawatt campuses, but the route from optioned land to energised data hall runs through grid studies, substation delivery, transmission upgrades, environmental assessment, planning politics, MEP procurement, and customer fit-out. Any weak point in that chain can move a project from near-term capacity into a long-dated option.
Established hubs carry the pressure most visibly. Dublin has already seen the political consequences of data centre load growth. Frankfurt continues to manage constraints around land, power, and heat. London faces connection queues and competing development priorities. Amsterdam and Paris are balancing density, sustainability, and public acceptability.
Colliers’ report reflects a wider shift in investor diligence. A site with theoretical demand is worth less than one with contracted or realistically deliverable power. Connection timing, planning status, and grid certainty now affect land pricing, platform valuations, and the credibility of forward development pipelines.
Self-generation changes the risk
The move towards self-generation and microgrids is not only a technical trend. It is evidence that large digital loads are outpacing the ability of public networks to connect them quickly in some markets.
Developers are assessing gas engines, fuel cells, batteries, private-wire renewables, or hybrid systems as bridges to grid capacity. Those approaches can unlock projects, but they also change the risk profile. Gas-backed generation raises emissions and permitting questions. Private-wire renewables still need land, balancing, and connection arrangements. Batteries help with flexibility and resilience, but do not create firm energy supply on their own.
Energy-linked development strategies will therefore become more complex. Developers that can combine land, grid access, generation, heat reuse, and flexible demand may have an advantage, although those projects will face deeper scrutiny from utilities, regulators, and local authorities.
Power infrastructure is also moving deeper into the capital plan. Substations, transmission works, backup systems, procurement contracts, and on-site energy systems can no longer be treated as peripheral utility matters. They affect project phasing, customer commitments, financing assumptions, and exit values.
Secondary and emerging hubs may gain ground where they can offer land, capacity, and planning alignment, but not every power-rich location becomes a data centre market. Fibre, latency, labour, water, political support, customer demand, and resilience requirements still have to align.
The European market is entering a more selective phase. Capital is available and demand is rising, but developers will be judged on delivery credibility rather than pipeline scale alone.
The next wave of EMEA data centres will be decided at the grid interface as much as in the leasing market. Demand may fill the models. Power will decide which projects leave the ground.
