Executive Summary
- The growth of Europe’s data centre market continues to accelerate at pace, however, operators are increasingly encountering a different challenge to their North American counterparts.
- Predominantly, these differences present themselves in how operators secure resilient power where grid capacity is constrained and connection timelines can extend for years.
- Here, Emmet Taggart, Director for Data Center Markets Europe, from Cummins Power Generation, discusses differing infrastructure across continents, and the practicalities of differing backup technologies.
Power has become one of the defining factors in whether new data centre developments progress, with increased power demand driven by rapid growth of AI Compute. With the IEA (International Energy Agency) projecting DC (Direct Current) power to more than double to ~945 TWh by 2030, operators are having to consider energy infrastructure much earlier in the planning process than ever before.
In this context, North America provides a valuable point of comparison. While Europe’s regulatory environment and electricity networks differ, the operational lessons from these mature markets are increasingly relevant as European capacity expands.
Europe’s constraints are changing development decisions
Grid constraints in some European markets are increasingly shaping data centre power decisions much earlier than in previous development cycles. In some locations, grid connection timelines can exceed three to seven years, creating a significant challenge for operators looking to meet growing demand.
Site selection increasingly depends not only on land availability, but also on grid connection times, future network reinforcement plans and the long-term ability to expand capacity.
At the same time, increasing rack densities also present power challenges. Traditional air cooling systems can become less effective at higher densities, driving a shift towards liquid cooling and adding another consideration for operators alongside connectivity.
With AI racks scaling from around 120 kW today towards 600 kW by 2027 and potentially reaching megawatt-class levels before 2030, power and cooling can no longer be planned by a single asset, requiring operators to consider them as part of an integrated infrastructure strategy. This in itself puts additional strain on the power requirements for new data centres and existing sites with plans for growth.
Regulation is changing operational priorities
Alongside grid availability and power demands, regulation is reshaping how facilities are designed and managed.
North America has a much more fragmented approach, whilst requirements across the Middle East and APAC vary significantly by market and often rely more on national efficiency standards, permitting and voluntary ESG disclosure.
In comparison, European reporting requirements, which apply to facilities with an installed IT power demand of 500 kW or higher (with France and Germany applying lower thresholds), are placing greater emphasis on whole-site performance. This legislation is prompting operators to consider how cooling systems, standby generation, batteries and utility imports interact rather than assessing each asset independently.
European data centres are also progressively being recognised as nationally significant infrastructure, bringing greater expectations around resilience, transparency and environmental performance.
For operators, the challenge is designing power systems that support business continuity while adapting to changing regulatory expectations over the lifetime of a facility.
So what role is power generation and surrounding technologies playing in this landscape?
Onsite power as part of a wider power strategy
Historically, standby generators existed for one purpose; maintaining operations during an unexpected utility outage. Today’s environment is different.
Across North America, operators are increasingly designing facilities to accommodate delayed utility connections, constrained grid capacity, phased campus expansion and rapidly changing workloads. Rather than treating onsite generation as an insurance policy against outages, operators are more often considering a broader mix of onsite power technologies and how they can support power availability, resilience and growth.
In some developments, onsite generation enables facilities to begin operating while permanent grid capacity is completed, referred to as a bridge-to-grid approach. Depending on the requirements of the site, this can involve generator sets alongside technologies such as battery energy storage systems (BESS), microgrids, natural gas and alternative fuel solutions, creating a more flexible power system where grid reinforcement cannot keep pace with construction schedules or future demand growth.
Alongside generator sets, Battery Energy Storage Systems (BESS) can work in tandem to store and apply power in locations where fluctuations could impact the operation of sites. Using energy management systems (EMS), this technology monitors load profiles, grid conditions, electricity prices and, in some cases, renewable generation forecasts. This has clear benefits from a standby power perspective, but can support data centre operators manage spikes in demand driven by AI workloads.
Having this level of power available, where grid capacity cannot provide extra uplift, gives operators the option to turn to more off-grid power models. These systems combine generation, energy storage and controls to provide a reliable power source where supply is not available.
This is an area where European operators can draw practical lessons from the US experience and lead with innovative practises. Grid constraints across Europe mean power infrastructure needs to be considered as part of a broader resilience strategy, pushing European operators to take a multi path approach to powering sites.
Integration matters more than individual technologies
As power requirements become larger and more complex, the critical question is increasingly not which individual technology to deploy, but how the overall power architecture works together.
Generator sets, battery energy storage, and renewable energy assets all have a role to play. The challenge is ensuring those systems perform predictably and reliably together under real operating conditions.
The same engineering approach applies to fuel selection.
Diesel continues to provide the high power density and transient response expected for critical standby applications, making it an important layer within the wider power architecture. A resilient power strategy could combine gas and BESS for prime power, grid connectivity where available, and diesel as a fuel-secure backup independent of grid and gas supply, with Hydrotreated Vegetable Oil (HVO) capabilities helping to reduce the carbon impact of this critical resilience layer.
HVO can be used in compatible equipment with minimal tweaks, offering a potential lower lifecycle carbon alternative in many installations. Whilst fuel choice remains a site-specific engineering decision that depends on local availability, resilience objectives, operating strategy and regulatory requirements, the potential for HVO in Europe is exciting. Here is a fuel that can be produced at scale in Europe, for use in Europe, that can deliver diesel-like performance, greenhouse gas emissions reductions of up to 50% compared to conventional diesel with a near-complete reduction in CO2 well-to-wheel emissions.
Learning from global markets
Europe and North America will continue to develop differently because of their respective grid systems, planning frameworks and regulatory environments are fundamentally different. The same is true across other evolving global data centre markets, where grid availability, infrastructure capability, energy policy and access to different power sources are shaping how operators approach growth.
With global data centre electricity consumption expected to more than double from approximately 415 TWh in 2024 to around 945 TWh by 2030, these challenges are not just faced by particular regions in isolation.
As Europe’s data centre market evolves, operators that treat onsite generation, battery storage, controls and utility connectivity as a single, coordinated power architecture will be better positioned to deliver capacity without compromising resilience.
