This AVK and Wärtsilä Energy white paper examines constrained grid capacity as a major barrier to the expansion of European data centres, particularly facilities serving AI workloads. It proposes on-site microgrids combining dispatchable generation, BESS and renewables to accelerate site energisation before—or independently of—a grid connection. The paper compares power-system scenarios and four supply configurations for a representative AI data centre in terms of cost, reliability, land use and carbon footprint.
Key Takeaways
Limited grid capacity and lengthy connection timelines are becoming a direct constraint on AI data-centre deployment in European hubs; a microgrid can provide either interim or permanent power for a site.
A renewables-and-BESS-only model for continuously supplying an AI data centre requires substantial overbuild of generation and storage, with high land requirements and increased embodied emissions.
The authors consider a balanced architecture combining renewables, storage and dispatchable balancing generation to be more practical for reliable off-grid power supply to AI data centres.
In the modelling, modular medium-speed engines adapt more effectively to changing load than large CCGT plants, as they require less BESS capacity and maintenance reserve.
Once connected to the transmission grid, a data centre’s power capacity can become a system resource, supporting balancing, reserve provision and the integration of additional renewable generation.
Key Figures
Growth in data-centre capacity demand by 2030
approximately 25 GW
McKinsey estimate; current capacity is stated as the baseline.
Current data-centre capacity
10 GW
Baseline capacity in the McKinsey estimate.
Data-centre electricity consumption in 2024
62 TWh
Data-centre electricity use stated in the document.
Data-centre electricity consumption by 2030
over 150 TWh
Forecast data-centre electricity demand.
European data-centre pipeline
around 170 GW
Goldman Sachs estimate; compared with roughly one-third of current European electricity consumption.
AI data centres expected to face operational constraints by 2027
40 %
Gartner forecast; the constraint is driven by power availability.
Reference AI data centre in the modelling
80 MW
Base case used to compare supply configurations, LCOE, land use and embodied emissions.
Land requirement for renewables plus BESS for the reference AI data centre
10,087 hectares
Model 1; described as exceeding one hundred square kilometres.
Land requirement for ICE plus BESS for the reference AI data centre
1.6 hectares
Model 2.
Embodied emissions in Model 1
180 gCO₂eq/kWh
Renewables plus BESS; elevated by substantial overbuild across extraction, manufacturing, transport and construction.
Typical embodied emissions from renewable energy
around 10 gCO₂eq/kWh
For normal deployment without excessive overbuild.
BESS service life
12 years
The document notes the need to replace batteries relative to other power systems.
Typical lifecycle of other power-supply systems
25 years
Comparison point for BESS service life.
European data-centre market growth
12 → 23 GW
Forecast aggregate European market capacity over the stated period; conservative growth rate is also cited.
Investment in European data centres through 2030
€110 billion
Expected investment volume.
Expected European market revenue
€208 billion
Estimate for the same period through 2030.
Germany colocation capacity by 2030
3,572 MW
Forecast colocation capacity; the highest figure among the European markets cited.
Data centres’ share of metered electricity consumption in Ireland in 2023
21 %
Cited to illustrate the burden on the power system.
Grid connection applications around London
up to 400 GW
Cited as accumulated connection requests.
UK target AI cluster by 2030
at least 500 MW
UK government intention cited in the document.
Practical Value for Data Center Owners
For data-centre owners and developers, this paper supports early-stage site assessment where grid connections are constrained: the comparison should extend beyond connection tariffs and CAPEX to include an islanded or hybrid microgrid incorporating renewables, BESS and dispatchable generation. It identifies a specific design risk: a renewables-and-BESS-only configuration for a continuous AI data-centre load may prove unacceptable in terms of land requirements, reserve provision, battery replacement cycles and embodied emissions. It also outlines a procurement and ownership model in which a specialist partner builds and operates the power capacity, which can be monetised through grid-balancing services after grid connection is secured.
The paper introduces FissionReady, a scheduler for data centers supplied by a multi-module small modular reactor (SMR) plant. It jointly manages the output of each reactor module, deferrable batch workloads, and grid purchases while accounting for fuel age, iodine-xenon dynamics, and reactivity margin. In simulations, the approach eliminates forced outages and missed batch-job deadlines while reducing water use and grid-electricity costs relative to a comparable baseline configuration.
Small Modular Reactors (SMRs)Data Center Power SupplyGeneration Load FollowingData Center Microgrids
This research paper presents a method for quantifying the electricity-demand flexibility that AI data centers can provide to the power system. The approach reschedules batch computing workloads subject to constraints on delay, uninterrupted execution, and available CPU, GPU, and memory resources. Task-scheduling decisions are translated into a server power profile, which is then used to assess peak-shaving potential and the duration for which a load reduction can be sustained, using real GPU-cluster traces.
AI Data CentersDemand FlexibilityLoad ManagementBatch Workload Shifting
Wärtsilä’s presentation examines the use of modular gas and dual-fuel power plants alongside energy storage to supply data centers, including AI workloads. It links constraints on grid connections to growing demand for primary onsite and off-grid generation, and cites several projects in Ireland and the United States. The material emphasizes capacity availability, modularity, part-load efficiency, lower water consumption, and the potential to transition to sustainable fuels over time.
Data center power supplyAI workloadsOnsite generationOff-grid data centers