Artificial intelligence is creating a new kind of energy challenge. The rapid expansion of AI computing is driving the development of increasingly large, power-intensive data centres, placing significant pressure on electricity infrastructure.
The scale of the change is substantial. According to the International Energy Agency’s (IEA) 2026 analysis, global electricity consumption from data centres is expected to approximately double, from around 485 TWh in 2025 to 950 TWh by 2030. Electricity consumption from AI-focused data centres is projected to grow even faster, roughly tripling over the same period.
Goldman Sachs Research also forecasts global data centre power demand increasing by around 50% to 92 GW by 2027.
The question is therefore shifting from “Where should we build new data centres?” to “How can we provide sufficient, reliable power quickly enough?”
The challenge is not simply generating more electricity
Two interconnected challenges are emerging:
1. Generating sufficient power
AI infrastructure requires very large quantities of electricity, delivered reliably and often continuously.
2. Getting that power to where it is needed
Even where generating capacity exists, data centres may face long waits for grid connections, transmission upgrades, transformers and other infrastructure.
This second challenge is particularly important. A new data centre can potentially be developed in two or three years, while major transmission lines in advanced economies can require four to eight years to deliver. Lead times for critical grid components have also increased.
- 01AI growth
- 02More computing capacity
- 03Higher electricity demand
- 04Grid connection bottlenecks
- 05Demand for local generation, storage and smarter energy management
This mismatch in development timelines is encouraging data centre operators, utilities and technology companies to consider a much wider range of energy solutions.
From grid connection to the “power campus”
One important trend is the move towards more localised electricity generation.
Instead of relying entirely on electricity transported through the conventional grid, data centre campuses could increasingly combine grid connections with dedicated generation, energy storage and intelligent power management.
This does not mean abandoning the grid. Rather, the emerging model is a hybrid energy system in which multiple technologies work together to provide capacity, resilience and flexibility.
The IEA expects renewables to supply nearly half of the additional electricity required by data centres through 2030, while natural gas, nuclear and other dispatchable sources also play important roles.
Several technologies are attracting particular attention:
Onsite gas turbines
- Potential role for data centres
- Dispatchable electricity generation that could reduce dependence on delayed grid upgrades
- Key innovation opportunities
- Higher efficiency, lower emissions, hydrogen-ready systems, improved load-following
Small Modular Reactors (SMRs)
- Potential role for data centres
- Reliable, high-capacity electricity with low operational carbon emissions
- Key innovation opportunities
- Smaller reactors, modular manufacturing, simplified construction and integration
Large-scale batteries
- Potential role for data centres
- Short-term backup, peak management and grid flexibility
- Key innovation opportunities
- Lower-cost chemistries, longer duration, improved safety and power management
Renewables + storage
- Potential role for data centres
- Low-carbon electricity supported by energy storage
- Key innovation opportunities
- Hybrid systems, forecasting, storage optimisation and microgrids
Advanced geothermal
- Potential role for data centres
- Potential source of continuous low-carbon electricity
- Key innovation opportunities
- Enhanced geothermal systems, deeper drilling and closed-loop technologies
Grid-enhancing technologies
- Potential role for data centres
- Increase the usable capacity of existing transmission infrastructure
- Key innovation opportunities
- Dynamic line ratings, power-flow control, advanced conductors and AI-enabled grid management
Three technologies to watch
1. Onsite gas generation: Speed and reliability
Grid constraints are encouraging developers, particularly in the United States, to explore onsite natural gas generation. The attraction is clear: dedicated generation could allow some projects to progress without waiting for major grid reinforcement.
However, technical challenges remain. AI data centres can experience large and rapid changes in electrical load, requiring onsite generating equipment to respond dynamically while maintaining reliability and efficiency.
The IEA’s 2026 analysis also highlights significant design, supply chain and economic challenges. Many proposed onsite gas projects remain at an early stage, while shortages of gas turbines create further uncertainty around deployment times.
This creates innovation opportunities in turbine design, control systems, hybridisation with batteries, heat recovery and lower-carbon fuels.
2. Small Modular Reactors: Bringing nuclear closer to demand
Nuclear power has characteristics that align well with data centre requirements: high availability, large quantities of reliable electricity and low operational carbon emissions.
Small Modular Reactors (SMRs) could potentially offer another advantage: deployment in smaller, modular increments closer to major electricity users.
Interest in nuclear power from the technology sector is growing, with data centre developers exploring long-term electricity supply agreements and investment in advanced nuclear technologies. The IEA expects the first SMRs to enter operation around 2030, although deployment timelines remain uncertain.
The challenge now is accelerating commercialisation, reducing construction and regulatory risks, and demonstrating attractive economics.
3. Batteries: From backup power to active energy management
Batteries are evolving from a supporting technology into a strategic part of data centre energy infrastructure.
Large battery systems can absorb excess electricity, manage peaks, smooth the output of onsite generation and renewables, provide rapid-response power and potentially allow data centres to interact more flexibly with the grid.
Rapid advances in lithium-ion manufacturing, scale and cost have helped make applications such as grid storage and data centres more viable.
According to the IEA, the global lithium-ion battery market exceeded USD 150 billion in 2025, with batteries becoming strategically important across power grids, data centres, transport, robotics and other industries.
The IEA also estimates that around 20–25 GW of battery storage could be installed in data centres globally by 2030, supporting not only reliability but potentially greater flexibility in how data centres interact with electricity networks.
The next innovation challenge: Integrating everything
The most effective solution is unlikely to be a single technology.
Instead, the future AI data centre may operate more like a miniature energy system, combining:
-
Grid electricity -
Renewable generation -
Dispatchable onsite power -
Battery storage -
Intelligent energy management -
Flexible computing loads -
Potentially nuclear or advanced geothermal power
Software will be as important as hardware. AI itself could help optimise when electricity is generated, stored, purchased and consumed, dynamically balancing computing requirements against electricity availability, price and grid constraints.
The challenge is integrating these different technologies into a reliable, efficient and economically viable system. This requires coordination across electricity generation, power electronics, storage, control software and grid infrastructure.
The result could be a fundamental change in how large electricity users approach energy: from simply consuming electricity to actively generating, storing and managing it.
Where NineSigma can help
NineSigma works across power generation technologies, energy storage, grid technologies, efficiency and emerging approaches to industrial energy supply.
Through its global innovation network and technology scouting capabilities, NineSigma helps organisations identify emerging technologies, innovative suppliers, startups, research organisations and partners to address complex technical challenges, including solutions originating outside traditional energy and data centre industries.
As AI infrastructure places new demands on electricity systems, identifying the right combination of technologies and partners will be essential to developing reliable, scalable energy solutions.
Are you exploring how to supply, store or manage the next generation of high-demand electricity loads?
Get in touch with NineSigma to explore how technology scouting and external innovation partnerships could support your next project.
Nicolas Wojnarowski
Director, Innovation &
Business Development
WESTERN EUROPE · AMERICAS · AUSTRALIA · MIDDLE EAST
20 years in innovation and business transformation across energy, healthcare and high-tech.
Yiqi Zhang
Innovation & Strategic
Business Development Manager
NORTHERN, CENTRAL & EASTERN EUROPE · UK · ASIA PACIFIC
Drives cross-industry innovation across Western and Asian markets for high-impact initiatives.
Other topics :
