Data centres are the backbone of the digital economy, supporting cloud computing, AI, and the rapid expansion of data-driven industries. They are also energy-effective facilities, with AI server sizes now running at 10-20x power levels of speed servers & continuing to require round-the-clock power to maintain operations. With electricity making up a significant portion of running costs & sustainability taking centre stage in corporate strategies, a key challenge emerges: how can data centres ensure reliability while integrating renewable energy sources that are inherently intermittent?
The answer is in a strategic combination of energy storage, backup power, and smart grid participation. By adopting innovative solutions, data centres can not only improve their uptime and energy security but also play an active role in supporting grid stability and the broader transition to a low-carbon energy system.
How data centre computational power is evolving
Data centres are requiring more power to sustain operations as advances in computation power mainly rise with AI. The graphic below shows a comparison of how data centre processing power has evolved in the last decade and compares this to the processing power of the human body, our conscious mind, subconscious mind and sensory systems.
Data centres of the 2010s vintage depend on Central Processing Unit (CPU) based servers, which handle tasks sequentially. 2020s AI data centres now depend on Graphics Processing Unit (GPU) based servers, which use parallel processing for graphics and complex calculations simultaneously. This is ideal for video editing, machine learning, and gaming applications. The computational power of an individual human system is mainly registered compared to the 2010s and 2020s AI data centres.
Overcoming renewable energy intermittency
Addressing the challenge of reliability requires a systems thinking approach to energy solutions for this highly advanced computational infrastructure. Battery storage is an important solution to bridge the gap between changing usable generation & continuous data centre demand. By storing excess usable energy during peak production periods and discharging it during low-generation periods, battery storage helps ensure a consistent, high-quality power in the short term up to the duration of the battery. SLR is seeing longer durations of battery storage coming to the market, for example, in recent Capacity Market rounds. Ongoing advances in longer-duration storage technologies will have wider application for data centres in the future.
On-site renewable energy and microgrids are being used with on-site solar and wind power to supplement grid electricity. As microgrids, these facilities can operate independently or semi-independently of the power grid, balancing renewable generation with storage and backup systems. Offsite renewables can also be accessed by data centres through long term power purchase agreements providing longer term price security and reducing reliance on fossil fuel power.
Hydrogen and alternative fuels such as green hydrogen, produced using surplus renewable electricity, are a longer-duration energy storage solution. When stored & used in fuel cells, it gives a long-duration backup power source that complements battery storage over a longer time. Similar fuels such as biofuels & renewable natural gas also give lower-carbon backup options compared to traditional diesel & gas engine generators.
Artificial intelligence & predictive analytics play an important role in managing energy supply & demand. AI can forecast demand, predict renewable energy availability, & use battery charging & discharging schedules. These algorithms are also often used by energy collectors & electricity traders. This dynamic energy management ensures maximum efficiency.
Backup storage and grid ancillary services
Energy storage and backup power assets can also participate in grid-supporting services. These services markets typically enable the grid network operator to help balance the grid in terms of energy supply & power quality by paying market participants for these services. By participating in these markets, the data centre owner can earn more diverse revenue streams and maximise the utilisation of the assets during their economic life. By giving flexibility to the grid, data centres can support huge energy system stability in a lower-carbon-intensity grid. Some of the main markets with frequency regulation & demand response, where battery systems can give frequency regulation services by injecting or absorbing power to maintain grid stability.
Demand response programmes allow data centres to adjust power use in response to grid conditions, minimizing demand during peak demand periods. In certain markets, this allows the data centre to minimize its electricity bills. Peak shaving & load shifting, examples of demand response, are delivered by discharging energy during peak demand, which can help data centres minimize their reliance on expensive grid electricity & ease grid congestion. This is a characteristic SLR sees in many of its markets, which helps the battery economics and maximizes the use of renewable energy.
A virtual power plant (VPP) may be formed from more than one data centre, with significant energy storage capacity. This has the benefit of transforming data centres from being an otherwise passive demand to an active grid participant; using the backup storage capacity to help grid operators balance the grid as we make the transition to net zero.
Conclusion
As the digital economy uses data centres are becoming both important network & significant energy users. The increasing computational demands driven by AI & advanced applications require round-the-clock, high-quality power. Meeting these demands sustainably is no longer optional. By integrating battery storage, on-site renewables, smart grid capabilities, & using technologies like hydrogen fuel cells, data centres can improve energy use while contributing to a cleaner, more flexible energy system. Far from being passive consumers, data centres are poised to become active partners in the energy transition, supporting grid stability, minimizing carbon use, & enabling a net-zero future.
