TJ Surbella, Strategic Planning Director, and Ron Beck, Senior Director, AspenTech, discuss how self-sufficient energy systems help enhance operational efficiency, sustainability and resilience, offering data centres greater control over their energy use while addressing growing electricity needs and environmental goals.
Data from Stocklytics projects the global data centre market will grow by 30%, reaching a value of over US$430 billion by 2028 – but this positive projection still begs the question: where will the electricity required to run these data centres come from?
The US Energy Information Administration (EIA) projects that global energy consumption will increase through 2050, which is driven by energy-intensive sectors like data centres. The demand created by AI scaling will only increase. Today there are hundreds of millions of daily queries on ChatGPT. This many queries can cost around 1GWh each day – the equivalent of the daily energy consumption for about 33,000 US households.
Analysis by the Center for Advanced Manufacturing at Clemson University, projects that global electricity demand will be increasing by 10-12% annually while the growth rate in global electric energy generation is only 2% per year. This puts a squeeze on the rate of the addition of new data centres.
Additionally, companies need to meet their sustainability targets leading them to explore the usage of renewable energy resources (such as solar, wind and geothermal) and low-carbon electricity to satisfy their electricity demand. Addressing this dilemma requires a transformative approach, one that ensures resilience and sustainability. Microgrids are emerging as a pivotal solution available today to provide data centres with reliable and lower-carbon power that complements and augments regional power grids.
Moving away from a centralised grid solution
Today’s power grids are already vulnerable to disruptions stemming from the increasingly complex and dynamic nature of renewable generation and grid storage, climate events, ageing infrastructure, cyberthreats and geopolitical tensions. These factors challenge the traditional uni-directional generation and distribution grids to reliably adapt to support data centres’ rapidly expanding needs.
This strain is particularly evident across parts of Europe, where grid stability risks deteriorating under intensifying demand and increasing the proportion of dynamic wind and solar generation needing to be transmitted at a distance to population centres.
For data centre operators, dependence on centralised grids has now become risky and costly. In some areas, utilities are unable to guarantee meeting the load demand of new data centres leading to their delay or cancellation altogether. Data centre operators, who see big opportunity but face operational risk from even occasional outages and who face rising energy costs across Europe, are exploring alternative options to meet their energy needs with an increasing sense of urgency.
Microgrids are self-sufficient energy systems that can operate alongside or autonomously of the main grid. They enable data centres to generate and manage their power locally concurrently with power supplied by the grid, increase electricity availability as well as contribute to the resilience of the adjoining power grid.
By reducing reliance on regional grids, microgrids provide data centres with greater control over energy use, supporting operational resilience, cybersecurity, the ability to address decarbonisation requirements and economic benefit. During grid outages, an effective microgrid can anticipate and react in real-time to the event and ‘island’ itself, enabling the data centre to continue operating without interruption. Therefore, large data centre developers are turning to low-carbon captive generation, renewables and microgrids as a kind of package solution to achieve faster permitting and approval of data centre projects.
Cue a strategic solution
For data centres, the benefits of microgrids are clear. They include accelerated schedules for bringing new facilities online, ensuring uninterrupted power for AI and business-critical computations, optimising energy utilisation and bolstering resilience, sustainability and economics. By generating power on-site, data centres reduce dependence on the often-increasingly complicated dynamics of grid-supplied electricity, minimise costs associated with peak-demand pricing and bid excess electricity back into the grid.
When exploring the installation of their own generation assets, customers typically focus on electricity price, reliability and carbon footprint – in that order. On-site conventional and renewable energy sources as well as Power Purchase Agreements (PPAs) are often the trigger of the desire to address this with a microgrid.
Advanced microgrids increase the efficiency of renewable energy integration, enable the tracking of green power content and align operations with sustainability goals. Renewable energy sources are intermittent and weather-dependent, which creates challenges, particularly when utility power is unavailable for extended periods. To ensure reliable power, facilities need to incorporate lower-emission generators, such as those powered by natural gas, green hydrogen fuel cells, or battery storage as a backup source to meet demand.
Microgrid customers will find that they face the kind of challenges traditionally managed by utilities: balancing electrical generation with load demand. On the generation side, this might include renewables, battery storage, natural gas-driven generators, fuel cells, or even small modular reactors. To address load demand, operational flexibility becomes essential. This can involve modulating non-essential processes, delaying certain computational tasks, or redistributing workloads to optimise energy efficiency. These measures ensure critical systems remain powered without interruption while maximising the use of available renewable energy.
How does a customer accomplish managing and optimising their microgrid? A supervisory control and data acquisition (SCADA) system can help the customer with this challenge. These are the same types of systems used by utilities today to help manage the challenge of balancing generation and load across an entire electric grid. These systems can provide the same benefits to microgrid users, allowing them to optimise their generation usage based on the lowest cost of electricity or emissions while ensuring power reliability 24/7 to all of their loads.
AI, in turn, becomes crucial in enhancing the capability of microgrids. Advanced systems powered by Machine Learning and predictive analytics are particularly effective in forecasting demand patterns (for example the periodic ebbs and flows of data centre computation) and renewable energy availability and pricing. Artificial intelligence-driven systems can analyse weather patterns to predict the amount of renewable energy that will be generated over the next several hours or days and estimate power loads. This enables data centres to plan energy usage more effectively, whether optimising for cost or emissions.
Real-time monitoring and prediction of energy flows ensures that excess energy generated during low-demand periods can be stored in battery systems or redirected to other operational needs within the facility, such as powering cooling systems or charging on-site electric vehicles. When feasible, excess energy can be sold back to the main grid or shared with nearby facilities, depending on local regulations. This then can be optimised to achieve the highest reliability, lowest cost and lowest carbon footprint – an approach which supports energy efficiency while enabling data centres to contribute to grid stability by managing peak load requirements.
This independence from traditional grid limitations empowers data centres to operate with improved flexibility and optimisation. As the demand for uninterrupted digital services continues to grow, microgrids provide a proactive and sustainable solution for the evolving data centre landscape.
Preparing for a resilient future
The benefits of microgrids extend beyond immediate operational gains. As power grids worldwide face mounting pressures from climate change and increasing energy demands, some geographies are embracing microgrids for commercial customers and localities as an important component of long-term smart grid solutions. Examples of this are in some jurisdictions in California, in the State of Western Australia and in Upstate New York, where microgrids have been incentivised.
Additionally, as regulatory requirements on carbon emissions become more stringent, data centres with microgrid systems are better positioned to comply with evolving standards. These facilities can manage their energy resources more dynamically, enabling them to meet sustainability targets while minimising costs; and with the data collection and reporting capabilities, they can demonstrate their net zero pathway progress.
In a world where energy demand is only set to increase, microgrids stand as a forward-thinking solution that not only meets operational needs but aligns with sustainability goals and additionally offers economic advantages. For data centres committed to reliable, efficient and environmentally responsible operations, microgrids are no longer optional – they are essential.
