Electricity consumption from data centers is projected to more than double by 2030, according to the IEA’s Report on Energy and AI from April 2025. With AI workloads expected to drive up to 30 percent of total demand by 2027, the global energy landscape is being redefined. In our September webinar, Powering the Digital Future: Grid Resilience Amidst AI Data Center Growth, we explored how this transformation is unfolding in Europe and the Middle East with insights from Dr. Bruce Stedall, COO at TAQA Transmission; Gregorio Ogliaro, MD Global Utilities at Accenture; and our Chief Strategic Growth Officer, Duncan Burt. This blog is a summary of the discussion and key insights.
AI’s rapid expansion is reshaping the grid
Gregorio Ogliaro, Accenture’s MD shared that in Italy alone, grid connection requests have surged from 1 GW to 60 GW in last four years, and that shift is being driven by the explosive growth of AI data centers, which are expected to grow 30 percent annually over the next five years.
Dr. Bruce Stedall from TAQA Transmission highlighted similar trends in the Middle East, where grid capacity is projected to triple by 2030. He emphasized that delivery timelines have been compressed from five to seven years down to just 24 to 36 months and this level of acceleration requires not only new technologies but also new frameworks for permitting and procurement.
Dr. Bruce Stedall further summed up the challenges that AI data centers pose to grid stability:
“Firstly, the AI data centers being developed now have very high and concentrated demand, that need to be addressed in a very specific area. Secondly, load variability is also an issue as the AI algorithms training workloads go through different phases. The third point is the need to manage the number of inverter-based connections. So, as well as addressing these from the supply side with clean energy migration of generation, there’s also the need to recognize how those are connected.”
These challenges demand innovative solutions, from real-time measurement to hybrid energy models. They also underscore the growing importance of real-time data and measurement technologies in identifying the variability of data center demand and effective integration of AI data centers into the grid.
The importance of measurement data
Duncan Burt addressed multiple questions related to the 1.5 GW event in North America in July 2024, when 60 data centers tripped simultaneously, causing significant voltage swings:
“The most important thing is getting measurement data on how your grid is behaving now.”
Duncan added that this data helps utilities understand whether their grid model is properly calibrated to actual system performance. He also mentioned that the grid event was so significant that the North American Electric Reliability Council (NERC) launched a Large Load Task Force, and in their reporting, most of their high priority recommendations were around measuring real-time grid stability in terms of voltage and frequency.
Gregorio Ogliaro echoed the points Duncan made and stated:
“Data is a strategic asset. To unlock its full potential, it’s essential to build solid foundations, the so-called ‘Digital Core’.”
He added that effective data management and the right operating model are key to fully leveraging the power of artificial intelligence.
Reactive Technologies’ GridMetrix® provides this much needed, real-time, high-resolution, measurement data and data-driven insights to grid operators around the world, and has been successfully deployed in the UK, the Middle East, Australia, Japan, Taiwan and North America. With the high-performance data from GridMetrix® platform, grid operators can monitor critical stability parameters and make informed decisions about renewable integration and system resilience.
Dr. Bruce Stedall continued:
“It’s clear that aspects like data are very important and we’re working with the likes of Reactive Technologies in that area. And it’s amazing how these technologies start to have other benefits.”
This shift opens the door to a new question: Can AI data centers move from being grid liabilities to becoming grid assets?
Integrating AI data centers into the wider energy system
The short answer is yes; AI data centers can be better integrated into the energy system and even become grid assets. Dr. Bruce Stedall referenced the pioneering hybrid model of the Stargate UAE project in Abu Dhabi, which combines 1 GW of centralized demand with 1 GW across decentralized sites. It is powered by 3.7 GW of solar PV and a 19 GWh battery storage platform, complemented by conventional generation for resilience. This model is a great example of how large-scale digital infrastructure could be integrated into the broader energy system.
Gregorio Ogliaro further emphasized how data centers are increasingly becoming flexibility providers. With AI and machine learning, they can adjust energy consumption based on grid conditions, shifting non-urgent compute tasks away from peak hours and participating in demand response programs. This marks a shift from them being passive consumers to active contributors to grid.
Key takeaways
Across Europe and the Middle East, utilities are adapting to the challenges posed by hyperscale AI data centers.
In Europe, grid saturation and connection bottlenecks are driving the need for better data and cross-sector coordination. Gregorio Ogliaro emphasized the importance of data as a strategic asset and building a digital core to unlock the full value of data and support strategic decision-making.
In the Middle East, Dr. Bruce Stedall pointed to the need for streamlined delivery frameworks and anticipatory investment to meet compressed timelines.
As Duncan Burt reminded us: “If you measure it, you can manage it.” Yet to truly manage the grid impact of AI data centers, utilities, regulators, developers and technology providers must go beyond measurement, building shared visibility and joint frameworks that turn high-resolution data into coordinated action.
Learn More
To explore how real-time grid measurement data can support your planning and operations, or to see GridMetrix® in action book a quick call with me or visit our website for more information.
