FEATURE Energy & renewables
DELIVERING 99.99% AVAILABILITY THAT AI OPERATIONS DEMAND
All images courtesy of Emerson
The question is no longer whether renewable energy can support AI data centres. It can. The real question is whether organisations can intelligently orchestrate renewable generation, storage, conventional power, and grid resources into a resilient energy ecosystem. By Thomas Andersen, vice president of renewable technologies for Emerson
M
eeting AI data centres’ extreme uptime requirements requires more than generation capacity. It demands intelligent coordination
across every energy-producing and energy-consuming asset. Around the globe, artificial intelligence (AI) data centres are rapidly changing the energy equation. Traditional enterprise data centres were largely passive consumers of electricity. Modern AI facilities are different. Many operate at hundreds of megawatts or even gigawatts and are increasingly being viewed as large-scale industrial energy systems rather than commercial buildings. In parallel with this shift, many AI data
centre operators face growing pressure to reduce carbon emissions and increase renewable energy use. However, AI requires reliable power 24/7 even though renewable generation is inherently variable. Renewable energy can contribute, but to do so, it must be integrated reliably. AI data centres typically fall under one of
three frameworks: • Pure grid connection • Pure behind-the-meter (islanded) generation
• Hybrid solutions utilising behind-the- meter and grid generation
Each of these models has its own strengths and weaknesses, but regardless of the model, planning and coordination typically matter more than the specific energy source. Wind and solar can absolutely support data centre operations. In fact, in some cases, renewable projects can be deployed faster than traditional
16 October 2026 | Automation
generation technologies due to long lead times on gas turbines and other equipment. However, renewable energy output varies with changing weather and environmental conditions, but data centre operations do not. Accordingly, operators must account for energy storage needs, backup generation, and reserve capacity. They must also have effective methods for forecasting and planning, both around weather and energy availability. This is where advanced energy management and
control become critical.
A modern AI data centre technology stack and ecosystem is typically quite complex. A single facility might include renewable generation, battery energy storage systems (BESS), utility grid connections, traditional generation assets, and rapidly changing AI workloads. At any given time, operators are continuously responsible for answering questions such as which energy source to use, when to charge and discharge batteries, when to draw power from the grid, and when and how to build reserves for future needs. Further complicating operation, modern
organisations frequently operate many distributed sites and need a way to centrally manage all those locations. To successfully accomplish these goals, whether at a single facility or across many distributed resources, organisations need load forecasting, renewable energy forecasting, BESS scheduling, reserve management, economic dispatch optimisation, and more. The best energy management systems (EMS) bring all these inputs together and continuously determine the most effective operating strategy.
Another challenge common in AI data
centres is increased operational complexity. AI data centres can experience significant load swings in seconds or even milliseconds, much faster than operators can react to make changes manually. Moreover, not all power producing assets use the same control systems, with many essential elements relying on OEM controls. Teams need fast communications and a unified control layer to successfully maintain stability across all energy sources. Today’s most effective organisations accomplish this goal via asset performance management (APM) software that helps operators monitor equipment health, battery health, system degradation, reliability risks, and more. In addition, many teams are turning to digital twin systems, where they can simulate failures and validate operating strategies before problems occur (see above). As organisations strive to quickly design and deploy the most efficient, effective AI data centres around the globe, they will continue to face challenges and complexities surrounding energy use. The ultimate solution is not simply adding more generation. The more effective strategy is to develop tighter coordination between the data centre, onsite assets, energy storage, and the electrical grid. Modern technologies like EMS, APM, digital grid management, digital twins, and predictive analytics, empower operators to understand both electricity production and consumption in real time, and to allocate resources in the safest, most effective way possible.
Emerson
www.emerson.com
automationmagazine.co.uk
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40
