Unlocking latent grid value: Reconfiguring generation and storage connections for data centre power

The SLR data centre thought leadership series continues with a short three-part focus on the power grid. Part two of this power grid series explores how reconfiguring existing infrastructure can help unlock grid capacity for increasing AI compute demand.

As the build-out of AI digital infrastructure accelerates globally, particularly in energy-constrained markets, developers are increasingly encountering a fundamental challenge: how to access the power needed to meet growing AI compute demand in time frames that work with increasing market demand.

Repurposing existing infrastructure

With new grid capacity becoming harder to secure and existing queue times going beyond data centre developer target timelines, developers are beginning to ask a different question: What if we can reconfigure what already exists?

Routes that have already gained momentum include utilising existing legacy thermal power stations and legacy large industrial grid connections where that industry has gone to another country. Another question developers are asking is: What if we can reconfigure proposed connections in the future? Converting new proposed generation / storage grid connections for data centres or other large new industry demand is also a possibility.

However, this isn't a simple process. Repurposing grid connections from generation or storage to demand use has its own technical and commercial challenges. Whilst legacy grid connections may have been designed for firm operation, some of the new generation / storage proposed grid connections come with non-firm or interruptible agreements; and some will only have export capability. Electrically, the infrastructure may need to be re-designed to meet demand-side power quality, fault level, and resilience standards.

How SLR can help

SLR is supporting energy generators and data centre developers with feasibility studies helping to inform early engagement with the network operator. We believe this approach is essential prior to demand side grid connection applications. The demand side use case will typically trigger a material change of use, resulting in a fresh connection application. In constrained grid areas, that could be the difference between progressing or stalling development.

The economics change quickly, especially where grid rights or capacity sharing arrangements are still maturing. We are regularly advising clients on techno-economic-risk modelling that considers all these variables, including whether there’s any retained value in integrating behind-the-meter microgrids and storage to mitigate peak cost exposure or enhance resilience and uptime.

At a micro level, from a technical design perspective, we recommend a full review of load profile, cooling demand, switching transients, and the interaction between UPS systems and grid-side protection. This analysis is needed for a resilient, compliant, and future-proof power solution. Our Power Systems team is advising on power quality, which is essential for data centres and many large industrial demand users to avoid equipment failure – high capex equipment that often comes with very tight tolerances on power quality. This area can be often overlooked, especially when integrating intermittent renewable generation.

At a macro level, power market reforms also need to be considered. Globally, grid connection queue reforms, the increasing use of deliverability criterion, and growing interest in dynamic regional system planning are shaping how capacity is allocated. Projects that can demonstrate national significance, feasibility, and integration with flexibility mechanisms are more likely to progress. We see and welcome many markets evolving to better address the challenges of net zero in a more holistic manner.