Executive Summary

• As AI workloads continue to expand, operators are under pressure to safely and reliable deliver more computing power than ever before while balancing rising energy costs – and to meet these objectives, Colin Dean of Socomec writes, access to high-quality data is essential
• As equipment requirements increase, power quality monitoring has a continuous impact on power quality, potentially impacting equipment lifespan, maintenance costs and the energy efficiency of facilities
• Energy visibility not only helps operators understand how power usage patterns across their facilities, but also how their electrical infrastructure will cope in these environments

The UK’s data centre boom is well underway, as recent Ofgem figures indicate that 315 data centre projects are currently seeking grid connections, equivalent to 73GW of demand – more than one and a half times the UK’s peak national electricity demand. As AI workloads continue to expand, operators are under pressure to safely and reliably deliver more computing power than ever before, while balancing rising energy costs and demonstrating compliance with sustainability targets.

To meet these objectives, access to high-quality energy data is essential. Detailed visibility into energy use helps operators optimise Power Usage Effectiveness (PUE) – the key efficiency metric for data centres – while also providing the insight needed for effective capacity planning, helping to identify potential overload risks and support operational resilience.

Data centres need visibility beyond the utility meter

Improving PUE in data centres requires more than a high-level view of overall energy consumption. As competition for power intensifies, operators need a detailed understanding of where energy is being used across their facilities to make the most of available capacity, from cooling systems and power distribution equipment through to individual data halls and racks.

Granular power monitoring of electricity consumption by circuit, zone and use provides this visibility, enabling teams to establish energy baselines for individual systems, track performance over time, and identify irregularities that might otherwise go unnoticed.

Power quality monitoring is also essential for operators to gain full insight into their facility performance. As equipment requirements increase, this has a continuous impact on power quality, potentially impacting equipment lifespan, maintenance costs and the energy efficiency of facilities. With the energy transition driving growth in the levels of renewables and battery energy storage systems (BESS) throughout the grid, these can be another potential source of power quality issues on the incoming supply. It is therefore imperative that data centre operators have a precise understanding of the sources of any poor power quality to maximise the efficiency of their sites.

The result is a clearer picture of overall energy performance and a stronger foundation for improving efficiency in line with tightening sustainability requirements. Indeed, under the EU’s Energy Efficiency Directive (EED), data centres must disclose PUE along with key environmental performance indicators on an annual basis to demonstrate they’re working towards sustainability goals. However, the EED is more than just a compliance exercise. In the UK and elsewhere, the same good practices adopted by EU data centres to optimise PUE can be the drivers of innovation, competitiveness and growth.

Energy data supports informed infrastructure investment

Granular energy monitoring is also fundamental for scaling data centre operations. As facilities take on new customers, higher rack densities and more demanding AI workloads, operators need a clear understanding of whether existing power infrastructures can support this growth. By basing infrastructure upgrades and expansion plans on measured performance rather than estimates or aggregated energy data, operators can maximise available capacity and make informed decisions about where and when additional investment is needed.

A flexible, modular approach to metering can make this process easier. Monitoring points can be added as requirements evolve, allowing operators to focus first on the areas where greater visibility will have the biggest impact. This helps avoid disruptive, large-scale upgrades, which can be particularly challenging in legacy data centres.

For colocation providers, the value of accurate metering extends to fair and transparent tenant billing, while giving customers detailed insight into their own energy consumption and associated environmental footprint. To ensure confidence in the data, meters must comply with the Measuring Instruments Directive (MID), which sets strict requirements for accuracy. The highest classification, Class C, delivers accuracy of up to 0.5 percent across a wide operating range, ensuring reliable performance even as conditions change.

Managing overloads in power-dense environments

Finally, energy visibility not only helps operators understand how power usage patterns across their facilities, but also how their electrical infrastructure will cope in these environments.

The demands placed on facilities to support critical applications and multiple tenants mean that overloads and short-circuit faults can have far-reaching operational consequences, making electrical resilience a priority for operators today. In ultra high-density designs, precise monitoring frameworks directly correlate to fewer overcurrent blind spots, allowing facilities to maximise their power infrastructure safely without risking catastrophic structural downtime.

The way monitoring systems are integrated into a data centre’s electrical systems plays an important role in determining both the quality and reliability of the insight operators receive, as well as the resilience of the monitoring framework itself.

Visibility underpins resilient data centres

Managing the mounting demands on data centres begins with a clear understanding of how power is used and where pressure points may emerge. Operators would do well to prioritise energy monitoring as a key architectural consideration of data centre design, to ensure facilities are engineered to withstand increasingly demanding electrical conditions, while continuing to support the critical services that rely on them.

Picture credit: Created with Magnific

—

About the author: Colin Dean is managing director of Socomec UK – an independent industrial group specialising in the availability, control and safety of low-voltage electrical networks. With over 30 years’ experience in the electrical and data centre industry, Colin leads the UK and Ireland business, supporting the delivery of resilient power infrastructure for critical environments.