Executive Summary

• The UK is facing a near-20% surge in data centre capacity with over 100 new projects alongside the 450+ existing sites. And the question on everyone’s mind is: can the electrical infrastructure cope with the rapid AI data centre expansions?
• Traditional data centres were predictable: AI data centres are anything but that. AI workloads don’t just consume more electricity; they consume it differently, thus creating power quality challenges.
• These technical challenges arrive as UK data centres face significant grid capacity constraints. Industry analysts have declared that grid connection queues have stretched to unprecedented lengths, with some projects facing waits of five years or more.

 

With the world’s biggest companies announcing major AI data centre expansions across the UK, the headlines are publicising job creation and technological leadership, yet behind the scenes engineers are asking a more pressing question: can Britain’s electrical infrastructure cope?

The country faces a near-20% surge in data centre capacity, with over 100 new projects in the pipeline alongside its existing 450+ sites. This explosive growth, fuelled by soaring AI computing demands, is spawning clusters around London and the Thames corridor, as well as emerging hubs in Manchester, Leeds, Wales, and Scotland.

The answer reveals uncomfortable truths about the collision between AI ambitions and power grid reality. Across the UK, data centre operators are discovering that artificial intelligence workloads don’t just consume more electricity; they consume it differently. This creates power quality challenges that threaten equipment reliability, grid stability, and the viability of aggressive AI deployment timelines.

A New Electrical Landscape

Traditional data centres are predictable. Server racks hum along at steady loads, cooling systems maintain consistent draw, and power consumption follows recognisable daily patterns. But AI has shattered that predictability.

“The electrical behaviour we’re seeing with AI workloads is fundamentally different,” explains Roshan Rajeev, Vice President of Engineering at Janitza USA, who previously managed data acquisition and analytics at Meta’s hyperscale facilities. “Model training creates sustained loads in the megawatt range; massive base loads that stress utility infrastructure. But inference operations are even more challenging from a power quality perspective.”

Inference, the process of running AI models to generate outputs, creates what engineers call “burst activity”. A facility might spike from baseline to peak consumption in seconds as thousands of users simultaneously query large language models (LLMs). These high power, short-duration surges (or overvoltages), sometimes increasing by 100 kilowatts within ten seconds, generate voltage sags, transients, and flicker that propagate through electrical systems.

The challenge extends beyond raw capacity to the dynamic nature of AI loads. Traditional data centres might vary by ten or 15% throughout the day, following predictable patterns as businesses open and close. By contrast, AI facilities can experience 50% load swings within minutes, creating extraordinary difficulties for distribution networks designed around more stable consumption profiles.

As Roshan emphasises: “Live power quality data is non-negotiable, and AI workloads are constantly evolving. You need visibility into voltage harmonics, current distortion, and power factor variations, and you need to access this data remotely so you can respond immediately.” This visibility relies on modern, standards-compliant monitoring and analysis hardware, an area in which Janitza has established its global reputation.

The hardware compounds the problem. Graphics processing units and tensor processing units powering AI computation draw massive amounts of current in highly nonlinear patterns. This creates harmonic distortion, electrical “noise” that can damage transformers, interfere with sensitive equipment, and reduce system efficiency. In colocation environments hosting multiple tenants, aggregate harmonic distortion at the point of common coupling can exceed safe thresholds, affecting all customers.

Britain’s Infrastructure Reality

These technical challenges arrive as UK data centres face significant grid capacity constraints. According to industry analysts, grid connection queues have stretched to unprecedented lengths, with some projects facing waits of five years or more. The National Energy System Operator (NESO) for the UK has warned that data centre electricity demand could increase sixfold by 2030, requiring network reinforcement.

For AI facilities, the situation is acute. A single large-scale AI data centre can require 100MW or more, roughly the output of a small gas turbine plant. Distribution network operators, already managing connections for renewable energy projects, struggle to accommodate these enormous, unpredictable loads without risking grid stability.

“The power quality challenges we’re seeing in AI data centres represent a step change from traditional facilities,” says David Gilligan, VP Critical Power Solutions & Technology – Global at Janitza Electronics. “Operators must monitor electrical parameters that previously weren’t critical: transients as brief as 18 microseconds, voltage harmonics up to the 127th order, and rapid load fluctuations that can destabilise distribution networks. Here, Janitza’s UMG 801 power analyser resolves these challenges, making it a foundational tool for next-generation data centre management.”

The regulatory environment adds complexity. UK planning frameworks, designed for conventional development, struggle to accommodate the speed of AI deployment. Energy efficiency requirements under Building Regulations Part L and ESOS compliance create additional hurdles for operators trying to balance performance with sustainability commitments.

Solutions Taking Shape

Forward-thinking operators are implementing sophisticated approaches to manage these challenges. Real-time power quality monitoring has evolved from optional to essential, with comprehensive systems capturing electrical parameters at millisecond intervals to detect problems before they cascade into failures.

Janitza’s UMG 801 expandable modular power analyser offers the granularity required for AI environments, recording voltage harmonics up to the 127th order and detecting transients from 18 microseconds. With modular expandability to 92 current measuring channels and high-frequency sampling at 51.2 kHz, the UMG 801 delivers the visibility operators need across complex electrical topologies. Its dual Ethernet interfaces and OPC UA connectivity ensure secure integration with Building Management Systems and centralised platforms, such as Janitza’s GridVis® Power Grid Monitoring Software.

Integration with GridVis enables remote analysis and ISO 50001-compliant energy management, crucial for facilities where power quality events can occur anywhere across distributed infrastructure. GridVis supports ongoing compliance and increases operational transparency, letting operators act quickly in response to anomalies.

“High availability in data centres demands continuous monitoring according to standards like IEC 61000-2-4 and IEEE 519,” explains David. “The challenge with AI workloads is that traditional monitoring can’t capture the speed and complexity of electrical events. You need devices that can record fast transients whilst simultaneously tracking harmonic distortion across dozens of measurement points.” Janitza’s modular system approach meets these evolving needs.

Battery energy storage systems (BESS) have emerged as a critical tool for load smoothing in datacentres. By absorbing demand spikes and releasing power during lulls, BESS installations help flatten the load profile presented to the grid, reducing stress on upstream infrastructure. Several UK facilities have deployed multi-megawatt battery systems specifically to manage AI workload variability.

On-site generation is gaining traction too. Combining grid power with natural gas generators or hydrogen-ready systems provides redundancy while reducing grid dependence during peak AI operations, with some operators are exploring hybrid models that integrate renewable generation with storage and conventional backup power.

Looking Ahead

As the UK government pursues AI infrastructure investment as part of broader digital economy ambitions, the power quality challenge will intensify. Projections suggest AI computing capacity must increase tenfold by 2028 to meet anticipated demand, requiring additional grid connections and advanced monitoring.

The sector’s response will determine whether the UK can realise its AI ambitions or lose ground to regions with more robust electrical infrastructure. Success requires collaboration across the ecosystem: investment in monitoring and mitigation technology from operators, network upgrades from utilities, flexible regulatory frameworks, and new talent pipelines.

“Every data centre behaves differently once you factor in cooling systems, tenant mix, and AI workload types,” says Roshan. “There’s no one-size-fits-all solution. What we need is flexible, modular approaches that adapt as conditions evolve.” Janitza’s modular, scalable hardware and software systems are designed for this adaptability, a necessary attribute in today’s market.

For colocation providers, the challenge is acute. Cost centre management and tenant billing must now account for power quality impacts, not just consumption. Meeting fire protection requirements through comprehensive residual current monitoring becomes more critical when electrical systems operate under stress, a challenge Janitza technology directly addresses.

The AI revolution promises transformation on every level, but it must first overcome the challenge of maintaining stable, efficient, and resilient energy systems in the data centres that power innovation.

For UK datacentres, that challenge has arrived, and the clock is ticking!