How high-resistance grounding keeps data centres online

In December 2025, the UK’s National Energy System Operator (NESO) overhauled the grid connection process, replacing first come, first served with a readiness-based approach. The reform prioritises projects that are demonstrably deliverable, especially large demand connections such as data centres.

Data centres were designated Critical National Infrastructure in September 2024, placing them on “an equal footing with water, energy and emergency services systems”. 

At the same time, electricity demand is forecast to rise sharply as transport, heating and industry electrify. 

National Grid’s Future Energy Scenarios have repeatedly indicated that there will be substantial growth in peak electricity demand by the early 2030s, while connection queues in recent years have run into the hundreds of gigawatts across generation and demand. 

In that environment, transmission and distribution operators are under pressure to ensure that every new connection supports rather than destabilises network performance.

The connections context

Under the Connections Reform, NESO now prioritises projects that are first ready and needed. Many transmission-connected demand schemes, including data centres, have been allocated firm capacity blocks with connection dates extending into the next decade. 

The intention is to align connection dates with realistic delivery schedules and system need, but it also raises the technical bar.

For utilities, readiness is now both a contractual milestone and an engineering one. Developers must demonstrate design maturity, credible delivery plans and robust risk management. 

Increasingly, network operators are scrutinising protection philosophies, fault contribution studies and system strength assessments at an early stage. 

The question is not only whether a facility can draw power, but how it will behave during a disturbance, how it will respond to an earth fault and whether it could inadvertently trigger wider disconnection.

Large data centres can represent tens or even hundreds of megawatts of concentrated demand. Their internal generation, battery storage and power electronics can alter local fault levels and transient behaviour. 

In a grid that is already integrating growing volumes of inverter-based renewables, predictable fault management becomes essential. Utilities assessing these connections are therefore looking for clear evidence that protection systems will operate selectively and that first faults will not escalate into wider outages.

High-resistance grounding

One mechanism central to grid-friendly design is high resistance earthing, also known as high resistance grounding.

Although industry benchmarking shows that overall outage frequency has declined over the past decade, data from the Uptime Institute consistently indicates that power remains the leading cause of the most impactful incidents.

In its recent survey, more than half of operators cited power issues as a primary cause of major outages.

The financial impact of significant downtime can run into six figures per incident, with reputational and contractual consequences that extend far beyond the site boundary.

In high resistance grounding systems, the transformer or generator neutral is connected to earth through a high-value resistor.

By introducing resistance, the system limits single line to earth fault current to a controlled value, typically a few amps at low voltage or tens to hundreds of amps at medium voltage.

In contrast, a solidly earthed low-voltage system can experience earth fault currents of several thousand amperes.

The result in a high resistance grounded system is that a first earth fault becomes an alarm-only condition rather than an immediate trip.

Operations teams can continue supplying critical loads while locating and clearing the fault. For utilities, this behaviour reduces the likelihood that a transient or localised insulation failure will propagate upstream and operate network protection.

High resistance grounding also avoids the disadvantages of alternative strategies. Compared with solid earthing, it limits single line to earth current and reduces arc flash energy for that fault type.

Compared with ungrounded systems, it maintains a defined neutral reference, suppresses sustained transient overvoltages and improves fault detection sensitivity.

In data centre environments, where multiple uninterruptible power supplies, standby generators and static switches operate in parallel, coordination margins are tight. Nuisance tripping during a generator transition or a minor insulation fault can cascade rapidly.

By constraining first earth fault current and maintaining system stability, high resistance grounding supports selective protection and reduces the risk of wider disturbance that could be visible at the network level.

The critical component

An NER is the enabling component of a high resistance grounded system because it establishes the magnitude and duration of the earth fault current. By selecting the resistance value and thermal rating carefully, designers define how the system will behave under fault and how protection devices will respond.

For utilities reviewing connection applications, the specification of the NER provides insight into the maturity of the overall protection philosophy. The resistor must be rated for the expected fault current and duration, whether short time duty measured in seconds or continuous operation in specialised schemes.

Thermal performance, enclosure design and environmental resilience are not cosmetic details but factors that influence long-term reliability in plant rooms, substations and exposed compounds.

Protection coordination is closely linked to resistor selection. Earth fault relays must be set with sufficient sensitivity to detect the limited current, while maintaining grading with downstream devices.

Current transformers, monitoring equipment and alarm circuits must be integrated so that a first fault is clearly signalled without ambiguity. If the resistor is underspecified, protection may fail to detect a fault. If it is oversized without coordination, unnecessary trips may still occur. In a readiness-based connection regime, such misalignment represents a tangible delivery risk.

From a network perspective, well specified NERs within high resistance grounding schemes contribute to predictable fault levels and improved insight. They reduce the likelihood that an internal earth fault will drive upstream protection to operate, thereby supporting continuity of supply not only for the facility but also for adjacent customers.

A grid-ready approach

As grid connections are reprioritised and electrification accelerates, utilities face the challenge of integrating large, complex loads without compromising stability or resilience. Earthing philosophy may once have been viewed as an internal design choice, but under a readiness-based framework it becomes part of the wider system conversation.

Well-engineered high resistance grounding schemes contain the first earth fault without immediate tripping. They limit fault current, support selective protection and demonstrate that the facility has been designed with predictable network interaction in mind. For data centres now recognised as Critical National Infrastructure, this is essential to maintaining continuous service.

For utilities operating an increasingly constrained and dynamic grid, it is evidence that new demand can connect without introducing avoidable risk.

In that sense, high resistance grounding is not simply a resilience measure within a single site. It is part of a broader shift towards grid-compatible design, where large energy users and network operators share responsibility for stability in an evolving electricity system.

About the author: Mike Torbitt is managing director of resistor manufacturer Cressall Resistors.