How Energy Supply Stays Resilient
Reliable supply is engineered, paid for and defended. The system keeps spare capacity for hard days, uses emergency tools when supply tightens, protects vulnerable consumers, and plans restoration if a major loss occurs. Gas and electricity resilience also depend on each other, because power stations use gas and gas networks need electricity.
Scope: how the energy system withstands and recovers from stress, from a cold-snap supply squeeze to a cyber attack or a major loss of supply. The markets that price energy day to day are on the markets page; the wires themselves are on the network page.
Sources and standards
Quantitative and regulatory claims resolve to a primary source: NESO's Winter Outlook and balancing reports, the Capacity Market parameters, National Gas operational data, Ofgem's consumer-vulnerability and price-cap material, the NCSC Cyber Assessment Framework and the NIS Regulations 2018, and the Electricity Act 1989 on legislation.gov.uk.
Security of supply: the margin and the reliability standard
The starting point is the margin: how much generating capacity is expected to be available above the level of demand on a tight winter day. NESO assesses it each year in the Winter Outlook; the de-rated margin for winter 2024-25 was around 6.3 gigawatts, the buffer that sits above peak demand once each technology is discounted for the chance it is unavailable when needed.7 The margin is not chosen for its own sake. It is whatever is needed to meet a reliability standard, set in statistical terms.
That standard is the Loss of Load Expectation: no more than three hours per year, on average across many simulated weather years and plant-availability outcomes, in which available supply cannot meet demand. Three hours is not a prediction that the lights will go out for three hours; it is a probability target that the system is planned to. Every capacity decision reduces, in the end, to whether the portfolio meets the three-hour bar.
What the three-hour standard means
The Loss of Load Expectation averages shortfall risk across thousands of simulated years. A standard of three hours per year does not mean three hours of cuts every year; it means that, over the long run, the expected total is about three hours:
most years: 0 hours | a rare severe year: many hours → long-run average ≈ 3 hours/year
The point of the standard is to size the margin against rare bad years, not average ones. The Capacity Market then buys whatever availability is needed to hold that bar.
The Capacity Market is the instrument that procures the availability. Through annual auctions, four years ahead (T-4) and one year ahead (T-1), it pays generators, storage, interconnectors and demand-side providers to be available when the system needs them, on a pound-per-kilowatt-per-year basis rather than for the energy they produce. The most recent auctions cleared around 40 gigawatts and several gigawatts respectively, at clearing prices published with each result.1 The market is the standing way the three-hour standard is met; flexibility, the ability to shift demand and discharge storage when the system is tight, is the growing complement to it, with the government's clean flexibility roadmap targeting a large rise in flexible capacity by 2030.5
| Measure | Figure | What it is |
|---|---|---|
| Reliability standard | 3 hours/year | The Loss of Load Expectation the system is planned to |
| Winter 2024-25 de-rated margin | about 6.3 GW | Expected spare capacity above peak demand7 |
| Capacity Market | around 40 GW cleared (recent T-4) | Availability bought to meet the standard1 |
| Rota disconnection | not used since the 1970s | The last-resort measure the margin exists to avoid |
When the system is tested: the three response tiers
When a cold snap or a sudden loss of supply tightens the system, the response escalates through three tiers, each more costly than the last. Good resilience keeps the response in the lower, cheaper tiers and reserves the top tier for genuine emergencies.
| Tier | What happens | Who acts | Relative cost |
|---|---|---|---|
| 1. Market | Prices rise; generators run more, flexible users shift demand, storage discharges, imports increase. Handles most stress. | Market participants | Lowest |
| 2. Balancing | NESO issues system warnings and buys extra reserve; National Gas calls on interruptible contracts and large-user reductions. | NESO, National Gas3 | Higher |
| 3. Emergency | If the gap is still open, emergency procedures apply, with rota disconnection as the last resort under the Electricity Supply Emergency Code. | Secretary of State, NESO | Highest |
The top tier rests on statute. Section 96 of the Electricity Act 1989 lets the Secretary of State direct measures to mitigate an emergency that threatens to disrupt electricity supplies, and the reserve powers in the Energy Act 1976 allow the use of electricity to be restricted, with the Electricity Supply Emergency Code setting out how rota disconnection, planned, rotating outages across regions, would be applied if it were ever needed.8 It has not been used since the 1970s, and the entire margin-and-market machinery exists to keep it that way. Flexibility is increasingly the cheaper alternative to the higher tiers: a system that can pay consumers to shift demand away from the tight hours needs less expensive standby plant to hold the same standard.5
The three response tiers as an escalation ladder
Based on NESO's system notices and balancing services and the gov.uk Electricity Supply Emergency Code (revised April 2026). Each tier is dearer and rarer than the one below, and is used only when the tier below cannot close the gap.
The market tier clears most stress through price signals and the voluntary Demand Flexibility Service, which has run year-round since 27 November 2024. If that is not enough, NESO buys reserve and issues warnings that escalate in order: a Capacity Market Notice when the margin falls below 500 megawatts about four hours ahead, an Electricity Margin Notice calling for about 500 to 800 megawatts more margin, the gas Margins Notice and Gas Balancing Notification, then High Risk of Demand Reduction and Demand Control Imminent. Only if the gap is still open is the statutory tier reached: under the Electricity Supply Emergency Code, rota disconnection splits each network into 18 load blocks of about 5 percent of demand, lettered A to U, each off for a nominal three hours in turn, with protected sites exempt. The legal basis is section 96 of the Electricity Act 1989 and the Energy Act 1976 reserve powers, and it has not been used since the 1970s.
Weather: the 2018 cold snap, a stress event the system passed
The clearest recent test was the cold spell of late February and early March 2018, the "Beast from the East". A weather system over Scandinavia pulled temperatures down sharply across Britain, and gas demand for heating spiked to near a single-day record. Liquefied natural gas arrivals thinned at the same time, and the large Rough storage facility had closed the year before, so the system's buffer was unusually thin. National Gas issued a Gas Deficit Warning, its first in over a decade.2
No household lost supply. Prices rose sharply, large industrial users were asked to cut their gas use, and the system held. The lesson was not that it was unsafe but that the buffer had thinned: the cushion of gas stored inside the pipes, normally several hours, fell to a fraction of that during the worst day. The reopening of Rough storage in 2022 and the growth of electricity-side flexibility through the 2020s are, in part, responses to that event. A cold, still, dark week, high heating demand, low wind, short days, remains the canonical stress case the margin and the response tiers are sized against.7
Gas and electricity lean on each other
The two systems are not independent. Gas-fired power stations still provide a large share of electricity, especially on cold, low-wind days, so a squeeze on gas is also a squeeze on power. The link runs the other way too: the gas network depends on electricity for its compressors and control systems. A cold, still day stresses both at once, because the same weather lifts heating demand (gas) and suppresses wind output (electricity), pulling harder on the gas-fired plant that bridges the gap.
Gas has its own resilience toolkit. Linepack, the gas held under pressure inside the transmission pipes, gives a short operating buffer through the day; storage, including the reopened Rough facility, holds a larger reserve; liquefied natural gas terminals and pipeline imports bring gas in; and interruptible contracts let large users be asked to cut demand first.2 Because the two systems are coupled, resilience has to be planned across both together: NESO's winter assessment looks at electricity and gas side by side, since the worst day for one is usually the worst day for the other.7
Beyond weather: cyber attack and restoring the system
Not every threat is the weather. The control systems that run the grid, the substations, the smart meters and the market platforms are all potential targets for a cyber attack, and a successful one could disrupt supply as surely as a cold snap. Energy operators are treated as operators of essential services under the Network and Information Systems Regulations 2018, which place security and incident-reporting duties on them, with Ofgem and the energy department as the regulators for the sector.10 They are assessed against the National Cyber Security Centre's Cyber Assessment Framework, which sets out what good security looks like across managing risk, protecting systems, detecting events and minimising impact.9 Security is built into the metering devices and the data pipe as well, as the digital infrastructure page sets out.
Resilience also means being able to recover. If a large part of the system were ever lost, it would have to be rebuilt from the ground up, a process called restoration, or black start. Because most power stations need electricity to start, a small set of restoration providers can energise without an external supply and then bring others back in a controlled sequence, rebuilding the network island by island until it reconnects into a whole. NESO contracts and exercises this capability and reports on it through its system and balancing reporting, and the statutory emergency powers above sit behind the restoration arrangements.38 Restoration is the backstop behind the backstop: the plan for the day the margin and the response tiers are not enough.
How the grid is restored after a total shutdown
Based on NESO's Electricity System Restoration Standard and Distributed ReStart programme. The six steps are the engineering sequence of a restoration, read left to right; only the two published ESRS deadlines are quantified in time.
From a total shutdown, a self-starting anchor generator, often a distributed resource, makes the first live voltage; a 33 kilovolt power island forms around it; demand is restored in small blocks; more generation and load grow the island; separately energised islands are phase-matched and joined; and the joined islands reconnect to the main grid until all demand is back. At every step generation and demand must match second by second with little inertia, so too large a block of load can collapse the island and force a restart. The Electricity System Restoration Standard requires NESO to restore 60 percent of regional demand within 24 hours and 100 percent of Great Britain's demand within 5 days, with full compliance by 31 December 2026.
Protecting people: vulnerable consumers and power cuts
A loss of supply does not fall on everyone equally. A household that depends on electricity for medical equipment, or that cannot keep warm, is harmed faster and more severely than most. The Priority Services Register is the free list that network operators and suppliers keep of people in vulnerable circumstances, so that they are contacted first during a planned or unplanned power cut, given extra help, and, where possible, protected from disconnection. Ofgem's consumer-vulnerability strategy sets out these protections and the duties on suppliers to identify and support people at risk.11
Everyday power cuts, a fault on a local line, a storm bringing down a cable, are handled by the regional network operator, which is held to standards for how quickly supply is restored and pays automatic compensation when it is not. Affordability is the other side of the same coin: a household that cannot afford to heat its home is exposed to cold-weather risk regardless of whether the system as a whole holds. The default tariff cap limits the price of a standard tariff,4 and the latest official statistics estimate 9.4 percent of households in England in fuel poverty, the population most exposed when the system is stressed.6 Keeping the system upright is, in the end, about keeping these households safe, which is why the margin, the response tiers and the consumer protections are read as one system rather than as separate problems.
Primary sources
The most load-bearing sources for how the system withstands and recovers from stress are listed below.
- Capacity Market final auction parameters and results, DESNZ and NESO. The instrument that procures availability to meet the reliability standard. https://www.gov.uk/government/publications/capacity-market-auction-parameters-letter-from-desnz-to-neso-february-2026/final-auction-parameters-t-1-and-t-4-capacity-market-auctions
- National Gas transmission operational data. Gas demand, linepack and the operational picture behind gas resilience. https://www.nationalgas.com/data-and-operations/transmission-operational-data
- NESO system and balancing reports. Operational forecasts, balancing actions and restoration reporting. https://www.neso.energy/industry-information/balancing-services/system-and-balancing-reports
- Default Tariff Cap, Ofgem, under the Domestic Gas and Electricity (Tariff Cap) Act 2018. The price ceiling that protects default-tariff consumers. https://www.ofgem.gov.uk/energy-policy-and-regulation/policy-and-regulatory-programmes/default-tariff-cap
- Clean Flexibility Roadmap, DESNZ and Ofgem, 2025. Targets a large rise in flexible capacity by 2030 to complement the margin. https://www.gov.uk/government/publications/clean-flexibility-roadmap/clean-flexibility-roadmap
- Annual Fuel Poverty Statistics in England, 2026, DESNZ. Estimates 9.4 percent of households fuel poor. https://assets.publishing.service.gov.uk/media/69c3af123ed0546101e0dc3e/Main_Report__2026_Fuel_Poverty_Statistics_Publication_.pdf
- NESO Winter Outlook. The annual assessment of the electricity capacity margin and the Loss of Load Expectation. https://www.neso.energy/publications/winter-outlook
- Electricity Act 1989, section 96, with the Energy Act 1976 reserve powers. The emergency powers behind the Electricity Supply Emergency Code and rota disconnection: section 96 lets the Secretary of State direct measures to mitigate a civil emergency disrupting electricity supplies, and the Energy Act 1976 reserve powers allow the use of electricity to be restricted. https://www.legislation.gov.uk/ukpga/1989/29/section/96
- NCSC Cyber Assessment Framework. The framework energy operators of essential services are assessed against. https://www.ncsc.gov.uk/collection/cyber-assessment-framework
- Network and Information Systems Regulations 2018, SI 2018/506. Security and incident-reporting duties on operators of essential services, including energy. https://www.legislation.gov.uk/uksi/2018/506/contents
- Ofgem Consumer Vulnerability Strategy. Protections for consumers in vulnerable circumstances, including the Priority Services Register. https://www.ofgem.gov.uk/publications/consumer-vulnerability-strategy-2025
The Capacity Market Rules (SI 2014/2043), the Electricity Supply Emergency Code, and NESO's restoration (black start) contracts are the named operational instruments referenced inline.