Market Overview
The energy transition in Great Britain is well under way. Fossil fuels are on the out, and each year renewables make up a larger proportion of the power mix; in 2025, 44% of Great Britain's electricity was generated using renewables. Renewables are of course cheap and clean, but one consequence of a more renewables-focused power mix is that there is less system inertia, the physical flywheel effect that is used to resist sudden changes in frequency.
Grid frequency has to stay within a whisker of 50 Hz. Matching supply to demand, all else equal, is slightly harder when there are less heavy turbines to provide inertia on the supply-side of the grid. Less inertia means things move faster when something breaks, and requires faster responses to avert disaster. This problem is perfectly suited for batteries: they can go from idle to full output in under a second, in either direction. That capability is what the frequency response markets buy and why batteries' roles in the grids of the future will continue to grow.
The shift that created the market
The generation mix is the clearest way to see why these markets exist and why they have grown. Wind and solar rise; gas and coal fall; the system carries less and less inertia.
The relative mix matters because it shapes the underlying risk of frequency deviation and thus the opportunity for BESS sites to step in: high wind with low demand pushes frequency high, which the High-side services answer by charging, while low wind with high demand can cause dips, which the Low-side services answer by discharging. The High and Low in DCH, DCL and the rest name the frequency excursion being corrected, not the direction the battery moves power.
This plot shows a 28-day rolling mean, smoothing day-to-day noise while still capturing seasonal swings.
Average share by fuel group
How batteries capitalise
A grid-scale battery in GB has three main routes to revenue: frequency response (ancillary services), wholsale arbitrage, and capacity markets. As capacity markets are longer-horizon auctions (one or four years out from delivery) with more site-specifc physical limitations, this page covers the market for both the shorter term markets: frequency response and wholesale arbitrage.
Frequency response: contracted availability. NESO runs daily auctions for capacity that must react within seconds when frequency strays from 50 Hz. Win one and you are paid a £/MW/h availability fee for every hour you are committed, whether or not you are actually called. Predictable, contracted income — but the committed capacity has to keep enough charge and enough headroom to deliver in either direction.
There are three services, split by how fast and how long they must respond:
| Service | Frequency band | Response | Sustained for |
|---|---|---|---|
| DC — Dynamic Containment | ±0.2–0.5 Hz | ~1 second | 15 min |
| DR — Dynamic Regulation | ±0.015–0.2 Hz | continuous | 60 min |
| DM — Dynamic Moderation | ±0.1–0.2 Hz | ~1 second | 30 min |
Each runs as two separate auctions: High, which responds to rising frequency by charging, and Low, which responds to falling frequency by discharging — the name is the frequency excursion being corrected, not the direction the battery moves power. Auctions clear per EFA block — six four-hour windows covering the day — so a battery's commitment can differ across the day.
Wholesale arbitrage: opportunistic trading. Separately, the battery can buy energy when it is cheap and sell when it is expensive. The profit is the price spread less round-trip losses and the wear cost of cycling.
The tension between the two is the subject of the Forecasting & Dispatch page: capacity committed to frequency response cannot be freely traded, so the operator must decide each day how to split it.
Frequency Response
GB frequency response is procured through three dynamic services, each split into High (charge — activated when frequency rises above 50 Hz) and Low (discharge — activated when frequency falls below 50 Hz) auctions.
| Service | Frequency band | Role |
|---|---|---|
| DC – Dynamic Containment | ±0.2–0.5 Hz | Arrests large deviations within ~1 second |
| DR – Dynamic Regulation | ±0.015–0.2 Hz | Maintains frequency in normal operation |
| DM – Dynamic Moderation | ±0.1–0.2 Hz | Moderates frequency during stressed conditions |
Auctions run daily for each EFA block (six 4-hour windows covering the full day). The clearing price is the marginal accepted bid for that block and service.
EFA block timings
EFA Block 1 spans midnight (23:00 the previous calendar day to 03:00). All times are local GB time.
Clearing prices — 28-day rolling average by service
Individual auction results are first averaged to a daily figure per service, then smoothed with a 28-day rolling window, so the trend for each of the six services is readable without daily noise obscuring the signal.
Key takeaways — clearing price trends
- 2022 peak then sharp compression. DCL clearing prices peaked at £15–20/MW/h in 2022 as NESO expanded DC procurement ahead of renewable growth. From late 2022 a rapid wave of new GB BESS capacity entered the frequency response markets, outpacing NESO's procurement volumes and driving prices steeply lower across all services.
- Discharge (Low) services generally clear above charge (High) services. Fleet-wide charge headroom tends to be more available than discharge headroom — particularly during high-wind periods — so High-side auctions typically clear lower.
- DRH and DRL behave differently from DC and DM. DR's sustained 60-minute delivery requirement couples the two sides operationally, which is why the DRL spread sometimes inverts relative to DCL and DML.
Price distribution
DCL shows the widest spread of outcomes, reflecting its role as the primary fast-discharge service and its early-market dominance at elevated prices. Evening blocks (EFA 5–6, 15:00–23:00) attract higher premia as demand peaks and wind output often eases; the overnight block (EFA 1) is typically cheapest to procure.
Summary statistics
High vs Low spread
Each service runs two separate auctions: High (rising frequency — BESS charges) and Low (falling frequency — BESS discharges). Clearing prices differ because available discharge and charge headroom across the fleet is rarely symmetric.
Spread = H clearing price − L clearing price. Positive means charge capacity was scarcer; negative means discharge capacity was scarcer. All three markets average negative, so the discharge leg is consistently the scarcer of the two.
Daily average H − L spread over time
Spread distribution by market
Average spread by EFA block
DC shows the widest range of spread outcomes and the median closest to zero, so its two legs are priced the most symmetrically of the three. DR sits firmly negative across both charts — positive in only 7% of blocks — confirming the structural inversion described above. DM occupies the middle ground. All three average negative, so discharge capacity is the scarcer side throughout. Evening blocks (EFA 5–6) show the most pronounced spreads, as demand peaks and the balance between available charge and discharge headroom is tightest.
H − L spread heatmap: EFA block × month
Each cell is the average H − L spread for that market, EFA block and calendar month. Red = charge capacity scarcer (H > L); blue = discharge capacity scarcer (L > H).
Wholesale & settlement prices
System Buy Price (SBP) and System Sell Price (SSP) are the cash-out prices used to settle imbalance in the GB Balancing Mechanism. Parties that are short pay the SBP; parties that are long receive the SSP. The gap between them incentivises self-balancing rather than relying on the system operator.
Wholesale price spread
Daily peak-to-trough APXMIDP spread — the raw arbitrage opportunity available to a battery on any given day, before efficiency losses and cycling cost.
Thin line is the daily spread; heavy line is a 28-day rolling average.