Grid Services & Revenue Stacking for UK Commercial Sites

Why this use case matters for UK businesses

Grid revenue sits on top of the bill savings, not instead of them — the predictable part is explained in our DUoS charges guide.

A commercial battery project should stand on bill savings first: peak-shaving, energy arbitrage and lifting solar self-consumption. Grid services are the layer on top. For a larger site that can spare some of its battery’s capacity, the electricity system will pay you to keep that capacity available and to respond when the grid is short. Done properly, this can add £20,000 to £100,000 or more per year to a single asset.

This is revenue, not a grant. National Energy System Operator (NESO) and the wider market buy flexibility because the grid needs fast-responding assets to balance an increasingly renewable system. A battery that can charge and discharge in seconds is well suited to that job. But the income is site-specific, it depends on which markets you can access and how you bid, and it is never guaranteed. Prices in frequency response and the Balancing Mechanism move with supply and demand, and a service that is lucrative one quarter can soften the next. Which schemes exist, with their minimum sizes, is set out on battery storage grants and funding. Where grid services sit among the other jobs a battery can do is set out in our commercial battery storage buying guide.

The honest framing is this: build the project so it stands up on bill savings, then treat grid revenue as upside that shortens payback rather than the reason to invest. A system designed only for trading income, with no on-site benefit, is exposed if those markets cool. Stacking grid revenue onto solid demand-side savings is what can shorten a larger asset’s payback. See our cost breakdown for how the capital figures work, and where grid-flexibility and other funding routes sit alongside them.

How the battery does it

The markets are set out one by one in our guide to which grid revenue markets a business battery can enter; this section is the mechanism.

Everything starts with your half-hourly meter data. A full year of half-hourly import and export shows your demand shape, your existing peak-shaving and arbitrage headroom, and how much battery capacity is left once those on-site jobs are done. Grid services only use the capacity your site is not already calling on.

That spare capacity is then made available to the market through an aggregator, sometimes called an optimiser or route-to-market provider. The aggregator pools your battery with others, forecasts prices, and bids your asset into the most valuable market each settlement period. In practice the same battery moves between jobs through the day:

• Capacity Market. You hold capacity available to deliver during system stress events and get paid an annual fee per kilowatt of de-rated capacity, won in a competitive auction. It is steady, contracted income that underpins the stack.
• Frequency response. Fast services such as Dynamic Containment, Dynamic Moderation and Dynamic Regulation pay the battery to nudge its charge or discharge up and down within seconds to hold grid frequency near 50 Hz. This rewards the battery’s speed and is often the highest-value daily service.
• Balancing Mechanism. Through a P415-registered aggregator, your battery can be dispatched as a Balancing Mechanism Unit, with NESO buying or selling power from it in real time to balance the system.
• Demand Flexibility Service (DFS). During tight winter periods, you can be paid to shift or shed load, which your battery does by discharging instead of importing.

The optimiser decides minute by minute whether the battery earns more by shaving your own peak, arbitraging the wholesale spread, or sitting in a frequency-response contract. You keep the bill savings; the grid revenue is shared with the aggregator under your route-to-market contract. If you also run energy arbitrage or peak-shaving, those streams are scheduled around the grid-service commitments rather than competing with them.

Sizing and economics

Grid services reward both how much power you can move and for how long, so the battery is sized on two axes. Quote the power (kW) and the energy (kWh) separately, because the ratio between them sets how long the battery can hold a service: see kW vs kWh. Round-trip efficiency and cycle warranties vary by maker and are stated on different bases, and a battery that trades hard every day makes the cycle limit matter more: see what eight makers publish on warranties.

Power rating drives access to fast frequency services and sets your Capacity Market and Balancing Mechanism bid size. Energy capacity determines how long you can sustain a response and how much arbitrage and peak-shaving you can do alongside. A 1 MW / 2 MWh asset can hold a meaningful frequency-response position and still cover the site’s peaks; a smaller 250 kW / 500 kWh battery participates but earns proportionally less.

Cost follows the published UK bands on the cost page, at around £140 to £240 per kWh for containerised systems of 500 kWh and above. Cost is driven by power electronics (inverters scale with kW), cell count (kWh), the DNO connection and any export limitation or Active Network Management equipment, switchgear, and metering to settlement standard. A battery is plant and machinery, and the Annual Investment Allowance covers the first £1m a year; above that, a company claims the 50% first-year allowance on special-rate spend, which is how this site treats a battery, as our tax guide explains. Commercial battery storage is standard-rated at 20% VAT, recoverable in full by a VAT-registered business — the 0% relief applies to domestic and charitable buildings only. With bill savings plus stacked grid revenue, typical annual benefit lands at £60,000 to £300,000, for a simple payback near six years.

A worked illustration

A worked illustration, not a customer result: grid revenue depends on delivery, so a business case that needs it to be high is fragile. Build the case on bill savings first and treat revenue as the upside. To test your own site, send us your details and we reply by email.

Compliance and grid connection

A grid-services battery in this size band is a G99 connection, submitted to your DNO. The G99 process can take a long time, so the application should go in early — it is often the longest item on the programme. The DNO may impose an export limitation scheme or require an Active Network Management connection that curtails export when the local network is constrained; the design has to follow whatever terms come back, and the inverters are configured accordingly. Trading still works behind an export limit because the most valuable services use the battery’s response capability, not unlimited export.

On top of the connection, market access carries its own requirements. The asset needs metering to settlement standard, Capacity Market pre-qualification before any auction, and registration as a P415-compliant Balancing Mechanism Unit through your aggregator. The route-to-market contract sets out revenue shares, dispatch rights and minimum availability.

Everything is engineered to the IET Code of Practice for Electrical Energy Storage Systems and to BS EN/IEC 62933 and IEC 62619. That means proper fire, thermal, detection and separation design around the LFP enclosure, full CDM 2015 management on site, and your insurer engaged before energisation rather than after. Before you sign, confirm the installer quoting for the work holds MCS certification for battery storage plus NICEIC, RECC or TrustMark registration, and that the workmanship warranty is insurance-backed rather than a company promise.

Two gates that decide whether any of this is available to you

Most writing about revenue stacking describes grid-scale assets — a standalone battery on its own connection, built to trade. If your battery sits behind your meter, on your existing supply, two constraints decide which of these revenue streams you can actually reach. Check both before you model a penny of income, because either one can close a stream off entirely.

Your export limit. A behind-the-meter battery is connected under a G99 agreement with your DNO, and many commercial connections are granted with export limitation applied under G100 — the battery is allowed to discharge to serve your own load, but the amount it may push back onto the network is capped, sometimes at zero. That distinction is decisive. Services that pay you to reduce import — peak shaving against red-band DUoS, avoided available-capacity charges, and demand-turn-down products — work perfectly well on an export-limited site. Services that require you to export power on instruction do not: if the connection will not let the energy out, the asset cannot deliver against the instruction, however well specified it is. Read your connection agreement for the export limit before you read anyone’s revenue projection, and if you intend to trade, treat lifting that limit as part of the project rather than an afterthought — a variation to an existing G99 agreement takes time and may trigger reinforcement costs.

Your warranty’s throughput allowance. Stacking revenue means cycling the battery harder, and a commercial battery warranty is not open-ended. Cover is typically expressed as a total energy throughput (an MWh figure over the term) or an equivalent full-cycle count, alongside a retained- capacity guarantee — commonly a floor the cell must still meet at end of term. Standard behind-the- meter duty of roughly one cycle a day sits comfortably inside that. A stacked strategy that adds arbitrage and frequency response on top can push you towards two or more cycles a day, and at that rate the throughput allowance is consumed in a fraction of the time. The revenue may still be worth it — often it is — but the calculation is not simply the extra income. It is the extra income minus the accelerated degradation, and minus the risk of trading your way out of warranty cover in year six. Ask your supplier to state the warranty in MWh of throughput rather than years, then divide it by the throughput your proposed strategy actually implies. If nobody will put that arithmetic in writing, that is your answer about the strategy.

Neither of these is a reason to avoid revenue stacking. They are the two questions that separate a revenue model that survives contact with your site from one that reads well in a proposal. Note too that the rates themselves — Capacity Market clearing prices, frequency-response and Balancing Mechanism revenues — move continuously with each auction and settlement period, so treat any per-MW figure you find online, including in the sources ranking above this page, as a snapshot rather than a forecast.

Who it suits — and who it doesn’t

This use case suits larger sites with a genuine half-hourly supply, a sizeable agreed import capacity, and demand patterns that leave room for the battery to take on grid work — typically manufacturing, cold storage, logistics, data and large commercial estates. It works best where there is already a solid bill-saving case from peak-shaving or arbitrage, because grid revenue then sharpens an already-sound investment. Sites with on-site PV can layer this on top of solar self-consumption as well.

It does not suit smaller sites where the battery’s whole capacity is consumed by on-site savings, leaving nothing to offer the grid. It is the wrong fit for any business that wants guaranteed, fixed returns: frequency-response and Balancing Mechanism income is variable by design, and we will not pretend otherwise. If grid revenue is the only thing making a project work, the project is too fragile, and a smaller, demand-led system is the sounder case.

Frequently asked questions

Is grid-services revenue guaranteed? No. The Capacity Market fee is contracted once you win an auction, but frequency response and Balancing Mechanism income vary with live market prices and dispatch. The sound approach is to test the case on bill savings first and treat grid revenue as upside.

Do I need an aggregator? For practical purposes, yes. Markets such as the Balancing Mechanism require P415 registration and settlement-standard metering, and bidding into frequency response daily needs a 24/7 optimiser. An aggregator handles registration, forecasting, bidding and dispatch under your route-to-market contract.

Can I still earn grid revenue if the DNO limits my export? Yes. The highest-value services reward the battery’s fast response and its ability to vary import, not unlimited export, so an export limitation or Active Network Management connection still leaves a strong stack of Capacity Market, frequency response and demand-side income intact.

What revenue stacking means for a UK commercial battery

Revenue stacking means one battery earning from several income streams rather than a single one: on-site bill savings plus payments from the electricity system. A UK behind-the-meter battery can shave site peaks, shift load, hold a Capacity Market agreement and sit in frequency response. Those streams are layered across the day, not all delivered in the same second.

The stack has two halves. The demand-side half is yours alone and needs no market access: avoiding red-band DUoS half-hours through peak shaving, load shifting on a time-of-use tariff, and lifting solar self-consumption. That half alone reduces the bill before any market revenue. The Triad regime has ended and network charging is now banded, so the saving comes from avoiding red half-hours rather than three winter peaks.

What cannot be stacked is the same kilowatt in two places at once. If 400 kW is committed to Dynamic Containment for a four-hour window, it is not available to cut your own peak in that window and cannot also be bid into the Balancing Mechanism. Energy is the second limit: a 250 kW / 500 kWh system is a two-hour asset, so a deep site peak and a long market call cannot both be served from one charge. The optimiser picks the highest-value job each settlement period, which is why stacking is a scheduling exercise rather than simple addition.

Stacking can shorten payback, but the outcome is site-specific and never guaranteed. Only the Capacity Market fee is contracted once won; frequency response and Balancing Mechanism prices move with the market. This site is independent of suppliers and manufacturers, so it favours no route to market. Check the demand-side case against real commercial battery storage cost figures first, then check twelve months of your own half-hourly data to see which streams your site can reach. Send us your details and we reply by email.

See also

• Nottinghamshire industrial sites
• South Yorkshire process loads