Commercial Battery Storage UK 2026: How to Size a BESS

Quick answer

Sizing a commercial battery correctly is worth more than any other decision in the project — get it wrong and you double your payback. The rule is simple: size to the job, not to your roof or your budget. A peak-shaving battery is sized in kW (the demand spike you need to shave); an arbitrage battery is sized in kWh (the energy you shift overnight from cheap to expensive periods); a back-up battery is sized to your critical load and runtime. Most UK SME systems land between 30kWh and 215kWh, pay back in 4 to 9 years, and need a G99 application to the DNO before connection — a process that takes 6 to 16 weeks and is usually the project's critical path, not the install. Below we work through every sizing method, the G99 and DNO gates, ROI scenarios, and how a battery earns grid-services revenue on top of bill savings.

The three jobs a commercial battery does

Before you size anything, decide what the battery is *for*. The three core use cases need completely different specifications, and most real-world systems blend two of them:

1. Peak-shaving — discharging the battery during your demand peaks to cut the kW your site draws from the grid. This reduces capacity (kVA) charges and avoids triggering higher DUoS red-band costs. Sized in kW of discharge power.

2. Arbitrage (load-shifting) — charging the battery overnight on a cheap rate (often 7-15p/kWh) and discharging during the expensive working day (22-30p/kWh). The profit is the spread × the energy shifted. Sized in kWh of usable capacity.

3. Back-up power — keeping critical loads running through a grid outage. Sized to your critical load (kW) × required runtime (hours), with an inverter capable of forming an island grid.

A fourth layer — grid-services revenue — sits on top of all three: a battery that is not fully committed to the above can be traded into frequency-response and balancing markets by an aggregator. This is income, not saving, and it changes the sizing maths because the optimiser wants headroom to trade.

The single biggest costly mistake we see in tender documents is a battery sized to "fill the solar roof" or to a round-number budget, with no reference to the site's half-hourly demand profile. That nearly always over- or under-sizes the asset.

How to size a BESS: peak-shaving vs arbitrage vs back-up

Each job is sized from a different number in your data. Here is the comparison that drives every Ecoaim feasibility study:

A worked sizing example: a light-industrial unit drawing a 90kW peak for two hours each afternoon, on a 28p/kWh day rate with a 9p/kWh overnight rate. For arbitrage you size to the energy you can shift — roughly 120kWh charged overnight at 9p and discharged at 28p, a 19p spread worth about £22.80 per full cycle, or ~£6,800/year over 300 cycles. For peak-shaving the same battery's 90kW+ discharge capability cuts the site's billed capacity, saving on kVA charges. The asset does both jobs from one investment — which is why the sizing has to model the load profile, not a single use case in isolation.

G99 application gates — what you must clear before connecting

Any commercial battery exporting or capable of exporting to the grid needs a connection agreement under Engineering Recommendation G99. The threshold is low: anything above 3.68kW per phase — which is effectively every commercial system — falls under G99.

There are two routes:

• G99 Fast Track — for systems up to 50kW (three-phase). A streamlined application with a decision typically inside 45 working days.
• Full G99 application — for systems above 50kW. The DNO assesses network capacity and may return a connection offer that includes reinforcement costs and conditions such as export limitation.

You can reduce the burden with an export limitation device (G100), which caps grid export so the DNO assesses a smaller connection — useful where the battery is for self-consumption and peak-shaving rather than export. The official framework sits with the Energy Networks Association G99 standard and your regional DNO. Ecoaim prepares and submits the full G99 (or Fast Track) package in-house, including the witness-test and commissioning documentation the DNO requires.

The practical point: the G99 process is a gate, not a formality. Start it early. We have seen otherwise-ready projects sit idle for two months waiting on a connection offer because the application went in late.

DNO connection timelines — plan for the critical path

The Distribution Network Operator (DNO) timeline is almost always the longest single item in a commercial battery project — longer than procurement or installation. Indicative 2026 timelines:

For a sub-50kW Fast Track system you can reasonably plan on 8-10 weeks from application to energised. For a larger system requiring a full application, 3-5 months is realistic, and longer if the DNO identifies the need for network reinforcement. Reinforcement is the wildcard — it is uncommon for moderate SME systems but can dominate the timeline for large arrays on constrained networks. The lesson is to submit the G99 application *before* you finalise procurement, so the DNO clock runs in parallel with everything else.

Commercial battery ROI scenarios

The economics depend on what jobs the battery does and whether it sits alongside solar. Three representative scenarios for a UK SME, using a 30p day / 9p night spread and 26p avoided import:

The standalone arbitrage-only case is the weakest — it relies purely on the tariff spread and capacity savings. The moment you pair the battery with solar (capturing surplus that would otherwise export at 4-15p and redeploying it at 26p) or stack grid-services income on top, payback drops into the 4-5 year range. Note the 100% Annual Investment Allowance in the final column: solar and storage qualify as plant and machinery, so the full cost is written off against taxable profit in year one — at 25% corporation tax that shortens net payback by roughly a year. Confirm the tax treatment with your accountant, as it depends on whether the asset is bought outright or leased.

If you are already considering panels, the case for adding storage is strongest — see our commercial battery storage feasibility process and our commercial solar payback worked examples for how the two assets combine.

Grid-services revenue — turning a cost centre into income

A battery sitting idle between peaks is wasted capital. Through an aggregator or optimiser, that spare capacity can be traded into the grid-services markets operated by National Grid ESO and accessed via Ofgem-regulated routes:

• Dynamic Containment / Dynamic Regulation / Dynamic Moderation — fast frequency-response services that pay for keeping the grid at 50Hz. The highest-value services for batteries.
• Balancing Mechanism — being dispatched to balance supply and demand in real time.
• Capacity Market — annual payments for guaranteeing availability at system-stress periods.
• Wholesale arbitrage — the optimiser trades the day-ahead and intraday markets on top of your own load-shifting.

Stacked, these can add £30 to £80 per kW per year, depending on the asset, the contract and how much headroom you leave the optimiser. The framework for these markets and the SEG export side sits with Ofgem. The trade-off: committing capacity to grid-services means less is available for your own peak-shaving, so the sizing has to balance self-use against income. This is exactly the kind of optimisation a proper feasibility study resolves — and it is why we always model self-consumption, arbitrage and grid-services together rather than in isolation.

How Ecoaim sizes and delivers your BESS

Every Ecoaim commercial battery project starts with your half-hourly consumption data — the only reliable basis for sizing. Our process:

• Load-profile analysis — we pull your HH data and model peaks, overnight idle windows, and the tariff spread you are actually on
• Use-case modelling — peak-shaving, arbitrage, back-up and grid-services modelled separately and stacked, with NPV for each
• Sizing recommendation — the kW/kWh that delivers fastest payback against *your* profile, not a roof-fill estimate
• G99 handled in-house — Fast Track or full application, export limitation where it helps the connection, witness testing and commissioning docs
• Funding routes — CapEx (with 100% AIA), asset finance/lease, or zero-capex PPA, modelled side by side
• MCS-certified install from our Livingston base across the Scottish Central Belt

We size the asset to do the job you need, clear the DNO gate early so it is never the bottleneck, and structure the funding so the numbers work on your balance sheet.

Bottom line

Commercial battery storage is a sizing problem first and an installation problem second. Size to the job — kW for peak-shaving, kWh for arbitrage, critical-load × runtime for back-up — using your real half-hourly data, never your roof area or a round-number budget. Clear the G99 application early because the DNO timeline (6-16 weeks, longer with reinforcement) is the project's critical path. Expect 4 to 9 year payback depending on whether the battery is standalone, solar-paired or stacked with grid-services income — and remember the 100% Annual Investment Allowance shortens net payback by about a year. For a system sized against your actual consumption, our commercial battery storage team provides a fixed-price feasibility study with worked ROI before you commit a penny.

Related Ecoaim guides:

• 🔋 Commercial battery storage — full guide
• 🏭 Commercial solar payback UK 2026: 100kWp to 500kWp worked
• 🏭 Commercial solar panels — full guide
• 💷 Green mortgages & business finance
• 🛠️ Postcode ROI calculator