Commercial battery storage ROI — a worked example for UK SMEs

Our battery storage guide makes the qualitative case: a commercial battery only earns its keep when it's doing a specific job — shifting solar into the evening, shaving demand charges, exploiting a time-of-use spread, or keeping you running through an outage. That guide tells you whether to look at storage. This one shows you how the numbers actually work once you do.

Below is a full worked example — capex in, revenue out, payback at the bottom. The figures are illustrative, chosen to be representative of a mid-sized UK SME in 2026 rather than a quote for any real site. The point isn't the answer; it's the method. Swap in your own numbers and the structure holds.

The business in our example

A light-manufacturing unit in the Midlands, already running — or about to install — a commercial solar array:

• Existing solar: 120 kWp on a south-facing roof
• Annual generation: ~107,000 kWh (120 kWp × ~890 kWh/kWp at this location)
• Daytime self-consumption before a battery: ~72%, so roughly 30,000 kWh/year is currently exported
• Import tariff: 30p/kWh (a realistic all-in 2026 non-domestic rate once standing charges, DUoS and levies are included)
• Export tariff (SEG): 6p/kWh
• A demand (capacity) charge on the bill, because it's a half-hourly metered supply

That last point matters. This site is on a half-hourly (HH) meter, so we have real data on when power is used — which is the only honest basis for sizing a battery. If you don't have that data yet, our half-hourly data guide explains how to get it before you model anything.

Step 1 — Size the battery to the job, not the roof

The single most common way to ruin battery ROI is to oversize. A battery only earns on the energy that flows through it, so we size it to the genuine daily surplus and the evening demand it can usefully serve — not to "match" the 120 kWp array.

From the half-hourly data in this example, the site has:

• A daily exported surplus averaging ~80–100 kWh in summer, far less in winter
• An evening/early-morning load (cleaning shift, charging, refrigeration) of ~50–70 kWh that currently comes from the grid at 30p

A 100 kWh usable battery fits that shape: big enough to soak up most of the summer surplus and cover the shoulder-hour load, small enough to cycle close to fully most days. A 200 kWh unit would sit half-used for much of the year and roughly double the capex for little extra benefit.

Step 2 — The capex

Commercial battery storage in 2026 runs roughly £200–£450 per kWh installed, falling per-kWh as systems get larger. For a 100 kWh system we'll use £320/kWh:

• Battery capex: 100 kWh × £320 = £32,000 (ex VAT)

That's the headline number the rest of the example has to earn back. Note we're costing the battery alone — the solar array is a separate investment with its own, generally stronger, payback (see commercial solar payback). Mixing the two flatters the battery and hides whether it's actually pulling its weight.

Step 3 — The three revenue stacks

A commercial battery can earn from up to three sources at once. We'll build each up separately, then total them — and be honest that they overlap and compete for the same kWh, so you can't simply add the theoretical maximum of each.

Stack 1 — Self-consumption shift (the core)

Today, ~30,000 kWh/year is exported at 6p instead of being used on site at 30p. The battery captures a share of that surplus and discharges it into the evening load, converting low-value export into high-value self-consumption.

Realistically the battery round-trips about 18,000 kWh/year of that surplus (summer surplus is abundant; winter has little to store, and round-trip efficiency loses ~10%). Each shifted unit is worth the difference between the import price avoided and the export income foregone:

• Value per shifted kWh: 30p − 6p = 24p
• Stack 1 revenue: 18,000 kWh × £0.24 = £4,320/year

Stack 2 — Peak (demand) shaving

This site's bill includes a capacity / demand charge based on its highest half-hourly draw. By discharging the battery during predictable demand spikes, the metered peak drops and so do the capacity and DUoS red-band charges.

Say the battery reliably shaves 30 kW off the chargeable peak, and demand-related charges run ~£60/kW/year on this tariff:

• Stack 2 revenue: 30 kW × £60 = £1,800/year

Peak shaving is often the quietest line on the page and one of the strongest. It doesn't depend on solar at all, so it keeps earning in midwinter.

Stack 3 — Time-of-use arbitrage

On a time-of-use tariff, the battery can also charge from cheap overnight grid power and discharge at the expensive evening peak. But here it competes with Stack 1 — the same capacity can't both store midday solar and charge from the grid overnight. We therefore credit only a modest residual:

• Stack 3 revenue (residual, illustrative): £600/year

Total annual benefit

• Self-consumption shift: £4,320
• Peak shaving: £1,800
• Arbitrage (residual): £600
• Total: ~£6,720/year

Step 4 — Simple payback

• Capex: £32,000
• Annual benefit: £6,720
• Simple payback: £32,000 ÷ £6,720 ≈ 4.8 years

On its own that looks healthy. But a battery isn't a panel — it degrades, it has a warranty cliff, and the cash arrives over years. Two adjustments make the number honest.

Step 5 — The adjustments installers underplay

Degradation and cycle life. Lithium commercial batteries are typically warrantied for around 10 years and a set number of cycles (often ~6,000), retaining perhaps 70% of capacity by end of warranty. Cycling our 100 kWh unit roughly once a day is ~3,650 cycles over 10 years — comfortably inside warranty, but capacity (and therefore annual benefit) fades over time. Modelling a gentle decline rather than a flat £6,720 every year is the realistic approach.

It may need replacing before the panels do. Your solar array is warrantied ~25 years; the battery likely needs replacing at year 10–12. A 25-year view of the site must budget for a second battery (at a lower future price), not assume one unit lasts the distance.

Tax treatment — the Annual Investment Allowance. Commercial battery storage installed by a business generally qualifies for the Annual Investment Allowance (AIA) — a 100% first-year capital allowance, up to the £1m annual cap. For a company paying 25% corporation tax, that's worth up to £8,000 off the £32,000 (25% × £32,000), cutting the effective net capex to around £24,000. We cover the reliefs in full in our grants and funding guide — and you should always confirm eligibility with your accountant, as the treatment depends on your tax position and how the asset is owned.

Step 6 — The honest payback

Putting those together for our example:

• Gross simple payback: ~4.8 years
• Post-AIA simple payback (effective net capex ~£24,000 ÷ £6,720): ~3.6 years
• Allowing for capacity fade (benefit declining gently year on year): a more realistic effective payback of ~4 to 4.5 years post-tax
• Discounted payback (applying, say, a 6% cost of capital so future savings are worth less than today's): roughly 5 to 5.5 years

So the spread of defensible answers for this illustrative site is roughly four to five-and-a-half years, depending on whether you take simple or discounted, gross or post-tax. That range — not a single hero figure — is what an honest proposal should show you. A unit warrantied for ten years paying back in four-to-five with peak shaving and tax relief doing real work is a genuinely good case. The same battery on a daytime-only site with no demand charge — losing Stacks 2 and 3 and most of Stack 1 — would push past eight years and fail the test.

How to re-run this for your own site

The arithmetic is the same five lines every time:

1. Usable capacity × £/kWh = capex. Get the installed, ex-VAT figure for the size that fits your surplus and evening load.
2. Self-consumption shift: annual kWh the battery moves from export to on-site use × (import rate − export rate).
3. Peak shaving: kW shaved off your chargeable peak × your £/kW demand charge. Zero if you have no capacity charge — check your bill.
4. Arbitrage: only if you're on a time-of-use tariff and the battery isn't already full of solar — usually a small residual.
5. Adjust for AIA (cut effective capex by your tax rate × capex), capacity fade, and discounting.

Model the solar first on our calculator, then layer the battery on top using your half-hourly data. The battery should be judged on its own payback, separate from the panels.

Sanity-check, in one line each

1. Size the battery to your surplus and evening load, not the array — an oversized battery sits idle and never pencils.
2. Cost the battery standalone; don't let it hide inside the (stronger) solar payback.
3. Build revenue from three stacks — self-consumption shift, peak shaving, arbitrage — but don't double-count overlapping kWh.
4. Peak shaving can be the strongest line and is the one installers most often omit.
5. Adjust for AIA (100% first-year allowance), capacity fade, and a discounted-payback view.
6. A four-to-five-year payback on a ten-year-warranty unit is strong; eight-plus means the case is thin — defer and retrofit later.

The bottom line

Commercial battery ROI isn't mysterious, but it is easy to fudge — and the fudges nearly all run in the seller's favour. Done honestly, it's five lines: capex in; self-consumption shift, peak shaving and arbitrage out; then adjusted for tax, degradation and the time value of money. For a site with surplus solar, an evening load and a demand charge on the bill, a right-sized battery pays back in the mid-single-digit years and earns for a decade. For a daytime-only operation with no demand charge, it doesn't — and the right answer is to bank the solar savings and add storage later if your usage changes. Insist on seeing the workings, not the headline.

To model the solar case first, run the calculator. To decide whether a battery is even the right tool for your business, read battery storage for commercial solar. And to size it on real evidence, get your numbers straight with our half-hourly data guide. For monthly, plain-English intel written for finance directors, subscribe to the Brief.

General information, not financial or tax advice. The figures above are illustrative — model storage against your own half-hourly consumption data and confirm capital-allowance eligibility with your accountant.