Should You Add More Solar Panels?

Yes — every extra panel lowers your electricity bill, but by less each time: in our worked example, growing a 3.7kWp array to 7kWp saves about £429 a year at a 12p export rate, but only about £168 at a 3p rate, because most of the extra electricity is exported rather than used at home.

A kWh you use yourself is worth the 26.32p you would have paid to import it. A kWh you export is worth only your export rate. The first panels mostly cover daytime use; each panel after that adds output at times when the house already has enough, so more of it goes to the grid.

Worked example: a household using 3,500kWh of electricity a year, with no battery and no electric car, at the Ofgem price cap rate of 26.32p. All panels face south.

At a 12p export rate each extra kWp saves about £134 a year going from 3.7 to 5kWp, falling to £122 from 7 to 10kWp. At 3p — the lowest large-supplier rate in Sunsave's comparison, updated 26 August 2026 — it is £57 falling to £37. The 12p rate is Outgoing Octopus, checked September 2026; your export rate decides most of the answer.

How the share used at home is worked out. Our other guides use industry estimates — 30% to 50% of solar used at home without a battery, around 70% with one — but those are single figures that don't change with array size, so they can't show what a bigger array does. Here we use the method in the government's SAP 10.2 energy rating, where the share falls as generation grows relative to what the home uses, month by month. It was fitted to monitoring data from 33 homes (BRE, SAP 10 Technical Paper S10TP-07). It gives lower self-use than the industry estimates: 33% rather than 40% for the 3.7kWp array. Both are estimates, not measurements, and both become less reliable the further an array's output runs ahead of the household's use. There is also a gap between our figures and the paper's own worked example, set out under the battery question below.

Generation comes from the European Commission's PVGIS model for a south-facing 3.7kWp array near Derby, scaled to each size — see how solar works on a new build.

For a household using about 3,500kWh a year, more panels make a home battery only slightly more worthwhile — its saving rises from about £182 to £212 a year between 3.7kWp and 10kWp — but in a smaller household the battery saves less as panels are added, because the roof covers more of the electricity the battery would otherwise replace.

A battery is pulled two ways by array size. With a small array it is short of surplus solar to store. With a large one, the panels cover more of the home's daytime use directly, leaving less imported electricity for the battery to replace. A 5kWh battery can deliver at most 1,825kWh a year if it fills and empties once a day.

Why the ceiling doesn't bite in this household: even with 10kWp of panels, the home still imports about 1,893kWh a year — evenings, nights and winter days the panels can't reach — which is more than the battery's 1,825kWh. The battery keeps finding electricity to replace. Its saving tops out at about £212 a year at around 10kWp. Without a car the household uses a standard tariff with no battery and Economy 7 with one, whichever is cheaper.

On a standard variable tariff, where the battery can only store solar, the two pulls are clearer: the saving rises from about £108 at 3.7kWp to a peak of about £135 at 7.3kWp, then falls to £113 at 12kWp.

In a smaller household the ceiling arrives immediately. For a home using 2,000kWh a year, the battery saves about £152 at 2kWp, £137 at 3.7kWp, £130 at 5kWp, £125 at 7kWp, £122 at 10kWp. Every extra panel leaves the battery a little less to do.

With an electric car on an EV off-peak tariff, array size makes no difference in this example: the battery fills from the cheap overnight rate, and even at 10kWp the home still imports more than the battery can cover. See how much a battery saves for why.

These battery figures use the SAP 10.2 self-use method, so they are lower than in our battery guides, which use the industry estimates: at 3.7kWp without a car, about £182 a year here against £201 there. The split between storing solar and charging off-peak is indicative, because the one-cycle-a-day limit is applied over the whole year.

Where these figures may be optimistic. We check our version of the SAP method against the two examples printed in the technical paper that derived it. Its 1kWp example reproduces closely — we get 0.54, 0.84, 0.96 for no battery, a 5kWh battery and a 10kWh battery, against its 0.54, 0.83, 0.96. Its larger example does not: with a 5kWh battery we get 0.71 where the paper prints 0.65.

The paper's three figures for that example can't all be right together: matching its no-battery figure of 0.40 leaves its 5kWh case at 0.69, not 0.65. So part of that example looks approximate. But if the paper's figure is the better one and ours is high by 0.06, this page is optimistic in one specific place: the share of solar a battery captures, and with it the "from storing solar" column above. That column would be about 23% lower — £83 instead of £108 at 3.7kWp. The total battery saving would fall by less than that, because a battery holding less solar has more room to charge off-peak instead. Nothing else on this page depends on it: the off-peak column, the electric-car figures and the value of the panels themselves are unaffected.

The grid connection is usually the first obstacle: a home on a single-phase supply can connect up to 3.68kW of generation and notify the network operator afterwards under the G98 rules, but anything more — including a battery with its own inverter — needs the network operator's assessment under G99 before it is installed.

G98 and G99. Under Engineering Recommendation G98, the limit is 16A per phase — 3.68kW on a single-phase supply, 11.04kW on three-phase — and the installer notifies the network operator within 28 days of commissioning. The limit counts every generator and battery at the property together, and for inverter-connected systems it is the lesser of the inverter rating and the panels' rating. Above it, G99 applies and the network operator must be consulted before installation. It can refuse, set conditions such as an export limit, or require network reinforcement that may be chargeable (Energy Networks Association, EREC G98 and G99 Issue 2, G99).

One G99 route is designed for homes: total capacity under 32A per phase (about 7.36kW) with an approved export limitation scheme holding export to 16A. The network operator replies within 10 working days, and nothing may be installed before then. Outside that route, SSEN says a quotation can take up to 45 working days (SSEN). An export limit also means some output on sunny middays can't be exported at all; our figures don't allow for that, so they flatter a large array under an export limit.

The inverter. The developer's inverter was chosen for the developer's array. Panels beyond its rating add little at peak times, because the inverter can't pass the extra through. The alternatives are a larger replacement inverter or a second inverter, and either counts towards the 3.68kW limit. If you already have a battery on its own inverter, the property may already be past it.

The roof. The developer will normally have used the best-facing roof. Extra panels often go somewhere worse. PVGIS gives the same system facing east 79% of the south-facing output, facing west 76%, and facing north 50%. Adding 3.3kWp on a west-facing slope instead of south saves about £332 a year rather than £429 at a 12p export rate; on a north-facing slope, about £218. Shading from chimneys, dormers and neighbouring houses reduces it further.

Planning. In England, roof panels on a house are usually permitted development, but not if they would stick out more than 0.2 metres from the roof slope, sit higher than the highest part of the roof, or go on a listed building, and there are extra limits in conservation areas (General Permitted Development Order, Part 14). Planning permissions for new estates can remove permitted development rights by condition, although government guidance says that needs clear justification — check your home's decision notice before you commit.

Building regulations and warranties. The Planning Portal says building regulations will normally apply: the roof's ability to carry the extra weight must be checked, and the electrical work must comply too. An installer registered with a competent person scheme for microgeneration can self-certify it. The MCS solar installation standard says the roof warranty provider should be consulted where the roof covering is under warranty — on a new build, ask your developer and warranty provider before anyone drills into the roof.

There is no published price for adding panels to an existing system, but government data from MCS records puts small domestic retrofit installations — which include extensions — at an average of about £2,109 per kWp in the year to March 2026, so adding 3.3kWp would cost roughly £7,000.

That average comes from DESNZ solar PV cost data (published 28 May 2026, checked September 2026). It covers the equipment, installation, grid connection and VAT where applicable, across 60,713 domestic retrofit installations of up to 4kW. It is an average, not a quote. For comparison, installations on new builds averaged £1,518 per kWp, but that price isn't on offer once the house is finished. A small addition carries the same fixed costs as a bigger job, such as scaffolding, connection and paperwork, so it may cost more per kWp than the average.

What it pays back: at about £6,960 for 3.3kWp, rounded up to whole years:

• South-facing, 12p export: saves £429 a year — about 17 years.
• West-facing, 12p export: saves £332 a year — about 21 years.
• South-facing, 3p export: saves £168 a year — about 42 years.
• West-facing, 3p export: saves £134 a year — about 52 years.

These use today's prices and assume the full extra output can be exported; an export limit under G99 would lengthen them.

You can keep getting Smart Export Guarantee payments after adding panels — Ofgem’s guidance says you can install additional capacity to an eligible installation — but you must tell your export supplier as soon as reasonably possible, and the new panels need to be certified.

Ofgem requires export contracts to include a term obliging you to tell the supplier about extensions (Ofgem, SEG guidance for generators). Octopus Energy's export terms, for example, say you must let them know as soon as reasonably possible if you install additional capacity, and accept MCS, Flexi-Orb or an equivalent certification scheme.

MCS doesn't publish a rule on whether an extension gets its own certificate or one covering the whole system. Ask the installer, before work starts, what certificate you will receive, and ask your export supplier what they need to see.

If the developer's panels were never MCS certified, nothing in MCS or Ofgem guidance says that adding certified panels makes the original system eligible. Sort out the existing system first — see MCS certification explained.

Adding solar panels is a bad idea when your export rate is low, the only roof space left faces north or is shaded, the extra capacity would need G99 approval with an export limit, your household uses little electricity, or the existing panels aren’t yet earning export payments.

In each case most of the extra output is exported cheaply, clipped by the inverter or export limit, or never generated. At a 3p export rate, 3.3kWp of south-facing panels in our example takes about 42 years to pay for itself.

It is also a weak case if you plan to move within a few years, or if the addition would take the property past the 3.68kW G98 limit on a supply where the network operator is likely to set conditions.

For most new build owners, neither pays for itself within the 10 years of a typical battery warranty, and your export rate decides which is better: at 12p, adding panels to a good roof saves more than a battery (about £429 a year against £182 without a car), but at 3p the battery saves more (about £266 against £168) — and a worse roof, a G99 export limit or a smaller household all tip the balance away from more panels.

Side by side, for the 3,500kWh household with the developer's 3.7kWp array, payback rounded up to whole years:

• Add 3.3kWp of panels (about £6,960): 17 years south-facing and 21 years west-facing at 12p export; 42 and 52 years at 3p.
• Add a 5kWh battery (£3,500–£5,500): 20 to 31 years without a car at 12p export; 14 to 21 years at 3p; 12 to 18 years with an electric car on an EV off-peak tariff.

Every one of those is longer than a typical 10-year battery warranty. On a good south-facing roof with a 12p export rate and no grid hurdle, extra panels are the better bet on these numbers. Change any of those — a lower export rate, a west or north slope, an export limit, a small household — and they aren't. The battery case is strongest with an electric car.

The honest order of priorities: first make sure the panels you already have are earning — an MCS certificate and a good export tariff are worth more than any addition. Then, if you still want more, get the grid connection question answered and a written quote before you compare it with a battery. Past the developer's array, the right answer for many homes is not to add panels at all.

Try your own numbers in our electricity and battery calculator, which lets you change the array size.