Do solar panels work on a north-facing roof in the UK?

How much less does a north-facing roof generate?

About half to three-fifths of what the same panels would make facing south — and that is the installers’ own figure, not a sales estimate. Every MCS-certified installer has to estimate your output with the method in Appendix B of MIS 3002: kWp × Kk × shade factor, where Kk is the kWh each kWp makes in a year, looked up by postcode zone, roof pitch and orientation from the MCS Irradiance Datasets workbook. That workbook has a column for every 5° of orientation from due south, out to 175° — the nearest it goes to due north — and a row for every degree of pitch. Read directly from it (v2.0) at a 35° pitch, unshaded:

Across all 25 zones in the workbook the north figure sits between 51% of the south one (Brighton) and 59% (Kirkwall). These are AC figures — what would reach your sockets — because MIS 3002 notes the table data “is drawn from the Climate-SAF-PVGIS dataset and multiplied by 0.8”, the 0.8 being the method’s built-in allowance for real-world losses. If the arithmetic behind that is new to you, how do installers turn kWp into kWh? walks through it.

What would a 4 kWp system make facing north?

Multiply the north column by four. An unshaded 4 kWp array at 35° facing north comes to about 2,160 kWh a year in the London zone, 1,990 in the Manchester zone and 1,890 in the Glasgow zone, against roughly 3,940, 3,460 and 3,330 for the same array turned to face south. Put another way, a north-facing 4 kWp system makes about what a south-facing 2.2–2.3 kWp system would in the same zone. That is the honest shape of the answer: the panels are not wasted, but you would need nearly twice as many of them to get the energy a south roof gives you — and how many you can fit is a fact about the roof, not a choice about the panels.

Does the pitch of a north-facing roof matter?

More than anything else about it, and in the opposite direction to a south roof. A south-facing roof does best at around 35–40°; a north-facing one does best as flat as possible, because the steeper it is the more it turns its back on the sun. From the London zone’s table, comparing north with south at the same pitch:

The other zones follow the same curve: across all 25, a north face at 10° makes 81–85% of south, at 20° 67–73%, at 45° 43–52% and at 60° 34–45%. So a low-pitched north-facing extension roof and a steep north-facing gable are different propositions, and the difference is bigger than the difference between Brighton and Lerwick. MIS 3002 defines a pitched roof as one “with an angle greater than 10° but less than 70° to the horizontal”, requires the installer to record inclination “to the nearest 1°” and orientation “to the nearest 5°”, and puts both on the estimate you are given — so if your quote says “north-facing” without a pitch, it has left out the number that matters most.

Is a flat roof better than a north-facing one?

Yes, in every zone and at every pitch. The workbook’s flat-roof figure (0° pitch, where orientation makes no difference) is 81–86% of the south-facing 35° figure across all 25 zones — 828 kWh per kWp in the London zone, against 540 for north at 35° and 745 for north at 10°. There is no pitch in the table at which a north-facing roof catches up with a flat one. And a real flat-roof installation usually does better than the table’s flat figure, because the panels go on frames tilted towards the south, and the MCS method then uses the frame’s pitch and orientation rather than the roof’s. One caution on this site’s own estimate: Google’s Solar API methodology states that panels are “assumed to be mounted flush with the roof, including on flat surfaces”, so the roof card’s figure for a flat roof is the flush one, and an installer’s framed design would beat it.

Is east or west better than north?

Much better, and closer to south than to north. At 35° the workbook puts an east- or west-facing roof at 78–82% of a south-facing one across all 25 zones — in line with the Energy Saving Trust’s line that “a system facing east or west tends to get around 15-20% less energy than one facing directly south”. A north-east or north-west face (135° from south) comes in at 59–66%: nearer north than east. The more useful point is that a roof which faces north almost always has a face that does not. A plain gabled roof has two slopes facing opposite ways, so if the front of the house faces north the back faces south; a hipped roof has four faces, two of them east and west. “Which way does my house face?” is a question people answer with the front door, and panels do not care where the front door is.

What does the Energy Saving Trust say?

That you should not. The Trust’s solar panels advice, last updated 27 August 2026, reads: “An unshaded, south-facing roof is ideal for maximum performance. East or west facing roofs still work, but we don’t recommend installing panels on a north facing roof.” For a roof whose only usable face points north, the tables above say the same thing in numbers. What the advice cannot tell you is whether north is your only face, or whether the roof you are calling north-facing is really the front slope of a house with a south-facing back — which is the question worth settling before deciding anything.

Is a north-facing roof worse in Scotland?

Not by more than it is in England — a little less, in fact. The gap between north and south narrows as the zones go north: the Brighton zone’s north face makes 51% of its south face at 35°, London’s 55%, Glasgow’s and Belfast’s 57%, Lerwick’s and Kirkwall’s 59%. The absolute figures are lower — 437 kWh per kWp for north at 35° in the Lerwick zone against 575 in Brighton — but the penalty for facing the wrong way is smaller. The workbook does not say why; it only carries the numbers, and this is what they show.

Does shading matter as much as which way the roof faces?

It can, and it is the one factor a table cannot tell you. MIS 3002’s Appendix B opens with: “Inclination, orientation and shading are the three main site factors that influence the performance of a PV system. While drawings, maps or photos are a suitable means to determine inclination and orientation, an accurate estimation of any shade effects will typically require a site visit.” The shade factor in the estimate is 1.00 “where there is an obvious clear horizon and no near or far shading”, and otherwise 1 minus the estimated loss — the standard’s own worked example is an 11% loss giving a factor of 0.89, assessed by the method in MCS’s MGD005. A south face behind a tall tree can lose more than a west face loses to orientation, so a roof’s direction is never the whole answer. The Energy Saving Trust adds that “optimisers” can limit the effect of one shaded panel on the rest of the array, but “if your roof doesn’t have shading, optimisers won’t help you generate more electricity”.

Would a battery make up for a north-facing roof?

No. A battery stores what the panels make and moves it to a time you can use it; it does not add a kilowatt hour to the year’s total. If a north-facing array makes 2,160 kWh, that is the ceiling with or without a battery. What a battery changes is the split between what you use yourself and what you export, which is a question about when your household uses electricity — and the reason which number does the export payment actually use? has the same answer whichever way the roof faces.

How does this site work out which face of my roof is worth using?

By measuring every face from aerial imagery, then placing the panels where the light is. The model behind the roof card is Google’s Solar API, which splits your roof into segments and records each one’s pitch and compass direction — in its own reference, “0 = North, 90 = East, 180 = South” — together with the sunshine each point on the segment receives, with shadows “cast by nearby structures and trees” and “historical cloud and temperature patterns” already taken into account, per its methodology. It then lays panels out in what the reference describes as “usually, though not always, in decreasing order of annual energy production”, so a nine-panel estimate is what the nine best-lit positions make, whichever faces they sit on. The roof card lists every face of your roof and names the best-lit one — chosen by sunshine rather than by size, because on a plain gabled house the two slopes are near enough the same size that “the biggest” would be decided by measurement noise. The annual figure is then reduced by the 0.85 DC-to-AC derate Google documents, and the workings panel says so. What the imagery cannot see is a tree that has grown since the photograph, or the inside of the roof — those are the surveyor’s, as MIS 3002 says they should be.

What should I do if my only roof faces north?

Check that it is, first: walk round the house and count the faces, including the garage, any extension and any flat roof, because the tables say a flat roof beats a north-facing one at any pitch and an east or west face beats it by a wide margin. If north really is the only face, ask any installer for the estimate MIS 3002 obliges them to give you before a contract — with the postcode zone, the pitch and the orientation printed on it, so you can check the Kk against the workbook yourself — and treat a shallow pitch as a materially better case than a steep one. What you should not accept is a “north-facing” verdict with no pitch and no number behind it, in either direction: a roof that makes 2,000 kWh a year is a roof that makes 2,000 kWh a year, and whether that is worth having is a decision for you to take with the figure in front of you, not one for a rule of thumb to take for you.