The optimal angle and the profitable angle are not the same
Every guide on this subject arrives at 35 degrees facing south. That is correct and it is also close to useless, because on a pitched roof you cannot change the angle, and on a flat roof, where you can, building at 35 degrees is the wrong decision by a wide margin.
The honest position on roof angle is that it matters far less than the volume of coverage implies. Anything between 20 and 50 degrees facing south loses under 5% of annual output, and most British roofs are pitched between 30 and 45 degrees already. The interesting questions are what happens when you can choose the angle, and what happens when maximising kilowatt-hours stops being the same thing as maximising money.
- 35 degrees facing south is optimal, and the penalty for missing it is small. A 30 degree roof and a 45 degree roof are within 2% of each other. Most UK roofs are already in the right range.
- On a flat roof, shallow beats optimal by about 47%. Tilting at 35 degrees forces wide row spacing to avoid self-shading. At 10 degrees you fit 64% more panels and generate roughly 47% more from the same roof.
- A flat roof outperforms an east or west facing pitched roof. Flat delivers around 88% of optimal, east and west around 85%. Most people assume the opposite.
- Latitude tuning is a waste of attention. Optimal tilt ranges from roughly 36 degrees in Cornwall to 40 in Shetland. That four degree spread is worth well under 1% of annual output.
- Shading dominates everything. A chimney shadow crossing two panels for three hours a day costs more than a 15 degree tilt error, and on a string inverter it can cost ten times more.
The short answer
For a fixed installation anywhere in the UK, a south-facing array tilted at 35 degrees captures the most annual energy. The useful range is 30 to 40 degrees, and the acceptable range is considerably wider than that.
If you have a pitched roof, this is largely academic. Changing the pitch of a roof means rebuilding it, which no amount of solar gain justifies. Your installer will mount flush to whatever pitch you have and spend their design effort on orientation, shading and array layout, which is the correct order of priorities.
The angle question becomes a real decision in three situations: a flat roof, a ground mount, and a household choosing between two roof planes. Those are where this guide concentrates.
- Tilt
- The angle of the panel from horizontal. A flat roof is 0 degrees, a vertical wall is 90.
- Azimuth
- The compass direction the panel faces. Due south is the reference point in the northern hemisphere.
- Angle of incidence
- How close to perpendicular the sunlight strikes the panel. Output peaks when this approaches zero.
- Diffuse light
- Sunlight scattered by cloud and atmosphere, arriving from the whole sky rather than one direction. It is far less sensitive to tilt, which is why British angle penalties are milder than expected.
- Self-shading
- One row of panels casting a shadow on the row behind. The binding constraint on flat roof design.
- Row pitch
- The centre-to-centre distance between rows on a flat roof. Rises sharply with tilt angle.
- Specific yield
- Annual kilowatt-hours per kilowatt-peak installed. The figure to compare between orientations.
- Self-consumption
- The share of generation used in the house rather than exported. Angle and orientation both affect it.
What every angle actually costs
The table below gives annual output as a percentage of the maximum, which is a south-facing array at 35 degrees. Every combination of tilt and orientation is shown, because the two interact rather than adding up independently.
| Tilt | South | SE / SW | East / West | NE / NW | North |
|---|---|---|---|---|---|
| 0° (flat) | 88% | 88% | 88% | 88% | 88% |
| 15° | 96% | 94% | 88% | 80% | 74% |
| 30° | 100% | 96% | 86% | 74% | 64% |
| 35° | 100% | 96% | 85% | 72% | 62% |
| 45° | 98% | 94% | 82% | 68% | 57% |
| 60° | 92% | 88% | 76% | 61% | 49% |
| 90° (wall) | 70% | 66% | 56% | 44% | 35% |
Read across the flat row first. At zero degrees the compass direction stops mattering entirely, because a horizontal panel faces the whole sky. That single row contains one of the more useful facts in this guide.
A completely flat array produces more than an east or west facing pitched roof: 88% against 85%. People routinely reject flat roofs as unsuitable while accepting a west-facing pitched roof without hesitation. The physics says the flat roof is the better surface.
Tilt versus orientation
Of the two variables, orientation carries considerably more weight. Moving from south to west at a fixed 35 degrees costs around 15%. Moving from 35 degrees to 45 degrees facing south costs around 2%.
That produces a rule worth remembering when comparing roof planes: a south-facing roof at an imperfect pitch beats a west-facing roof at the perfect one, every time. If a house offers a 48 degree south plane and a 35 degree west plane, the steep southern one wins by roughly 16%.
The reason British penalties are milder than intuition suggests is diffuse light. On an overcast day, which describes a great deal of the UK year, light arrives from the whole sky rather than from the sun’s position, and a panel pointing slightly the wrong way loses very little. Even a north-facing array still returns 62%, which is poor but not the zero people expect.
Why latitude tuning is pointless
A popular rule sets optimal tilt equal to your latitude, or latitude minus ten to fifteen degrees for annual yield. Applied to the UK it produces a spread of about four degrees, from roughly 36 in Cornwall to 40 in Shetland.
Four degrees of tilt, at these angles, is worth well under 1% of annual output. It is smaller than the year-to-year variation in British weather, smaller than the effect of a dusty spring, and far smaller than the measurement error in most roof surveys.
Be wary of the latitude plus or minus fifteen version of the rule, which is often illustrated with examples from lower latitudes. Applied to London it recommends 66 degrees in winter and 36 in summer. That is a real optimisation for a tracker or an adjustable ground mount, and meaningless advice for a fixed roof array.
The flat roof inversion
This is where the standard advice actively misleads, and it is the section worth reading twice.
On a flat roof you can choose any tilt you like. The instinct is to build at 35 degrees and capture 100% per panel. The problem is that tilted rows shade the rows behind them, so the steeper the tilt, the further apart the rows must sit, and the fewer panels fit on the roof.
Work it through for a landscape-mounted panel roughly 1.13 metres across. At 35 degrees the back edge rises 0.65 metres, which at typical UK winter sun angles demands a row pitch of about 2.7 metres. At 10 degrees the rise is 0.20 metres and the row pitch falls to about 1.65 metres.
11 rows, 55 panels on a 30m by 10m roof, around 20,400 kWh a year
18 rows, 90 panels on the same roof, around 30,000 kWh a year
Each panel at 10 degrees produces about 90% of what it would at 35. But the roof holds 64% more of them, so total generation rises by roughly 47%. The angle that is optimal per panel is decisively wrong per roof.
Two further factors push the same way. Wind uplift rises sharply with tilt, so steeper arrays need substantially more ballast, which many flat roofs cannot carry. And since December 2023, permitted development rules allow panels to project up to 0.6 metres above a flat roof surface. A panel at 35 degrees rises around 0.75 metres including its frame, which breaches that limit and turns a straightforward job into a planning application.
The planning rules and the physics agree. On a flat roof, tilt between 10 and 15 degrees, accept the 10% per-panel loss, fill the roof, and stay inside permitted development. East-west facing rows at 10 degrees, back to back with almost no gap, pack even more capacity onto the same area and are now the standard commercial approach.
The one caveat is self-cleaning. Below about 10 degrees, rain does not run off effectively and soiling accumulates, costing a few percent a year and occasionally requiring cleaning. Ten to fifteen degrees is the balance point.
East-west: fewer kilowatt-hours, more panels
East and west facing roofs are treated as a consolation prize. That framing misses the structural advantage they carry.
A typical house with a south-facing roof plane offers one usable surface. A house running east-west offers two, and both are usable. If the south plane holds ten panels and the east and west planes hold ten each, the east-west house fits twenty panels at 85% each, which is the equivalent of seventeen south-facing panels against ten.
Put differently: an east-west roof produces 15% less per panel and can often carry twice as many. The householder with a south-facing roof usually runs out of roof before they run out of budget. The householder with an east-west roof frequently does not.
The generation profile also differs in a way that matters. South-facing arrays produce a sharp midday peak. East-west arrays produce a broader curve with a morning and an afternoon shoulder, which lines up better with when people are actually at home.
The angle that makes money
Total generation is not the objective. The objective is value, and a kilowatt-hour you use is worth more than one you export. At the July to September 2026 price cap of 26.11p and a competitive export rate of 12p, a self-consumed unit is worth 2.2 times an exported one.
That changes the arithmetic between configurations. A south-facing array concentrates output into the middle of the day, when many households are out, so a larger share is exported at the lower rate. An east-west array spreads the same generation across morning and evening, raising self-consumption.
| Configuration | Generation | Self-consumption | Annual value |
|---|---|---|---|
| South, 35° | 3,400 kWh | 32% | £561 |
| East-west split, 35° | 2,890 kWh | 45% | £531 |
| South-west, 35° | 3,260 kWh | 36% | £550 |
| Flat, 10° | 3,060 kWh | 34% | £515 |
| North, 35° | 2,110 kWh | 38% | £366 |
South still wins, but the margin collapses. A 15% generation penalty for east-west becomes a 5% money penalty, because the shape of the output curve partly compensates for its size. Add the fact that an east-west house can usually fit more panels and the ranking inverts in practice.
If you are choosing between roof planes rather than accepting what you have, ask the installer to model annual value at your import and export rates, not just annual kilowatt-hours. The two rankings are not the same, and the one printed on most quotations is the wrong one.
Shading beats angle every time
Everything above is worth a few percentage points. Shading is worth tens of them, and it is the variable most often waved through at survey stage.
A tilt error of 15 degrees costs around 3%. A chimney shadow crossing two panels for three hours a day, on a system wired with a string inverter, can cost 10% to 30% of the whole string, because panels in series operate at the level of the weakest one.
- Photograph the roof at 9am, 1pm and 5pm in both summer and winter if you can. Shadows move a long way between seasons.
- Look for chimneys, dormers, aerials, flues and satellite dishes before you look at trees. Fixed obstructions on your own roof are the commonest cause.
- Check neighbouring buildings and trees to the south, south-east and south-west. Anything to the north is irrelevant.
- Ask whether optimisers or microinverters are being proposed and why. They cost more and are worth it only where shade cannot be designed out.
- Consider moving or omitting a panel rather than accepting a shaded one. Nine clean panels frequently beat ten with one in shadow.
Seasonal adjustment
Adjusting tilt twice a year, steeper for winter and shallower for summer, gains roughly 4% to 5% of annual output. On a roof array that means getting on the roof twice a year, which no sensible person will do and no insurer will enjoy.
On a ground mount it is more defensible, because the adjustment happens at waist height with a spanner. Even there, 4% of a 4 kWp system is around 140 kWh a year, or about £25 to £35 of value. Judge whether two trips to the bottom of the garden are worth that.
Tracking systems, which follow the sun continuously, gain 15% to 25% but add moving parts, maintenance and cost. They remain rare in UK domestic installations for good reason: the extra output rarely covers the extra capital at British irradiance levels.
Roof types rated
Scores are out of five and are our editorial judgement, weighted for annual output, how much capacity the surface can carry, and design complexity.
Flat roof at 10 to 15 degrees
The configuration most often designed wrongly, and the one where getting the angle right is worth the most.
The benchmark, and the configuration most British houses already have. Anywhere in this pitch band you are within 2% of the theoretical maximum, so there is nothing to optimise. Flush mount, spend the design effort on shading and layout, and ignore anyone offering to improve the angle with brackets.
Watch for: Chimneys and dormers on the same plane, which cost far more than pitch ever will
Within 4% of due south, which is inside the noise of British weather. A south-west plane has a mild advantage over south-east for most households, because its afternoon bias overlaps better with people returning home, lifting self-consumption slightly at the same generation.
Watch for: Nothing in particular. Treat this as a south-facing roof
Underrated, and the only common roof where the angle is a real decision rather than a given. At 10 to 15 degrees a flat roof beats an east or west facing pitched roof per panel and carries far more capacity per square metre than a steeply tilted array would. Ballast weight and roof membrane condition are the real constraints.
Watch for: Quotes specifying 30 degrees or more, which cost capacity and may breach permitted development
Loses 15% per panel and often gains 100% in usable roof area, which is a trade most households should take. The broader generation curve raises self-consumption, cutting the money penalty to around 5%. Two strings on separate MPPT inputs, or microinverters, keep the two planes from dragging each other down.
Watch for: Single-MPPT inverters, which handle two orientations badly
Common on Victorian and Edwardian properties and on modern town houses. At 55 degrees facing south you retain around 94%, which is a small price. The compensations are excellent self-cleaning, better winter output when electricity is dearest, and snow that slides off rather than sitting.
Watch for: Access and scaffolding costs, which rise with pitch more than the yield falls
Around 62% of optimal at 35 degrees, and worse the steeper the roof. Rarely worth it as the only surface, but reasonable as an extension to a system that has already filled the good roof, particularly on a shallow pitch where the penalty narrows. North-east and north-west planes do better than true north and are sometimes viable.
Watch for: Payback calculations that assume south-facing yield without saying so
Pros and cons of chasing the optimal angle
- Choosing between roof planes on orientation first, pitch second
- Setting flat roof tilt at 10 to 15 degrees to maximise capacity
- Setting ground mount tilt to 35 degrees, where it is free to do
- Modelling annual value rather than annual kilowatt-hours
- Removing or relocating panels that sit in shade
- Using both planes of an east-west roof rather than one
- Tilt frames on a pitched roof to correct a few degrees
- Latitude fine-tuning, worth under 1% across the UK
- Seasonal adjustment on a roof array
- Rejecting a flat roof as unsuitable for solar
- Sun trackers on a domestic UK installation
- Rebuilding a roof to improve its pitch
Check orientation before pitch. It carries roughly seven times the weight.
Fill a flat roof at a shallow angle rather than part-fill it at the optimal one.
Survey shading across a full day and across seasons before finalising layout.
Ask for output modelled in pounds at your actual import and export rates.
Pay for tilt frames on a pitched roof already between 25 and 50 degrees.
Dismiss an east-west house. It usually offers twice the roof.
Accept a flat roof design at 30 degrees or steeper without asking why.
Let a supplier quote a postcode-specific optimal angle as if it mattered.
Frequently asked questions
What is the best angle for solar panels in the UK?
Around 35 degrees from horizontal, facing south, for maximum annual output. The useful range is 30 to 40 degrees, and anything from 20 to 50 degrees facing south stays within about 5% of the maximum. Most UK pitched roofs sit between 30 and 45 degrees, which is already close enough that there is nothing to gain by adjusting.
Does roof pitch matter much for solar panels?
Less than orientation. A 30 degree roof and a 45 degree roof facing south differ by about 2% of annual output. Moving from south to west at the same pitch costs around 15%. When comparing two roof planes, orientation should decide the answer and pitch should be a tie-breaker.
What angle should panels be on a flat roof?
Between 10 and 15 degrees, not 35. Steeper tilts force wider row spacing to prevent self-shading, so fewer panels fit. At 10 degrees each panel produces about 90% of optimal but the roof carries roughly 64% more panels, giving around 47% more total generation. Shallow tilts also need less ballast and stay within the 0.6 metre permitted development limit.
Is a flat roof worse than an east or west facing roof?
No, it is slightly better. A flat array produces around 88% of optimal against 85% for an east or west facing pitched roof at 35 degrees. A horizontal panel faces the whole sky, so compass direction stops mattering. Flat roofs also allow the array to be pointed and tilted freely, which a pitched roof does not.
Should I adjust my solar panel angle seasonally?
Not on a roof. Adjusting twice a year gains roughly 4% to 5% of annual output, which does not justify getting on a roof twice a year. On a ground mount the adjustment is safe and easy, but 4% of a 4 kWp system is only around £25 to £35 a year. Fixed at 35 degrees is the sensible default almost everywhere.
Does the optimal angle change across the UK?
Barely. Optimal tilt runs from roughly 36 degrees in the far south-west to about 40 degrees in Shetland. That four degree spread is worth well under 1% of annual output, less than year-to-year weather variation. Treat 35 degrees as a national figure and ignore postcode-specific claims.
Are north-facing solar panels worth it?
Rarely as the only surface. A north-facing array at 35 degrees returns around 62% of optimal, falling further on a steeper roof. It can make sense as an extension once the better roof is full, especially on a shallow pitch where the penalty narrows. North-east and north-west planes perform better than true north and are sometimes viable.
What matters more than angle?
Shading, by a wide margin. A 15 degree tilt error costs around 3%. A chimney shadow crossing two panels for three hours a day can cost 10% to 30% of a string on a string inverter, because panels in series operate at the level of the weakest. Orientation, array size and inverter configuration all outrank tilt as well.
Updated 22 August 2026 with the Ofgem July to September 2026 price cap, flat roof permitted development limits in force since December 2023, and revised value modelling comparing annual output against annual return.
Stop optimising the number you cannot change
If you have a pitched roof between 25 and 50 degrees facing anywhere from south-east to south-west, the angle question is closed. You are within a few percent of the theoretical maximum, tilt frames will cost more than they return, and the design effort belongs on shading, layout and array size instead.
Where the angle is a real choice, the standard advice inverts. On a flat roof, building at the optimal 35 degrees costs you roughly a third of the electricity the roof could have produced, because row spacing eats the space. Ten to fifteen degrees fills the roof, cuts the ballast, and stays inside permitted development.
Judge roof planes on orientation first and pitch second, treat a flat roof as a better surface than an east or west facing one, and ask your installer to model the array in pounds rather than kilowatt-hours. The rankings differ, and the one on most quotations is not the one that pays you.
METHODOLOGY: Output percentages modelled from plane-of-array irradiance for UK latitudes between 50 and 59 degrees north, indexed to a south-facing surface at 35 degrees. Flat roof row spacing calculated for a 1.13 metre landscape-mounted panel against typical UK winter sun elevations, giving row pitches of approximately 2.7 metres at 35 degrees and 1.65 metres at 10 degrees. Value modelling uses the Ofgem July to September 2026 price cap of 26.11p per kWh for import and a 12p export rate, with self-consumption shares varying by generation profile.
DISCLAIMER: Scores are our editorial judgement. Actual output depends on shading, local weather, panel specification, inverter configuration and roof condition; ask an MCS-certified installer for a calculation specific to your property. Permitted development rights vary for listed buildings, conservation areas and properties subject to Article 4 directions; confirm with your local planning authority. We do not provide financial advice.
