Solar Panel Tilt Angle & Orientation: The Honest Numbers
Tilt roughly at your latitude, aim at true south or north, and stop worrying — the loss curve is gentler than the internet claims. Here is what every compromise actually costs, in percent.
Tilt and orientation are the two free levers of solar design: they cost nothing to get right at install time and quietly tax you forever if you get them wrong. The good news, which this guide quantifies honestly, is that the penalty curve is gentle — most reasonable roofs land within 10% of perfect. The point of knowing the real numbers is not to chase the last percent, but to know which compromises are cheap and which ones actually hurt.
The latitude rule — and why it works
For maximum annual energy from a fixed array, tilt the panels up from horizontal by roughly your latitude. A site at 30° latitude wants about 30° of tilt; at 50° latitude, about 50°. The logic is geometric: over a full year the sun’s average midday height in the sky is set by your latitude, and a latitude tilt points the panel at that average. Fine-tuning by a few degrees for local climate (haze, monsoon season, winter fog) shifts the optimum slightly, but the rule of thumb gets you within 1–2% almost everywhere.
Annual optimum: tilt ≈ latitude
Summer setting: tilt ≈ latitude − 15°
Winter setting: tilt ≈ latitude + 15°Is seasonal adjustment worth the ladder trips?
Here is the honest answer that adjustable-rack marketing avoids: a fixed array at latitude tilt harvests only about 3–5% less per year than one you re-tilt twice a year, and four adjustments per year adds barely another 1%. For a grid-tied home, that is rarely worth the hardware and the climbing. The exception is winter-critical off-grid systems: if your batteries must survive December, set the tilt for the worst month (latitude +15° or steeper) and cheerfully give up some summer surplus you did not need. Full sun trackers gain 25–35% but add cost, moving parts and maintenance — nowadays it is almost always cheaper to simply add two more panels.
Orientation: true south, not compass south
In the northern hemisphere, fixed panels yield most facing true south (azimuth 180°); in the southern hemisphere, true north. Note the word true: a magnetic compass can be off by up to 15–20° depending on where you live, because magnetic declination varies across the globe. Use a declination-corrected compass app, satellite imagery, or the sun at local solar noon to find true south. A 15° aiming error costs little (more below), but there is no reason to donate even that.
Get your exact numbers. Enter your latitude and roof direction to see optimal tilt, seasonal settings and the real cost of your roof’s compromise.
Open the tilt angle calculator →What deviation really costs: the penalty table
Typical annual losses versus a perfectly tilted, perfectly aimed array at mid latitudes (25–50°):
| Configuration | Typical annual loss |
|---|---|
| Optimal tilt, true south/north | 0% (reference) |
| Tilt off by ±15° | 1–3% |
| Azimuth off by ±45° (SE or SW roof) | 4–8% |
| Flat (0° tilt) | 8–12%, plus ongoing soiling losses |
| Due east or due west | 10–18% |
| Vertical façade, equator-facing | 30–40% |
Read that table twice, because it carries the guide’s main message: the loss surface is a shallow bowl. Anything within 15° of optimal tilt and 45° of optimal azimuth is a rounding error compared with shading, soiling or an undersized array. Model your own roof in the solar output calculator before rejecting it.
East-west arrays: sometimes the smart choice
Splitting an array half-east, half-west costs 10–15% of annual energy versus due south — and is still often the right call. The production curve flattens: strong output from early morning to late evening instead of one sharp noon peak. That matches household self-consumption far better, which matters when export rates are poor. On flat commercial roofs, low-tilt east-west racking also packs 20–40% more panels into the same area because rows do not shade each other. Check what your roof can hold with the roof area calculator.
Minimum tilt, rain and snow
Keep at least 10° of tilt whatever else you do. Below that, rain no longer sheets off effectively: water ponds along the frame edge, dust and pollen accumulate into a stubborn grime line, and you trade a small geometry gain for a permanent 3–5% soiling loss plus regular cleaning duty. In snow country, go steeper: 35–40° or more lets snow slide off hours after a storm instead of days, and conveniently coincides with the winter-optimized tilt that low winter sun rewards anyway. See how panels perform in winter and cloudy weather for the seasonal detail.
Ground mounts and adjustable racks
Ground mounts free you from the roof’s geometry: you pick the exact tilt and azimuth, cleaning is easy, and cooling airflow is better. If you buy an adjustable-tilt rack, the twice-a-year schedule is simple — around the spring equinox set latitude −15°, around the autumn equinox set latitude +15°. Just be realistic about whether you will actually do it; a fixed latitude tilt you never touch beats an adjustable rack left in the wrong season.
Worked example: a home at 30° latitude
Take a 5 kW south-facing array at 30° latitude in a sunny climate producing about 8,000 kWh per year at its best tilt. Modeled annual output by tilt:
- 0° (flat): ≈ 7,280 kWh — about 91% of optimum, before extra soiling losses
- 15°: ≈ 7,840 kWh — about 98% of optimum
- 30° (latitude): ≈ 8,000 kWh — the reference optimum
- 45°: ≈ 7,680 kWh — about 96%, with a winter-shifted production curve
The spread between 15° and 45° is barely 4%. A typical 20–25° roof pitch at this latitude is, for practical purposes, already optimal.
Why azimuth matters less than people fear
Homeowners routinely dismiss a southeast or southwest roof as ‘wrong’, yet the physics says a ±45° azimuth error costs only about 6–8% annually at mid latitudes. The sun sweeps roughly 180° or more across the sky every day; a panel aimed 45° off still faces the sun squarely for hours, just shifted earlier or later. A west-facing roof even earns a bonus where late-afternoon electricity is priced highest. The configurations that genuinely hurt are the extreme ones — vertical walls, polar-facing slopes and unwashed flat panels — not ordinary imperfect roofs.
Common mistakes
- Aiming panels with an uncorrected compass and building in a 10–20° magnetic declination error for free.
- Mounting panels dead flat for looks, then losing energy to ponding water and a permanent grime stripe.
- Buying an adjustable rack, adjusting it once, and leaving it on the summer setting through two winters.
- Rejecting a good southeast roof over a 5% azimuth penalty while ignoring the chimney shadow that costs 15%.
- Using summer-optimized tilt on an off-grid cabin that actually lives or dies by December production.
When in doubt, run your coordinates through the tilt angle calculator — the right answer for your roof is a two-minute lookup, not a guess.
Frequently asked questions
Should panels face magnetic south or true south?
True south (true north in the southern hemisphere). Magnetic declination can put compass south 15–20° off true south depending on your location, so use a declination-corrected app, satellite imagery or the sun at solar noon to aim the array.
Is it worth adjusting tilt through the year?
Usually not for grid-tied homes: twice-yearly adjustment recovers only about 3–5% over a fixed latitude tilt. It can pay on off-grid systems where winter output is critical, or where an adjustable ground-mount rack already exists and adjusting takes minutes.
Can I install panels completely flat?
Avoid it. Below about 10° of tilt, rain stops self-cleaning the glass, water ponds at the frame and soiling losses of 3–5% become permanent. On flat roofs, use low-tilt racking — often east-west — to keep at least 10°.
How much does an east- or west-facing roof lose?
Typically 10–18% annually versus an optimal south-facing array at mid latitudes, and a deliberate east-west split loses roughly 10–15%. The flatter morning-and-evening production curve often matches household usage better, so the effective financial loss can be smaller than the energy loss.