What angle should solar panels be? Tilt and direction, priced in kilowatt-hours
Two numbers decide how much sun a fixed solar array catches: tilt, how far the panels lean back from flat, and azimuth, the compass direction they face. The textbook answer is short: tilt near your latitude and face true south. For most houses that is the whole story, and the tables in this guide show why. Near the top of the curve, a few degrees either way costs almost nothing, while a full swing in direction costs about a fifth of the year's output.
Every production figure below is kilowatt-hours per kilowatt of panels per year, from PVGIS 5.2 runs for Phoenix and Chicago. It scales to any system size: multiply by the kilowatts on your roof. Dollar figures appear once, in the worked example, and use the site's Phoenix example rate of 16 cents per kilowatt-hour as an illustration, not a price. The two cities sit far enough apart in latitude and weather to show which patterns hold everywhere and which ones belong to one city's sky.
Two angles, one decision
Tilt is the angle of the panel surface above horizontal: 0 degrees is flat on the roof, 90 degrees is a vertical wall. Azimuth is the compass direction the panel face points, measured clockwise from true north: 0 is north, 90 is east, 180 is south, 270 is west. A fixed array has no moving parts, so these two numbers are set once, at install, and they shape every kilowatt-hour the system makes for the next 25 years.
The sun does two things over a year that make both angles matter. It rises higher in summer and hangs lower in winter, so tilt sets how square the panel sits to the sun across the seasons. And it crosses the southern sky from east to west each day, so azimuth sets which hours the panel works hardest: east-facing panels peak in the morning, south-facing at midday, west-facing in the afternoon.
The good news is the shape of the curve. Output rises to a broad, flat top at the ideal tilt and direction, then falls away slowly at first and faster later. That flat top is why a roof that misses the ideal by 10 degrees can still be a good solar roof, and why the direction question deserves more of your attention than the tilt question. The U.S. Department of Energy puts the sweet spot for the continental United States at a south-facing array tilted between 15 and 40 degrees, a range wide enough to cover most American roofs as they stand.
Tilt: start at your latitude
The standard rule for a fixed array is to set the tilt close to your latitude. The reasoning is geometric: over a full year, a panel tilted at its latitude sits square to the average sun position. Miami at 26 degrees north wants about a 26-degree tilt; Chicago at 42 wants about 42. The table below gives the rule for ten cities, and your latitude is a search away if your city is not listed.
Treat the rule as a starting point, not a law. The true optimum usually lands a few degrees shallower than latitude, because summer days are longer than winter days and the extra summer sun rewards a slightly flatter panel. In the Chicago run below, a 27-degree tilt beats a 42-degree tilt by about half a percent over the year. The difference is small enough that no roof decision should turn on it, but it is a useful check on anyone selling precision: if a quote promises an exact optimum tilt, the gain behind that precision is a fraction of a percent.
| City | Latitude | Suggested fixed tilt |
|---|---|---|
| Miami, FL | 25.8° | 26° |
| Houston, TX | 29.8° | 30° |
| Phoenix, AZ | 33.4° | 33° |
| Los Angeles, CA | 34.1° | 34° |
| Atlanta, GA | 33.7° | 34° |
| Denver, CO | 39.7° | 40° |
| New York, NY | 40.7° | 41° |
| Chicago, IL | 41.9° | 42° |
| Minneapolis, MN | 45.0° | 45° |
| Seattle, WA | 47.6° | 48° |
The spread across the country is the point: the whole continental United States sits between about 25 and 49 degrees north, so the ideal tilt for any American roof falls in a 25-degree band. That is a narrow target, and most roofs already land inside it.
What the tilt curve actually looks like
A rule of thumb says where the top is. The curve says how much the slope costs. These are PVGIS 5.2 runs for a 1 kW DC array, 14 percent system loss, building-integrated mounting, due south, with PVGIS-NSRDB radiation and ERA5 weather. Phoenix first, because its clear skies show the geometry with the least weather in the way.
| Tilt | Annual output (kWh per kW) | Share of the best tilt |
|---|---|---|
| 0° (flat) | 1,503 | 87% |
| 10° | 1,616 | 94% |
| 20° | 1,689 | 98% |
| 33° | 1,722 | 100% |
| 45° | 1,691 | 98% |
| 60° | 1,571 | 91% |
Flat is the worst tilt in the table and it still keeps 87 percent of the best output. Anything from 10 to 45 degrees lands within 6 percent of the maximum. That is the whole tilt story for a homeowner: unless the roof is dead flat or nearly vertical, tilt is a second-order question.
Chicago shows the same shape with a shallower top. Its best tilt in this run is 27 degrees, not the latitude figure of 42, and the output at 42 degrees is 99 percent of that best. A roof at 15 degrees still keeps 97 percent.
Direction: south wins, and by how much
Direction is the bigger lever. The same PVGIS settings, now with tilt held at the latitude figure and the direction swung around the compass:
| Direction | Azimuth | Annual output (kWh per kW) | Share of south |
|---|---|---|---|
| South | 180° | 1,722 | 100% |
| Southeast | 135° | 1,633 | 95% |
| Southwest | 225° | 1,620 | 94% |
| East | 90° | 1,398 | 81% |
| West | 270° | 1,382 | 80% |
| North | 0° | 903 | 52% |
The pattern is forgiving near south and punishing at the edges. Thirty degrees either way costs 5 or 6 percent. Due east or west costs about a fifth. Due north costs about half. The Department of Energy's Solar Energy Technologies Office gives the same shape in rounder numbers, quoted by EnergySage: panels 10 degrees off south lose under 1 percent over a year, an ideal east-west setup loses 10 to 15 percent against a perfect south-facing system, and a typical rooftop east-west setup loses about 20 percent. In the office's San Diego example, a south-facing system on a typical 18-degree roof loses 2 percent against the ideal 30-degree tilt, while east- and west-facing panels at the same tilt lose 15 and 10 percent.
The same compass, a cloudier city
Chicago's run agrees on the shape and adds a warning about local weather. The same settings, the same tilt rule, a different sky:
| Direction | Azimuth | Annual output (kWh per kW) | Share of south |
|---|---|---|---|
| South | 180° | 1,162 | 100% |
| Southeast | 135° | 1,168 | 100% |
| Southwest | 225° | 995 | 86% |
| East | 90° | 984 | 85% |
| West | 270° | 782 | 67% |
| North | 0° | 569 | 49% |
Southeast edges out due south by a hair, 1,168 against 1,162 kWh per kW, and west-facing output falls to 67 percent of south while east-facing holds 85 percent. That asymmetry is weather, not geometry: the ERA5 weather file behind the run has cloudier Chicago afternoons than mornings. Phoenix, with clearer afternoons, puts east and west within a point of each other. The lesson is to run your own city before you price a west-facing roof, because the west number is the one local weather moves most.
Your roof pitch is already a tilt
Most homeowners never choose a tilt at all. The panels mount flush to the roof, and the roof pitch is the tilt. A roof quoted as 6/12 rises 6 inches for every 12 inches of run, and the angle that makes is the arctangent of 6 over 12, or 26.6 degrees. The conversions:
The common American roof pitch, 4/12 to 6/12, is 18 to 27 degrees. In the southern half of the country that sits at or near the ideal tilt. In the north it runs 10 to 20 degrees shallow, which the Chicago table prices at 1 to 3 percent of annual output. Steeper roofs, 10/12 and above, overshoot the ideal in the south and land close to it in the north. Either way, the pitch you have is usually within a few percent of the pitch you would choose.
That is why tilt-up legs, the brackets that prop panels above the roof plane to a steeper angle, rarely pay on a pitched roof. On a flat roof they are the whole design, and the flat-roof spacing guide works through their layout. On a pitched roof they buy back 1 to 3 percent of annual output while adding wind load, extra roof penetrations, and a visible rack. If an installer proposes them on a pitched roof, ask for the kilowatt-hour gain in writing and compare it with the added cost.
| Roof pitch | Tilt from horizontal |
|---|---|
| 3/12 | 14.0° |
| 4/12 | 18.4° |
| 5/12 | 22.6° |
| 6/12 | 26.6° |
| 7/12 | 30.3° |
| 8/12 | 33.7° |
| 9/12 | 36.9° |
| 10/12 | 39.8° |
| 12/12 | 45.0° |
Seasonal tilt and adjustable mounts
The sun sits lower in winter, so the winter ideal is steeper and the summer ideal flatter. The rule of thumb is latitude plus 15 degrees in winter and latitude minus 15 in summer: a Chicago array would sit at 57 degrees in January and 27 in July. Two adjustments a year, at the equinoxes, capture most of the gain.
Most of the gain is small. EnergySage, quoting the Department of Energy, puts the reward for a perfect azimuth and tilt at 1 to 2 percent more energy over a year than a good fixed setup. Seasonal adjustment splits that further, so each adjustment is worth well under 1 percent on a system that is already close. On a fixed roof mount it is not worth climbing up twice a year. On an adjustable ground mount, where the tilt changes with a wrench and no ladder, two adjustments a year is a free habit. The ground-mount guide covers when that trade is worth making.
Trackers, which follow the sun through the day, are a different purchase with a different payback, and they belong on ground mounts, not roofs. For a roof array, the decision is fixed tilt, and the tables above are the whole argument.
When west beats south: the time-of-use exception
Kilowatt-hours are not dollars. A west-facing array makes about a fifth fewer kilowatt-hours than a south-facing one in the Phoenix run, but it makes them later in the day, when the sun is still up and the house is cooling down, cooking dinner, and running its evening load. Under a time-of-use rate plan, afternoon and evening electricity costs more than midday electricity, so the west-facing kilowatt-hours can be worth more per unit than the south-facing ones they replace.
Whether that trade wins depends on your utility's peak window. The time-of-use guide checks six utilities' peak hours hour by hour. If your peak runs 4 to 9 p.m., a west or southwest array deserves a real comparison against south, priced in dollars at your actual rates, not in kilowatt-hours. If your utility still pays a flat rate for exports, south wins on both measures and the argument is over.
The same logic covers east-west split arrays: half the panels face the morning sun, half face the afternoon sun, and the daily curve flattens out. Averaging the east and west figures in the Phoenix run puts a split array at about 81 percent of a south-facing array's annual output, but its output lands in the morning and evening, when a house without a battery actually uses power. For a home that self-consumes most of what it makes, that shape can be worth more than the missing kilowatt-hours. The battery guide covers when storage changes the same calculation.
True south is not magnetic south
Every number above is measured from true south, the direction of the geographic South Pole. A compass points at magnetic south instead, and the gap between the two is magnetic declination. Across much of the central United States the gap is a few degrees and can be ignored. In the Pacific Northwest and New England it runs into double digits, enough to turn a roof that reads south on a compass into a southeast-facing array in the model.
Phone compass apps usually correct for declination on their own, using the phone's location, but the correction is silent and easy to miss. Check the number for your address on NOAA's magnetic field calculator before you trust a compass reading, and compare it with the azimuth on the production estimate in your quote. If the two disagree by more than a few degrees, ask which one the installer used. The production estimate guide shows what that document should state.
Satellite views settle most arguments faster than a compass. Pull up the roof, draw a line along the ridge, and the roof faces are usually obvious. A south-facing roof plane in the northern hemisphere is the one that catches the midday sun; in a satellite photo taken near noon it is the bright one.
The decision, in order
Angle decisions go wrong when they are made in the wrong order: a tilt kit before the azimuth is known, or a direction picked before the rate plan is checked. Work the list in order and each step uses the number the last one produced.
- Read the roof. Get the azimuth of each roof plane from a compass app or a satellite view, corrected for declination, and the pitch from the building plans or the installer. Write both numbers down; every step after this one uses them.
- Look up your latitude and compare. If the pitch sits within 15 degrees of the latitude and the azimuth within 30 degrees of south, the roof is close to ideal and the rest of this guide is reassurance, not a decision.
- Run PVWatts at the real tilt and azimuth, then again at south-facing and latitude tilt. NREL's free PVWatts calculator takes an address and returns both numbers in minutes. Keep the two outputs; the gap between them is the price of the roof you have.
- Check the rate plan. If your utility's peak hours run into the evening, price a west or southwest array in dollars before you rule it out. The time-of-use guide has the peak hours for six utilities.
- Ask the installer for the production estimate at the actual azimuth and tilt, in writing, and check that the estimate's azimuth matches the roof you measured. A quote that models a south-facing array for an east-facing roof is the angle error to watch for in solar sales.
A worked example: 8 kW in Phoenix, south against west
Take an 8 kW system in Phoenix, priced against the DOE Q1 2025 residential benchmark of $2.95 per watt DC, which puts the system at $23,600 before any incentive. This is an illustration built from the PVGIS figures above, not a quote. South-facing at a 33-degree tilt, the run gives 1,721.5 kWh per kW per year, so the system makes 8 times 1,721.5, or 13,772 kWh a year. The same 8 kW facing due west at the same tilt makes 8 times 1,381.5, or 11,052 kWh a year. The direction alone costs 2,720 kWh a year, about a fifth of the output.
At the site's Phoenix example rate of 16 cents per kilowatt-hour, that gap is worth about $435 a year if every kilowatt-hour displaces grid power at that rate, which assumes full self-consumption or flat-rate export at the same price. Under a time-of-use plan the west array's afternoon kilowatt-hours are worth more than 16 cents and the south array's midday kilowatt-hours less, so the dollar gap shrinks. That is the point of the example: the kilowatt-hour gap is fixed by geometry, and the dollar gap is set by the rate plan.
For scale, the tilt gap is smaller. The same south-facing system laid flat makes 8 times 1,502.8, or 12,022 kWh a year, which is 1,750 kWh a year behind the 33-degree tilt, or about $280 a year at the same example rate. Tilt is worth getting right when it is free, as it is on a flat roof or a ground mount, and worth ignoring when the roof has already decided it.
The tilt-up upsell
The angle mistake worth watching for in solar sales is not a wrong angle but a paid correction of a right one. An installer proposes tilt legs on a 5/12 pitched roof to lift the panels from 23 degrees to the 33-degree latitude ideal, and the proposal reads like precision engineering. The Phoenix table prices that correction: moving from 20 to 33 degrees gains about 2 percent of annual output. The legs add wind load, extra roof penetrations, and a rack that stands proud of the roofline, and the gain is smaller than the year-to-year weather variation in the same run. On a flat roof, tilt legs are the design. On a pitched roof, they are usually money spent on the flattest part of the curve.
Before you sign
- Measure the azimuth and pitch of every roof plane the quote uses, and correct the compass reading for magnetic declination.
- Compare the pitch with your latitude and the azimuth with true south, and note the gap in degrees for each.
- Run PVWatts at the real tilt and azimuth and at the south-facing, latitude-tilt baseline, and keep both outputs.
- Check your utility's peak hours before ruling out a west-facing array.
- Ask the installer for the production estimate at the actual azimuth and tilt, in writing.
- Skip tilt legs on a pitched roof unless the installer shows the kilowatt-hour gain they buy.
- Confirm the estimate's azimuth matches the roof you measured, not the roof the quote assumed.
Angle and direction questions
What is the best angle for solar panels?
A tilt close to your latitude, facing true south. The Department of Energy puts the sweet spot for the continental United States at 15 to 40 degrees. The true optimum is usually a few degrees shallower than latitude, and anything from 10 to 45 degrees lands within about 6 percent of the maximum in the Phoenix run, so a roof that misses the ideal by 10 degrees is still a good solar roof.
What direction should solar panels face?
True south, an azimuth of 180 degrees, in the northern hemisphere. Panels 10 degrees off south lose under 1 percent over a year, and anything from southeast to southwest, 135 to 225 degrees, stays within about 6 percent of the maximum. Direction past that range is where the losses start: due east or west costs about a fifth of annual output.
Do solar panels work facing east or west?
Yes. In the Phoenix run, east- and west-facing arrays at a 33-degree tilt make 81 and 80 percent of what a south-facing array makes. They peak in the morning and afternoon instead of at midday, which suits homes that use power in the morning or evening, and a west-facing array can earn more under a time-of-use plan even while it makes fewer kilowatt-hours.
Can solar panels face north?
They produce power, but about half of what a south-facing array makes: 52 percent in the Phoenix run and 49 percent in the Chicago run at a latitude tilt. North-facing panels only start to make sense at a very low tilt in a sunny place, and even then the payback stretches. If north is the only roof plane available, a ground mount or community solar is usually the better comparison.
Should I adjust my solar panel tilt seasonally?
Only on an adjustable mount, which usually means a ground mount. The rule is latitude plus 15 degrees in winter and latitude minus 15 in summer, but the gain from a perfect tilt is 1 to 2 percent over a year, so each seasonal adjustment is worth well under 1 percent. On a fixed roof mount it is not worth climbing up twice a year.
Does roof pitch matter for solar?
Less than most people expect. The common American roof pitch of 4/12 to 6/12 is 18 to 27 degrees, which sits at or near the ideal tilt in the southern half of the country and 10 to 20 degrees shallow in the north, costing 1 to 3 percent of annual output. Steep roofs overshoot the ideal in the south and land close to it in the north. The pitch you have is usually within a few percent of the pitch you would choose.
How do I find my roof's direction?
Use a phone compass app or a satellite view, and correct the compass reading for magnetic declination with NOAA's magnetic field calculator. A compass points at magnetic south, not true south, and the gap runs into double digits in the Pacific Northwest and New England. Compare the result with the azimuth on the production estimate in your quote; if the two disagree, ask the installer which one the estimate used.
Sources and verification
- EnergySage, What's the Best Angle for Solar Panels?: Department of Energy Solar Energy Technologies Office figures quoted in the article: 15 to 40 degree tilt range for the continental United States, under 1 percent loss for panels 10 degrees off south, 10 to 15 percent loss for an ideal east-west setup and about 20 percent for a typical rooftop east-west setup, San Diego example of a 2 percent loss for an 18-degree south-facing roof against a 30-degree ideal and 15 and 10 percent losses for east- and west-facing panels at the same tilt, and 1 to 2 percent more energy from a perfect azimuth and tilt. Checked November 8, 2026.
- European Commission, Joint Research Centre, PVGIS 5.2 (PVcalc): Author's own runs, November 8, 2026: PVGIS-NSRDB radiation with ERA5 weather, 1 kW DC crystalline silicon, 14 percent system loss, building-integrated mounting. Phoenix (33.45 N, -112.07 W) and Chicago (41.88 N, -87.63 W), tilt and azimuth sweeps as tabulated in the guide. Checked November 8, 2026.
- National Renewable Energy Laboratory, PVWatts Calculator: Free calculator for modeling a specific address, tilt and azimuth; the guide's decision steps tell the reader to run their own roof through it. Checked November 8, 2026.
- NOAA National Centers for Environmental Information, Magnetic Field Calculators: Look up magnetic declination for an address before trusting a compass reading for true south. Checked November 8, 2026.
Production figures are PVGIS 5.2 model runs (PVGIS-NSRDB radiation, ERA5 weather, 1 kW DC, 14 percent system loss, building-integrated mounting) for Phoenix (33.45 N, -112.07 W) and Chicago (41.88 N, -87.63 W), run November 8, 2026. They are planning figures, not a roof-specific design: they leave out shading, snow, and soiling beyond the flat 14 percent loss term, and the Chicago west-facing figure reflects afternoon cloud in the ERA5 weather file, not a general rule. City latitudes are rounded to one decimal and roof-pitch angles are computed as the arctangent of rise over run. The dollar illustration uses the site's Phoenix example rate of 16 cents per kilowatt-hour and the PVGIS figures, not a quote or a local price survey. The cost line uses the DOE Q1 2025 residential benchmark of $2.95 per watt DC. No federal residential solar tax credit is in any figure. Nothing here is engineering or financial advice.