South-facing at roughly your latitude captures the most annual kWh — but "most kWh" and "most savings" aren't always the same thing. In time-of-use markets, a west-facing roof can be worth more per kWh because it produces during the 4–9 p.m. peak when grid rates spike. East-facing roofs hit 80–85% of south-facing production and remain viable. Tilt is more forgiving than most people think. Your roof is almost certainly closer to optimal than the conventional wisdom suggests.
Why south-facing at your latitude wins on total kWh
A south-facing roof tilted at roughly your latitude — about 30 degrees in Houston, 35 in Phoenix, 40 in New York — produces the most annual kilowatt-hours in the continental U.S. The sun crosses the southern sky through every season, and a south-facing surface catches the most direct light through the year.
The production gap between south and other orientations is real but smaller than people assume. Compared to a south-facing roof at ideal tilt, east-facing produces roughly 80 to 85 percent. West-facing produces 85 to 90 percent. North-facing produces 40 to 50 percent and rarely makes sense.
That's annual total. The picture changes when you look at when the production happens. South-facing peaks at midday. West-facing peaks in late afternoon. East-facing peaks in late morning. The same total kWh has different financial value depending on what your utility charges at those hours.
If you're on a flat-rate tariff with classic full net metering, total kWh is what matters and south wins. If you're on a time-of-use rate or a NEM 3.0-style successor tariff, when you produce matters more than how much you produce in total. That's where conventional wisdom breaks down.
West-facing can beat south-facing in time-of-use markets
West-facing panels produce about 10 to 15 percent less total energy than south-facing, but the production they do create lands in the afternoon when utility peak rates are highest. In California, Arizona, parts of Texas, and a growing list of other TOU markets, that timing premium can outweigh the lower volume.
Run the math. A south-facing 8 kW system in a TOU market might produce 12,000 kWh per year, with about 60 percent of that landing during off-peak or mid-peak hours valued at 18 to 25 cents per kWh. A west-facing 8 kW system on the same roof produces about 10,500 kWh, but roughly 70 percent of that lands in peak hours valued at 45 to 55 cents per kWh. Annual offset value can favor west-facing even though total production doesn't.
This is utility-specific. Ask your Solar Partner to model your actual tariff in their proposal software. If your utility uses flat rates or your peak window is short, south-facing still wins. If your peak window runs from 4 p.m. to 9 p.m. and rates double during those hours, west-facing often does.
The same logic applies to mixed installs. Two-thirds south plus one-third west often outperforms an all-south layout in TOU markets while keeping total kWh high.
East-facing is a viable second choice; north-facing rarely is
East-facing panels produce 80 to 85 percent of what a south-facing array does and concentrate that production in the morning, from sunrise through early afternoon. Morning kWh has lower TOU value than afternoon kWh, but it still beats most alternative orientations and beats not going solar.
East-facing also avoids one minor complication: late-afternoon heat. Panel efficiency drops about 0.3 to 0.5 percent per degree Celsius above 25°C. West-facing panels produce in the hottest part of the day, so they lose a bit more output to heat than east-facing panels. That's a small effect — usually a few percent — but it slightly narrows the east-versus-west gap in hot climates.
North-facing rarely pencils out. Production runs 40 to 50 percent of south-facing and lands across the whole day without any time-of-use timing advantage. Exceptions exist: very low-latitude locations like southern Florida or Hawaii, very low-pitch roofs where orientation matters less, and roofs with no other viable plane. For most homeowners with a true north-facing primary roof, look at east, west, or a ground-mounted system.
Tilt is forgiving from flat to about 45 degrees
Tilt is the variable people worry about most and that matters least. The ideal tilt — roughly equal to your latitude — is a smooth optimum, not a cliff. A roof tilted 10 degrees off ideal loses only a few percent of annual production. A roof tilted 20 degrees off loses around 5 to 8 percent.
Flat and low-pitch roofs work well with solar. Ballasted racking systems sit on the roof without penetrations and tilt the panels to whatever angle you choose — often 10 degrees for self-cleaning rain runoff. Production is close to optimal and you avoid the cost and risk of drilling into the membrane.
Steep roofs above 45 degrees do start losing meaningful production, especially in winter when the sun is low. A 60-degree roof loses 10 to 15 percent compared to ideal. For most A-frame and Cape Cod roofs that's still acceptable.
Tilt drives small percentage swings. Orientation drives larger ones. Shading drives the largest. If you're triaging factors, work backward from shading, then orientation, then tilt.
Most homes use multiple roof planes — and ground mount unlocks the rest
Modern solar design isn't a single-plane decision. Your home probably has two to four usable roof faces, and proposal software optimizes panel placement across all of them. You might end up with 12 panels south-facing on the main roof, 6 west-facing on an addition, and 4 east-facing on a garage. That hybrid maximizes both total production and peak-rate timing.
Ask your Solar Partner to show you the per-plane production breakdown in their proposal — not just the system total. You should see kWh per year for each roof segment. If they can only quote the whole-system number, they may not be running plane-by-plane optimization.
Ground-mount is the escape hatch when no roof plane works. You pick the orientation and tilt. Some ground-mount systems even include seasonal tilt adjustment, which pushes annual production a few percent higher. The cost is real — typically $0.50 to $1.00 per watt more than rooftop, plus yard space and trenching. For homes with steep north-facing roofs, heavy shading, or roof age concerns, ground-mount often pays back better than forcing solar onto the wrong plane.