Solar panel output: what your system should produce, by size and location

Solar panel output is the number on the app that you cannot judge, because nothing tells you what it should be. A 10 kW system shows 6.8 kW at noon and 41 kWh for the day. Is that good? It depends on the month, the state, the direction your roof faces, and the temperature on the shingles, and the app knows none of that. This page gives you the expected figure for your size and location, a worked example each for Massachusetts and California, the way to read the Enphase app so you are comparing the right number, and the line between "low and normal" and "low and someone owes me."

The rating is not the output

Every system has two kW numbers and they are routinely confused. The DC rating (say, 9.6 kW) is the sum of the panel nameplates, measured in a lab at 1,000 watts per square meter of sunlight and a cell temperature of 77°F. The AC rating is what the inverters can pass to the house, usually 15 to 25% lower. Neither is what you see at noon on a real roof, for four reasons that are all normal:

• Heat. Panels run 40 to 60°F hotter than the air in summer and lose roughly 0.35% of output for every degree Celsius above the lab temperature. A 95°F afternoon costs about 10 to 12% by itself.
• Clipping. Microinverters and string inverters cap output at their AC rating. An IQ8M microinverter on a 400 W panel tops out near 325 W AC; on the brightest days that cap holds for hours and the app shows a flat top on the curve.
• Angle and sun position. The sun only lines up with your roof for a short part of the day. The rest of the time the light arrives at a slant and less of it is captured.
• Everything else. Wiring, dust, panel mismatch, and the small inefficiency of the inverter itself. NREL's PVWatts model rolls these into a default of 14% system losses.

Put together, a 10 kW DC system peaking at 7.5 to 8.5 kW on a clear June noon is exactly right. The useful comparison is never the instantaneous number. It is energy over time, in kilowatt-hours, against a modeled expectation.

Expected output by location: kWh per kW per year

The standard way to compare systems of different sizes is specific yield: how many kWh each installed kW of DC capacity produces in a year. It folds in the sunlight at your latitude, the local weather, and the losses above. The figures below are typical statewide values from PVWatts for a south-facing roof at a typical pitch, and they are the same values our free estimator uses.

Statewide averages, south-facing, typical tilt, PVWatts default losses. Your own roof's orientation, tilt and shade move the number, sometimes by 20% or more. That is why an honest audit models your roof, not your state.

Two worked examples

Massachusetts: 8 kW system, twenty 400 W panels

The average hides the season. In Massachusetts a south-facing array makes roughly 4 to 5 kWh per kW in a June month's best days and under 2 kWh per kW on a clear December day, with snow cover taking whole days to zero. Month by month, the same 8 kW system runs about 450 kWh in December and 1,150 kWh in June. A year with a cloudy May and a snowy February can land 5 to 8% below the long-run figure with nothing wrong. New England owners will find the seasonal shape and the SMART incentive context in our Massachusetts solar page.

California: 8 kW system, same panels

California's seasonal swing is gentler: roughly 750 kWh in December and 1,350 kWh in June for that system, with June-gloom coastal fog and inland heat pulling in opposite directions. Heat matters more here than in the Northeast: a Central Valley roof in August runs hot enough to give back 12 to 15% of what the extra sun would otherwise deliver. Under NEM 3.0 the dollar value of each kWh also depends heavily on when it is produced, which we cover in NEM 3.0 and the solar savings shortfall.

The same 8 kW system, the same panels, and a 3,200 kWh a year difference between the two states. That is why a national "average output per panel" figure is useless for judging your own roof, and why your contract's promised number should have been built for your address.

How to read the Enphase app so you compare the right number

Most owners look at the wrong screen. The home screen shows today, which is the noisiest possible number. Here is where the comparable figures live.

1. Energy tab, range set to Year. This shows production by month and a total. Scroll back so you have twelve full months, and add them if the app's window straddles a year boundary. That twelve-month total is the number to set against the table above or against your contract's annual figure.
2. Production versus consumption. If you have consumption monitoring, the app shows both. For a claim, only production matters. Consumption is about the house, not the panels, and a high bill with normal production is a different problem.
3. Array tab. One tile per microinverter with today's kWh or current watts. In daylight every tile should carry a number. A grey tile is a microinverter that is not reporting and, on most systems, not producing. Panel figures can lag 15 to 25 minutes because the gateway polls microinverters in rotation, so give it time before you worry.
4. Day view on a clear day. The curve should be a smooth hill, sunrise to sunset, with a flat top if you are clipping. A hill that quits early, dips at the same hour every day, or has a notch at a fixed time is shade, and shade that was not in the proposal is worth noting.
5. Lifetime and per-year. The System screen shows lifetime energy. Combined with your commissioning date, that gives you the per-contract-year totals a guarantee is assessed against.

For the full walkthrough of the app's screens and where the export lives, see reading and exporting Enphase Enlighten data.

When a low number is normal

Most "my output is low" worries fall into one of these, and none of them is a fault:

• Season. December output in the Northeast is 35 to 45% of June. That is geometry, not a problem.
• Weather. Back-to-back overcast days can cut a week to a third of the clear-sky figure. A month 15% below average is ordinary.
• Heat. The hottest, clearest afternoon of the year is not the best producing one. A mild, clear day in May beats a scorching one in July.
• Snow and soiling. A snowed-under array produces nothing until it sheds. Spring pollen and a dry summer can cost a few percent until it rains.
• Degradation. Panels lose about 0.5% a year. A ten-year-old system at 95% of its year-one figure is aging normally. See how fast solar panels degrade.

The tell in all of these is that they are already inside a correctly modeled expectation. PVWatts, run with your roof's geometry and the actual weather for the period, expects the December dip, the cloudy May and the hot July. What is left after you subtract that expectation is what matters.

When a low number is a claim

Three patterns separate a real shortfall from noise, and they show up in a per-year, per-panel comparison rather than on the home screen.

• A contract year below the guarantee threshold, against an honest expectation. Your contract promised 11,000 kWh with a 90% guarantee. The year came in at 9,100. A model of your roof under that year's weather says 10,400. The 1,300 kWh gap is not weather, because the weather is already in the 10,400. That is a claim in kWh and, at your rate, in dollars.
• Some panels flat while their neighbors work. Three tiles in the Array view reading a fraction of the rest, on the same roof plane, for weeks. That is equipment, it is usually under warranty, and the production guarantee covers the kWh lost while it went unfixed.
• A proposal that assumed a roof you do not have. The estimate modeled south-facing at 30° and the array faces east and west at 20° under a maple. The system is healthy and the promise was wrong. Under most guarantee clauses that is still the installer's problem, and it is the case a physical inspection can never find.

What to do with your number

1. Pull twelve full months of production from the Energy tab.
2. Run the free production estimator for your size and ZIP code. Within about 10% of the modeled figure, your output is normal. More than 15% under, keep going.
3. Find the annual kWh figure in your contract and the guarantee clause around it.
4. Get an independent expected figure for your exact roof and the weather of your most recent completed contract year, and compare. That is the solar performance audit: $79, PVWatts-based, exact shortfall in kWh and dollars for that year, earlier years as a range, with the spreadsheet behind every figure. Enphase systems today. If the answer is that your output is fine, the report says so, and that is a valid result.

OwlWatt sells no panels, installs nothing, and takes no referral fees or commissions from installers or inverter vendors. We measure. What you do with the report is up to you, and we never contact your installer.