Actual vs Estimated Solar Production: How to Compare Them the Right Way
A step-by-step method for checking your solar output against the installer's estimate, with normal ranges, red flags and what to do about a real shortfall.
To compare actual vs estimated solar production, add up your system's real output for a full 12 months from your monitoring app or production meter, then compare that total with the annual estimate in your contract or proposal. A single month can swing far from the estimate because of weather, so judge the system on yearly totals and on trends over more than one year.
Estimates are built for a typical weather year, and your actual year is never typical. The steps below help you separate normal variation from a real problem worth raising with your installer.
Key takeaways
- The PVWatts Version 5 manual says estimates based on typical weather can be off by up to about 10% for a year and 30% for a month, and a specific year can differ further.
- Compare 12-month totals, not single months or days.
- Expect slow decline. Long-term research by NREL (now NLR) found a median panel degradation rate of 0.5% per year, and Tesla tells its customers to expect about the same.
- Some contracts include a production guarantee with its own math. Palmetto's Energy Plan guarantee, for example, checks for at least 90% of estimated output every 36 months.
- A sudden drop, or one section producing much less than the rest, is a fault signal, not weather.
Where your solar production estimate comes from
Most estimates come from modeling software that combines your roof's tilt and direction, local sunlight records and a set of assumed losses. NLR's PVWatts is a common example. Its Version 5 manual (2014) lists default losses for soiling (2%), shading (3%), mismatch (2%), wiring (2%), connections (0.5%), light-induced degradation (1.5%), nameplate rating (1%) and availability (3%), with snow and age at 0%. Combined, they come to about 14%. They are multiplied together rather than added.
Two points matter for owners. First, the model uses typical-year weather, not a forecast of any specific year. Second, if your installer assumed less shade, less snow or less dirt than you really get, the estimate will run high every year. Ask for the design report so you can see the assumptions. For more on how these numbers are built, read what a solar production estimate is and how accurate it is.
How far apart should actual and estimated production be?
The PVWatts manual is unusually frank about uncertainty. It says results based on long-term typical weather may be off by as much as plus or minus 10% for annual totals and plus or minus 30% for monthly totals. For a particular year, actual output can deviate further from the long-term average, by up to about 20% annually and 40% monthly.
| Comparison period | Variation the PVWatts manual describes | How to read it |
|---|---|---|
| One day or week | Not addressed; weather dominates | Do not judge the system on this |
| One month (vs typical-year model) | Up to about 30% | Useful only for spotting trends |
| One month (specific year vs long-term average) | Up to about 40% | A gray winter month can look alarming but be normal |
| Full year (vs typical-year model) | Up to about 10% | The fairest single comparison |
| Full year (specific year vs long-term average) | Up to about 20% | Check whether your region had an unusual year |
These ranges describe model and weather uncertainty, not a promise from your installer. Your contract decides what counts as underperformance.
How to compare actual vs estimated solar production, step by step
1. Find the estimate
Look for an annual kWh figure in your proposal, contract or design layout. Tesla, for example, says expected performance information appears on the solar panel layout document. If a monthly breakdown is included, save it too.
2. Get 12 months of real output
Your inverter or monitoring app shows production by month. Use a full year so every season is counted. If monitoring has gaps, note the missing dates so you do not mistake an outage in the data for an outage in production.
3. Cross-check with a meter
Some homes have a separate one-way generation meter. Dominion Energy explains that its generation meter's register 01 tracks all solar production, and you subtract an earlier reading from a later one to get output for a period. A net meter is less useful here. San Diego Gas & Electric notes that a net meter shows only the electricity sent back to the grid, not your total solar output.
4. Adjust for age
A 2012 NREL review of nearly 2,000 measured degradation rates found a median of 0.5% per year. If your estimate is for year one and your system is in year five, expect output a few percent lower. Some estimates already include degradation, so check.
5. Calculate the percentage
Divide actual annual kWh by estimated annual kWh. Worked example (illustrative numbers): if your estimate is 9,000 kWh and you produced 8,280 kWh, you reached 92% of the estimate (8,280 divided by 9,000).
6. Compare with the next year
One weak year can be weather. Two or more weak years in a row, or a steadily widening gap, deserve a closer look.
Common reasons actual production falls short
- Weather: a cloudy or smoky year. Tesla notes clouds significantly reduce production.
- Season: Tesla says winter noon output is typically closer to 50% of the system's potential, versus up to 100% in summer. See solar production by month for what is normal.
- Snow: PVWatts assumes 0% snow loss by default. A 2017 NREL report modeled annual snow losses for fixed systems in the continental US from 0% up to about 16%, highest in Minnesota and Michigan.
- Shade: trees grow. Tesla lists tree growth as a cause of reduced output.
- Dirt: heavy pollen or dust. See is solar panel cleaning worth it.
- Equipment faults: a failed inverter, microinverter or optimizer, or a tripped breaker.
- Downtime: days when the system was off for repairs, a roof job or a grid outage.
Warning signs of a real underproduction problem
Normal variation is gradual and seasonal. These patterns suggest something is wrong:
- Output drops suddenly between similar sunny days.
- A flat line of zero production during daylight.
- In a panel-level app, one panel or string consistently lower than its neighbors.
- Error lights or codes on the inverter.
- Annual output below the estimate by more than the normal range, for two years running, with no unusual weather.
Production guarantees and your warranty
Production guarantees are contract terms, not a law, so read yours. One example: Palmetto's help center says its Energy Plan guarantee requires at least 90% of estimated output over each 36-month review period, pays a bank deposit for any shortfall at the agreement's energy price, and may exclude periods like grid outages, major weather events or repair downtime. Other companies use different percentages and periods.
Separate from guarantees, panels have performance warranties. Qcells' datasheet for its Q.TRON BLK M-G2+ panels, for instance, promises at least 90.58% of nominal power at 25 years. A panel warranty claim needs proof that a specific panel lost power, which usually requires testing by a professional. Our guide to product vs performance warranties explains the difference.
If you see a real shortfall, document monthly totals, screenshots and any error messages, then contact your installer or the company that holds your contract. If that company is gone, our solar orphans resource explains who can still help.
Next steps
Pull last year's production total today and divide it by your estimate. If you do not have an estimate, our solar panel calculator gives a rough benchmark for your area. If the gap is large and persistent, have a qualified technician inspect the system.
Frequently asked questions
How accurate are solar production estimates?
The PVWatts Version 5 manual says estimates based on typical weather may be off by up to about 10% for a year and 30% for a month. A specific year can differ more because weather varies.
Why is my solar system producing less than estimated?
Common causes are a cloudy year, more shade or snow than the model assumed, dirt, panel aging and equipment faults. A sudden drop usually points to a fault.
How much should solar panel output decline each year?
A 2012 NREL review of nearly 2,000 measured rates found a median degradation of 0.5% per year. Your panel's warranty states the exact rate it guarantees.
Can I use my net meter to check solar production?
Not fully. SDG&E explains that a net meter shows only what flows back to the grid, not total production. A separate generation meter or your inverter data gives the full number.
What if my installer guaranteed production and I am below it?
Read the guarantee terms for the percentage, review period and exclusions, then contact the company with your production records. Some guarantees pay automatically for shortfalls.
Sources
- NREL – PVWatts Version 5 Manual (2014, archived copy)
- OSTI – Photovoltaic Degradation Rates: An Analytical Review (Jordan and Kurtz, 2012)
- NREL – Integration, Validation, and Application of a PV Snow Coverage Model in SAM (2017)
- Tesla Support – Understanding System Performance
- Tesla Solar Owner Manual – System Performance and Maintenance
- Palmetto Help Center – Understanding your Palmetto Solar Performance Guarantee
- Dominion Energy – Understand My Meter
- SDG&E – Smart Meters and Net Metering
- Qcells Q.TRON BLK M-G2+ datasheet
Questions and comments
Be the first to ask a question or share your experience.