Why Is My Solar Production Lower Than Expected?
Last updated: 2026-04-06 · Solar Benchmark
Why Is My Solar Production Lower Than Expected?
The most common reason a solar system produces less than expected is shading that wasn't accounted for at installation, including trees, a neighbor's addition, or a rooftop obstruction that grew over time. Soiling (dust and debris on panels) is the second most common cause. A system producing 15–25% below its physics-based benchmark for its location and size almost always has one of six identifiable causes.
The Six Most Common Causes, Ranked by Frequency
| Cause | Typical Production Loss | How to Identify |
|---|---|---|
| Soiling (dust, pollen, debris) | 3–10% annually | Visible buildup on panels; drops worse in dry seasons |
| Shading (trees, obstructions) | 5–25% depending on degree | Production drops at specific times of day, consistent pattern |
| Inverter issue (fault, clipping, aging) | 5–30% | Monitoring app shows fault codes or unexplained gaps |
| Panel degradation beyond normal rate | 0.5–2% per year | Gradual multi-year decline exceeding 0.7%/year |
| Wiring resistance or connection problem | 5–15% | Consistent underperformance across all conditions |
| Panel damage (physical or cell-level) | Variable, 3–20% | Post-storm production drop; thermal hot spots visible to IR scan |
(Source: pvlib physics modeling, NREL field performance analysis)
How to Diagnose Each Cause
Soiling
Pull your monthly production data for the last two years. If you see a pattern where production peaks in spring and then declines through summer, but if the decline doesn't match the normal seasonal reduction for your region, soiling is likely. Dry climates (California, Arizona, Texas west of I-35) are most affected. Clean the panels and check whether production recovers. If it jumps 5–10% after cleaning, soiling was the cause.
Shading
Look at time-of-day production graphs. Shading creates a characteristic dip at a specific time each day that worsens through spring and fall as sun angles change. A system producing well at 10am but sharply dropping at 1pm in October is almost certainly shaded in the afternoon. Check for newly grown trees or new structures near the array.
Inverter Issues
Your monitoring app should show inverter status. Look for fault codes, error logs, or gaps in production data. A string inverter operating in "safe mode" after a grid event may reduce output by 20–30% without triggering a visible alarm in some systems. Check the physical inverter display if your monitoring app is inconclusive. Microinverter systems show panel-level data; look for any panel showing zero or persistently low output.
Degradation
Calculate your system's specific yield (annual kWh / system size in kW) for each year since installation. Normal PERC panel degradation is about 0.5%/year. A system declining at 1.2%/year over multiple years has a problem beyond normal aging. Accelerated degradation is more common in extreme-heat climates and in first-generation thin-film panels.
Wiring and Connection Problems
Wiring issues show up as consistent underperformance across all weather conditions, not just in extreme heat or at certain times of day. A wiring resistance problem reduces output by a fixed percentage regardless of irradiance level. If your system underperforms equally in spring and summer, wiring is a more likely cause than shading or soiling.
Panel Damage
Physical damage (hail cracks, delamination, cell-level failures) often isn't visible from the ground. An infrared scan during peak production hours reveals hot spots where cells have failed. After any significant hailstorm, track monthly production for two months. A permanent step-change drop of 3–15% after a storm event points to panel damage.
When to Contact Your Installer
Contact your installer when:
- Production drops suddenly by more than 10% month-over-month without an obvious weather or seasonal explanation
- Your monitoring shows fault codes that persist more than 48 hours
- You see production gaps (periods of zero output during daytime) that last more than one day
- Your system is less than 5 years old and specific yield is already 20%+ below regional benchmarks
Before calling, document: current production vs. same month last year, any recent events (storms, construction nearby, tree trimming), and your inverter fault code history if available.
Frequently Asked Questions
Q: My solar production dropped this month compared to last year. What should I check first?
A: Check three things in order: (1) Was this month's weather significantly cloudier or shorter on sunlight than last year? Compare your production to regional benchmarks for the month, not just year-over-year. (2) Are panels visibly dirty? Soiling is the most common reversible cause. (3) Does your monitoring show any inverter faults or zero-production periods? If weather explains the drop, it's not a system problem. If the shortfall persists after cleaning and weather was normal, investigate further.
Q: How much should my solar system produce compared to what my installer said?
A: Installer estimates typically use Typical Meteorological Year averages. In any specific year, actual weather diverges from TMY averages, causing real production to differ from estimates by 10–25%. A valid comparison requires a physics model using actual historical weather at your address, not the installer's estimate.
Q: My system is 7 years old and seems to produce less than when it was new. Is that normal?
A: Some decline is expected. PERC panels lose roughly 0.5%/year, so a 7-year-old system should produce about 3.4% less than its first year, not 15–20% less. If your decline exceeds 1%/year consistently, something beyond normal aging is reducing output. Pull annual production data and calculate year-over-year decline percentages.
Q: How do I get an independent benchmark to compare against my actual production?
A: A proper benchmark requires your system's actual production data and a physics model using ERA5 actual hourly weather data at your address. This gives a month-by-month expected production figure accounting for real weather, not long-run averages. Learn more at /resources/methodology.
Data: pvlib physics modeling + Open-Meteo ERA5 weather data | Last updated: 2026-04-06 | Solar Benchmark