How can I tell if my solar system is performing correctly?
Many homeowners install solar and trust monitoring without knowing what to look for. Here's a complete framework: what numbers matter, how to compare to expectation, and red flags that something's wrong.
The key metric: Performance Ratio (PR)
Performance Ratio = Actual production ÷ Expected production
- Healthy system: PR = 0.75 to 0.85
- Good system: PR = 0.85 to 0.92
- Excellent: PR > 0.92 (rare; usually means modeled expectation was conservative)
- Problem: PR < 0.70 (something is degrading or shaded)
- Critical: PR < 0.50 (major problem — investigate immediately)
Why PR isn't 1.00
"Expected production" is what your panels would produce in a lab. Real-world losses:
- Module mismatch: 1-3%
- DC wiring losses: 1-2%
- Inverter conversion: 2-4% (98-99% peak; lower at low loads)
- AC wiring: 0.5-1%
- Soiling: 2-5%/yr (varies by climate)
- Temperature: 5-15% lower output in summer in hot climates
- Shading: 0-30% depending on site
- Inverter clipping: 0-3%
- Annual degradation: 0.5%/yr typical
Total: 13-25% loss from STC nameplate. PR of 0.75-0.85 is "as designed."
How to calculate your expected production
Free tools (do this monthly)
- NREL PVWatts (pvwatts.nrel.gov): plug in your address, system size, tilt, azimuth, derate factors. Get monthly + annual expected kWh.
- Google Project Sunroof: shade-aware satellite estimate for your specific roof.
Your installer's design model
Your install contract probably included a "design model" giving expected first-year production (e.g., 13,500 kWh in year 1). Compare:
- Year 1: actual vs modeled. Within 90-100%? OK. Below 85%? Problem.
- Year 2-5: account for 0.5%/yr degradation (so year 5 = 97.5% of year 1).
- Year 10: 95% of year 1 typical. Below 88%? Investigate.
- Year 25: 87% of year 1 (12 years of degradation).
Daily monitoring routine
What to check daily (or every few days)
- Today's production matches recent days (controlling for cloud cover).
- Inverter status: green/operational. No persistent fault codes.
- Battery state of charge if applicable.
- Module-level production uniformity (Enphase Enlighten or SolarEdge per-optimizer).
Red flags during day
- One module/string consistently underproducing vs neighbors.
- Unexpected zero production during sunny periods.
- Inverter restart loops.
- Battery not charging or discharging as expected.
Weekly review
- Total kWh week vs same week last year (or vs PVWatts forecast for the week).
- Net consumption from grid: imports rising? Loads going up?
- Battery cycles: reasonable depth (30-90% typical for daily cycling)?
Monthly review
- Compare monthly kWh to PVWatts forecast for your zip code. Should be within +/- 10% in moderate-cloud months.
- Check utility bill for net consumption / export credits.
- Verify monitoring is reporting consistently: WiFi/cellular link OK? No data gaps?
Annual review (the big one)
Step 1: Total annual production check
- Last year's total: ___ kWh
- Original modeled (year 1) - 0.5%/yr degradation = expected ___ kWh
- Ratio = ___% (should be 90-110%)
Step 2: Module-level analysis
- (Enphase) Look at "Site" view: any individual module producing <90% of neighbors?
- (SolarEdge) Per-optimizer view: same check.
- (String inverter) Visit the inverter, log fault codes, look for IGBT/transformer issues.
Step 3: Visual inspection
- Walk around the array. Look for: cracked glass, brown/burn spots, dirt/debris, vegetation overgrowth, vibrating racking, cracked sealant, exposed wiring.
- (For roof installs, use binoculars from ground — don't climb on roof unless you must.)
- Check inverter LED status, time/date display, fault history.
Step 4: Battery health (if applicable)
- Capacity test: full discharge then charge, observe usable kWh.
- Check warranty cycle/throughput remaining.
- Most batteries report battery health % in app (e.g., Tesla shows 85% remaining capacity at year 5).
Step 5: Compare to baseline
- Year 1: lock in your baseline.
- Year 2: compare to year 1 - 0.5% degradation. If higher than 1% drop, investigate.
- Year 5: should be ~97.5% of year 1 baseline.
- Year 10: ~95%.
- Year 25: ~87%.
Common production shortfalls and what they mean
5-10% shortfall (year 1+)
- Optimistic install model.
- Higher-than-expected soiling or shading.
- Single module underperforming 10-20%.
- Microinverter or optimizer drop-out (1 of N missing).
- Compare your shading at install time to current (tree growth?).
10-25% shortfall
- Inverter clipping (DC oversize relative to inverter capacity).
- Major shading change (tree canopy expanded, new neighbor structure).
- Multiple module failures.
- String inverter fault.
- System hasn't been commissioned correctly.
25%+ shortfall
- Major system issue: inverter failure, multiple module damage, electrical fault.
- Service call needed promptly.
What to do when you spot a problem
- Document: screenshots of monitoring, photos of any visible damage, write down dates and observations.
- Compare: use PVWatts to confirm the shortfall is significant (not weather).
- Contact installer first. They diagnose under workmanship warranty (typically 5-10 years).
- If installer is unresponsive or out of business: see defunct installer playbook.
- If issue is at component level: contact manufacturer directly (Enphase, SolarEdge, Tesla, etc.) for warranty. See O&M / RMA guide.
- Production guarantees: if your contract had one, file claim in writing.
Annual maintenance recommendations
- $200-500/yr: hire a service technician to check the system once or twice/yr. Cost-effective vs lost production.
- Visual inspection: 1-2x/yr by you (binoculars from ground).
- Cleaning: usually unnecessary; rain handles it. Hire only if you see significant soiling and cost-justify.
- Inverter inspection: at year 10-15, expect inverter end-of-life conversation.
- Battery inspection: at year 12-15, plan for potential replacement.
Frequently asked questions
How do I know if my system is performing as designed?
Compare actual annual kWh to your installer's design model. Should be within +/- 10% in year 1, and degrade ~0.5%/yr after.
What if my installer's design was wrong?
Production guarantees in your contract usually obligate them to compensate the shortfall. Failing that, document and pursue. If installer is gone, check SolarInsure SI-30 or third-party warranty.
Should I hire someone to verify performance?
Annual performance audits cost $200-500 from independent service installers. Worth it if your contract had a production guarantee and you suspect shortfall. Otherwise, monthly self-review with PVWatts is sufficient.
How long does it take for production drift to be a real problem?
0.5%/yr is normal degradation. 1-2%/yr means problem within 5-10 years. 5%+/yr means immediate investigation.