A solar panel inspection is the search homeowners reach for when the bill is still high, the app looks vague, and nobody at the installer calls back. It is a reasonable instinct. It is also only half the answer, because there are two different questions hiding inside "something is wrong with my panels," and a person on the roof can only answer one of them.
The first question is is the equipment sound? Connectors, wiring, mounting, glass, the inverter. That is a physical inspection, and it needs someone up there. The second question is is the system producing what I was promised? That is a data question. It is answered by comparing the kilowatt-hours your system made against what a system of your size, on your roof, under the weather you had, should have made. Nobody needs to climb anything to answer it, and it is the one that tells you whether there is a gap worth pursuing.
This page walks through both, what each costs, who can do them independently, and the order to do them in.
What a physical solar panel inspection checks
A competent inspection of a residential array takes an hour or two and covers the things that fail or drift after installation. The items below are what a good inspector's checklist looks like, whether they are a licensed electrician, a NABCEP-certified technician, or a home inspector with a solar add-on.
| Area | What they look for | Why it matters |
|---|---|---|
| Mounting and flashing | Loose rails, lifted flashing, sealant failure, penetrations without proper boots | Roof leaks are the expensive solar failure nobody thinks about until the ceiling stains |
| Panel glass and frames | Cracks, delamination, snail trails, hot spots, browning of the backsheet | A cracked panel can keep reporting while producing a fraction of its rating |
| Wiring and connectors | MC4 connectors not fully seated, chafed conductors, UV-degraded insulation, rodent damage | Loose connectors are a top cause of both lost output and fires |
| Microinverters or optimizers | Units reporting zero or intermittently, error codes, corrosion at the DC input | One dead unit on a 30-panel roof is a silent 3% loss |
| Inverter or gateway | Status lights, fault log, firmware, grounding, ventilation, clipping behavior | The fault log often records what the homeowner never saw |
| Shading and soiling | New tree growth, chimney or vent shadows at key hours, pollen, bird droppings, leaf packs | A shade change since install year explains many "it used to be fine" cases |
| Electrical | Breaker sizing, disconnects, labeling, rapid shutdown, string voltages, IV curves on string systems | Code and safety, plus the one measurement that finds a weak string |
Thermal (infrared) imaging and IV-curve tracing are extras. They add to the price and they find faults the eye cannot.
A physical inspection is good at what it does: it finds broken things. It does not produce a verdict on whether your system, taken as a whole and over a year, delivered the number in your contract. An inspector can tell you a connector was loose. They cannot tell you that your 9.6 kW system produced 9,100 kWh last contract year when it should have produced 10,400, and that the 1,300 kWh gap is worth roughly $390 at your rate. That second sentence is what you bring to an installer, and it comes from the data, not the roof.
What an independent data audit checks
An audit of your production history asks one question: did the system make what it was supposed to? To answer it you need four things, and the reason many homeowners never get an answer is that they are only ever given one or two of them.
- The promised number. Your contract or proposal has an annual kWh figure, and often a guarantee clause around it: a percentage threshold, an assessment period, and a cure or payment term. Our guide to solar production guarantees explains where those clauses hide and what they commit the installer to.
- The actual number. Your Enphase (or other vendor) account has the production record. This is the installer's own vendor's data, which is why it is the number they cannot credibly argue with.
- An independent expected number. Not the installer's estimate, which was built to sell the system, and not last year, which had different weather. A model of your specific array: panel count, wattage, tilt, orientation, install date, and the sunlight your address received. The public model for this is NREL's PVWatts, and we explain how OwlWatt uses it for your roof.
- The comparison, by contract year. Because that is how the guarantee is assessed. A shortfall in kWh, then in dollars at your own electric rate, for each year the guarantee has run.
Done that way, the audit separates three things that look identical in the app: a system that is fine and just had a cloudy year, a system with a physical fault, and a system that was oversold. The first needs nothing, the second needs an inspection with a target, and the third needs a letter with the audit report as the attachment.
Why you may need both
Two common cases show where each one earns its keep.
The output is low and something is broken. A homeowner sees their annual total drift down for two years. An audit shows the shortfall is concentrated in one section of the roof and started on a specific date. A physical inspection finds corroded connectors on that section. The audit gives the claim its figure, the inspection gives it its cause, the workmanship warranty covers the repair, and the production guarantee covers the lost kWh. Both documents go in the letter.
The output is low and nothing physically finds a fault. The forum thread every solar owner eventually finds: a homeowner produces 7,800 kWh in year one against a promised 9,996, a 22% gap, on hardware that reports panel by panel. Experienced posters ask the right questions: was the array one orientation or split across several, had the shading been assessed including trees that were smaller when the contract was signed, was the guaranteed number even modeled for that specific roof. Nobody in the thread has an independently modeled expected number for that roof, so the thread never settles whether the gap is a real shortfall or a proposal that was wrong from the start. An audit exists to answer that question, by computing the expected number instead of arguing about it, and no roof visit would have found it either way.
The order matters. Audit first: cheaper, no appointment, and it tells you whether there is a gap and how large. Inspection second, if the audit says so, with the report in the inspector's hand.
What "independent" means, and how to check for it
Independence gives an inspection or an audit its value, and it is the thing most often missing. Your installer will inspect your system for free under the workmanship warranty. That inspection is worth having, and it is not independent: the inspector's employer is the party who would pay a claim.
Questions to ask anyone before you pay them:
- Do you sell, install, or get referral fees for solar equipment? If yes, they have a reason to find a repair rather than a shortfall.
- Will I get a written report with the measurements, or a verbal summary?
- For a data audit: what model produces the expected number, and will you list every input so my installer can check it?
- Will you ever contact my installer without my say-so?
OwlWatt's answers, for the record: we sell no panels, install nothing, take no referral fees or commissions from any installer or inverter vendor, the expected number comes from PVWatts with every input listed, and we never contact your installer. The report is yours to send or not send.
What you can do without anyone on the roof
Most of the useful checks are available to you today, from the ground and from your phone. Do them in this order.
1. Look, from the ground
On a bright day, walk the property. Heavy pollen or bird soiling, a branch that has grown over one corner of the array since install, visible cracked glass, a panel that looks a different color from its neighbors. Note what you see with a photo and a date. Do not go on the roof and do not open any electrical box.
2. Read the status lights
The gateway in your garage or basement has a legend in the manual. Steady green is normal. Amber, red or a blinking pattern maps to a listed fault. No lights at all points to a tripped breaker.
3. Check every panel is reporting
Microinverter and optimizer systems show a roof layout with one tile per panel. In daylight, every tile should carry a number. A grey or blank tile is a panel that is not phoning home and, on most systems, not producing. Our guide to panel-level monitoring shows what a healthy layout looks like.
4. Compare your annual total to an expected number
This is the check that answers the real question. Take your last twelve months of production from the app and compare it to what a system your size in your state should make. The free production estimator gives you a modeled range from six dropdowns and a ZIP code. If you are inside the range, the panels are probably fine. Well under it, and the next step is a proper audit with your real roof geometry and real weather.
The bill is a fifth witness, outside the app: if the app claims a healthy export and the bill shows no credit, one of them is wrong. The common reasons an electric bill stays high after solar are worth a read before you call anyone.
What each one costs
| Option | Cost | Independent? | Answers |
|---|---|---|---|
| Installer's warranty inspection | Free under workmanship warranty | No | Is something broken (that they are willing to find)? |
| Independent electrician or NABCEP technician | Varies by region and roof access. Thermal or IV-curve extra | Ask directly | Is something broken? |
| OwlWatt Solar Performance Audit | $79 one-time | Yes | Did it produce what you were promised in the most recent contract year, in kWh and dollars? (Earlier years appear as a range, and a subscription shows all of them.) |
Inspection prices vary by region and roof access. Get quotes locally. The audit price is fixed.
When a low number is not a problem at all
A fair share of audits find that the system is fine. A north-facing section, a hot summer, a cloudy spring, normal degradation of about half a percent a year, and a proposal built on a sunnier average than the year you got. All ordinary, and a correctly modeled expectation accounts for it. If your actual lands within a few percent of that expectation, there is no claim and no fault, and knowing so is worth the $79. It ends the question, and the report is worth keeping if you ever sell the house.
Start with the data. Send the inspector in second.
An independent solar performance audit for $79: exact expected versus actual for your most recent completed contract year, shortfall in kWh and dollars, PVWatts-based, with month-by-month charts verifiable against your own Enphase records. Earlier years shown as a range. Enphase systems today. No installer or vendor ties, and we never contact your installer.
See the audit → Or run the free estimator first
Owl