A solar system does not usually fail. It fades. Output slips a few per cent, then a few more, and because the monitoring app still shows green ticks and the bills still look lower than they were, nobody asks whether the number is the right number. Systems installed ten or fifteen years ago, under the Feed-in Tariff, are now the ones most likely to be quietly short, and the installer who would once have noticed has often gone.
The six usual causes, in the order worth checking
1. The inverter. The box that turns the panels’ direct current into mains electricity is the component most likely to fail, and it fails in two ways. Outright, which you notice because generation stops. Or quietly: it trips on hot days, restarts, and loses an hour of the best sun; or it cannot handle the full output of the panels and clips the peak. Inverters have a design life of ten to fifteen years. If yours is older than that and has never been replaced, start here.
2. A string or a panel that has stopped. Panels are wired in groups, called strings. A failed connector, a damaged cable or one dead panel can take a whole string out, and on a small system that is a third or a half of the array. If your monitoring only shows the whole system, you cannot see this. A step change in output from one month to the next, unrelated to the weather, is the signature.
3. Shading that was not there when it was installed. Trees grow. A neighbour builds an extension or puts up a flue. Shade on even one panel in a string can pull down the whole string, far more than the shaded area suggests. Compare a photo of your roof today with the installer’s original drawing or survey.
4. Dirt. Less than people think in most of the UK, because rain does most of the cleaning, but moss, lichen and bird droppings along the lower edge of a shallow roof, or under trees, can cost a few per cent and do not wash off.
5. Degradation. Panels lose output as they age, typically a fraction of a per cent a year. Over fifteen years that is several per cent, and it is normal. The warranty states a maximum rate; what is not normal is losing more than the warranty allows, which can be a claim.
6. The monitoring itself. A generation meter cannot lie, but an app can. Monitoring that relies on a clamp sensor, a wireless link or a cloud service that the installer no longer pays for can show nothing, or show the wrong thing. If the app and the meter disagree, believe the meter.
A ten-minute test from your own readings
You need two numbers: the size of your system in kWp, which is on the installation certificate or the FiT paperwork, and how much it generated over the last twelve months, which is the difference between two generation meter readings a year apart. FiT owners have been writing these down for years.
Divide the annual kWh by the kWp. In central Scotland, a well-placed south-facing roof produces roughly 800 to 900 kWh for every kWp a year; in the south of England, 950 to 1,050. East- or west-facing roofs produce about 15 to 20 per cent less; north-facing much less. Allow a few per cent off for age.
If your number sits in that range for your part of the country and orientation, the system is probably fine. If it is 15 per cent or more below, something in the list above is costing you money, and it is worth finding out which. If you have readings for several years, look at the trend: a steady slope is degradation, a step is a fault.
What to do next
Dig out your last twelve months of generation readings and do the division above. If the answer is low, or you cannot tell, that is exactly what the Health Check is for: a proper expected-versus-actual analysis from real weather data for your postcode, a diagnosis of where the gap goes, and a note of whether a warranty claim is available. It is desk-based and £150.