What Solar Underperformance Really Costs — and Why It Stays Hidden

There is no single study that says “faults cost X%.” The honest number is a composite: several independently documented loss buckets that stack. Taken together, they put avoidable underperformance in the high-single-digit to low-double-digit range of a system’s output — and in Pakistan’s dusty, high-soiling conditions, at the upper end of it.

The most important shift in the data is this: string-level faults are now the single largest loss category — about 27% of all observed loss in 2025, and rising. That matters because it is exactly the loss a plant-level view is worst at showing. An inverter app reports the inverter’s total; one dead string out of twenty is diluted into a number that still looks “roughly normal,” so it never raises a flag.

The problem compounds across brands. Any portfolio assembled over time or across different installers inevitably spans several inverter makes — the global market is now so fragmented that, outside the top two vendors, no single brand holds even 5% share. Each maker’s app is hardware-locked and only shows deep string-level detail for its own units, so an owner ends up juggling several dashboards, none of which sees the whole plant. The faults hide in the gaps.

The value of watching is measurable. Sites inspected once a year carry roughly 7% loss; sites inspected five times a year carry about 3%. Continuous monitoring is simply the logical extreme of that curve — the more often problems are caught, the less energy is lost between catches. Separately, the availability convention is telling: a default 3% availability loss falls to around 0.5% under active monitoring, so watching a plant is conventionally worth about 2–2.5 points of availability on its own.

None of this requires new hardware. The loss is already happening inside data the inverters already produce; it just needs an independent layer that reads every string, every brand, all the time — and tells you which string, on which day, started slipping.

Solar Performance Cloud (SPC) is built for exactly the losses the research says dominate. It inspects at string level — not just inverter totals — across every major inverter brand, continuously, as an independent layer with no reason to hide a problem. It grades each string against IEC 61724-1 monitoring practice and raises the specific fault (soiling, shading, a dead panel, a loose cable, a sensor fault), so the underperformance that would otherwise surface at the annual review surfaces the same day instead.

The boundary is honest: SPC recovers energy by making loss visible early; it does not manufacture generation. For commercial and industrial solar — where a few points of avoidable loss is real money every month — early visibility is the whole return.