Performance ratio from measured yield

The performance ratio is the standard yardstick of PV system health — the quantity monitoring practice, IEC 61724-style, is built around. It compares delivered energy against a reference: what the same nameplate would have yielded from the same sun if conversion were perfect. Because that reference already contains the site’s solar resource, the ratio measures the SYSTEM, not the weather — a cloudy region and a desert can both host an installation that scores well.

Within this cluster it is the closing of a loop. The sizing page divided by an assumed ratio; the yield page multiplied by the same assumption; this page derives the real one from a meter reading, using algebra the estimate would recognise — the same product of nameplate, sun hours and days, now sitting under the measured energy instead of beside an assumed factor. Setting the derived value beside the assumed one is the single most informative comparison a system owner can make.

What the ratio cannot do is say WHY. Elevated cell temperature, soiling, a string down for a month, inverter losses, creeping degradation and plain shading all pull it down through the same arithmetic, so a disappointing value is a flag, not a diagnosis. Its job is to say “look closer” cheaply and defensibly; the pack’s derate calculators — temperature, shading, soiling — are where the looking happens.

Because the sun is divided out, the ratio travels. Systems of different sizes, on different roofs, in different countries can be ranked on it directly, which raw kilowatt-hours cannot do at all and even specific yield — energy per installed kilowatt — can only do within one region’s resource. That portability is why fleet operators track it as the first-order health metric across a portfolio.

The measurement itself is unglamorous: one number from the system’s generation meter twelve months ago, one number today, and their difference. Monitoring apps will offer the same total with more ceremony. What matters is that the window is genuinely a year, that the meter is the system’s own rather than the household import meter, and that nothing in between — an inverter replacement, weeks of scaffolding shade — went unrecorded.

The recovered ratio is only as honest as the sun figure it is judged against. Use the same site peak-sun-hours value the original estimate used, not a value tuned until the answer looks respectable. The point of the measurement is to audit the assumption, and an audit that adjusts the evidence is a testimonial.

One measured year also sets a baseline that makes every future year cheap to judge: measure again next year and the comparison is ratio against ratio, with the site’s geography already cancelled out of both sides.

A disappointing measured ratio does not say what failed — it says that something did. Before blaming the hardware, interrogate the inputs: a measurement window that quietly excluded a strong month, a sun figure from the wrong side of a mountain range, a capacity figure that still includes the string that was never installed. Input errors are cheaper to find than inverter faults, and more common.

When the inputs survive interrogation, the ratio becomes a genuine diagnostic. A ratio that sagged gradually over years suggests soiling or degradation; one that stepped down suddenly points at a failed string or a curtailing inverter; one that was never respectable from the first year suggests the installation never matched its design — shading the survey missed, or an orientation compromise nobody re-modelled.

A suspiciously excellent ratio deserves the same scepticism in reverse. Ratios beyond the plausible band usually mean the reference sun figure is understated or the meter is counting something it should not. The calculator’s guards refuse the physically impossible; the merely implausible remains yours to question.

The engine below arrives pre-filled with the journey’s reference system — the pack’s own worked example, whose measured year matches what the yield lesson estimated. Substitute your own metered year and capacity, keep the sun figure honest, and the ratio that appears is your installation’s real report card.

Read the result against the band the sizing lesson called realistic: comfortably inside it, your system is healthy; below it, something specific is wrong and worth finding; above it, check the inputs before celebrating. Every figure is computed live by the verified engine.