Efficiency, Yield & Real Performance

Every PV module is rated in watts-peak (Wp) - its power output under Standard Test Conditions (STC), an internationally agreed laboratory benchmark. STC specifies an irradiance of 1000 watts per square metre (roughly bright midday sun), a cell temperature of exactly 25 degrees Celsius, and a defined light spectrum. Under those conditions, a '400 Wp' module produces 400 watts. The rating is honest, standardised and essential - it lets you compare modules and size systems. What it is *not* is a promise of the energy your building will receive.

The first thing to understand is the difference between power and energy. Watts-peak is instantaneous power under ideal light; the thing you actually use and pay for is energy - kilowatt-hours accumulated over time. A system's energy output depends on how much sun falls on it across a real year (the solar resource, orientation, tilt and shading of Module 1) multiplied by how efficiently the system converts that sun after all its losses. Two identical 5 kWp systems, one in sunny Rajasthan and one on a shaded Mumbai facade, have the same nameplate and wildly different annual energy. Nameplate tells you the size of the tap; it does not tell you how much water flows.

The second thing is that STC conditions almost never occur in the field simultaneously. A panel bright enough to hit 1000 W/m2 of irradiance is also sitting in the sun, so its cell temperature is far above 25 degrees - often 45 to 65 degrees or more, especially a poorly-ventilated BIPV panel pressed against the building. Since PV output falls as cells heat up, a module in real sunshine is already producing below its STC rating at the very moment it is brightest. The laboratory's cool spotlight is a condition the working panel rarely enjoys.

So the nameplate is a starting point, not a result. A literate designer treats it exactly as it is: a standardised comparison figure and a sizing input, from which real delivered energy is obtained by applying the site's actual sun and a chain of real-world losses. Anyone who quotes the nameplate, or the simple sum of the nameplates, as the energy a building will generate is - knowingly or not - over-claiming. The honest number is always downstream of the label.

Between the module rating and the energy at the meter lies a chain of real, largely unavoidable losses. Each is modest on its own; multiplied together they explain why delivered energy sits well below what the nameplate seems to promise. Walking the chain is the single most useful thing a designer can do to read solar honestly.

Temperature (heat). As covered above, cells run far hotter than STC's 25 degrees, and output drops with temperature - a real and continuous loss in hot climates, and worse for BIPV panels with little ventilation behind them. In an Indian summer this alone can cost a meaningful slice of output. Soiling (dirt). Dust, pollen, bird droppings and pollution settle on modules and block light; in dusty Indian conditions, unwashed panels can lose several percent or more, recovered only by cleaning. Shading. Even partial shade is disproportionately costly because cells wired in series limit each other - a shaded corner can drag down a whole string. Inverter losses. Converting the panels' DC to usable AC is efficient but not perfect; a few percent is lost in the inverter, plus more if it is poorly sized or runs outside its sweet spot. Wiring and mismatch. Resistance in cables and small differences between modules (mismatch) shave off more. There are also reflection, spectral and low-light losses at the module itself, and downtime when equipment faults.

No single loss is dramatic, but they compound: a chain of five or six factors each costing a few percent multiplies down to a system that delivers perhaps three-quarters to four-fifths of what a naive nameplate calculation suggests. This is not pessimism - it is the standard, expected behaviour of every real PV system, and good engineering plans for it rather than being surprised by it.

For the designer, the practical lessons are concrete. Ventilate PV where you can, because heat is a continuous tax - a real argument for a gap behind BIPV. Keep modules cleanable, because soiling is recoverable but only if you can reach them. Avoid shading ruthlessly at the design stage, because it is the most damaging and least recoverable loss. And never present a nameplate-based figure as delivered energy. The binding, quantified version of this loss chain - the actual derating factors for a specific system and site - belongs to the engineers' yield model and the manufacturers' data; the designer's job is to understand that the chain exists, design to minimise it, and read every yield figure knowing it is somewhere on that chain.

The industry has a clean way to capture the whole loss chain in one figure: the performance ratio (PR). PR is the ratio of the energy a system actually delivers to the energy it would deliver if it ran at its nameplate efficiency with no losses at all - in effect, the fraction of the theoretical maximum that survives heat, dirt, shade, inverter, wiring and the rest. A well-designed, well-maintained system typically achieves a PR of roughly 0.75 to 0.85; a poor or badly-sited one is lower. When you see a PR, you are seeing the honesty of the whole system compressed into a single number, and a PR near or above 0.8 signals a system doing its job well.

A companion figure makes yield comparable across systems: specific yield, the energy generated per unit of installed capacity - kilowatt-hours per kilowatt-peak per year (kWh/kWp/yr). Because it divides out system size, specific yield lets you compare a rooftop in Chennai with a facade in Delhi on equal terms, and it exposes underperformers: two systems of the same nameplate with very different specific yields are telling you something about their siting, orientation, shading, heat or maintenance. The related capacity factor expresses the same idea as the fraction of the year the system effectively runs at full power - low for solar (the sun is not always up), and a useful reality check against anyone implying a panel generates around the clock.

These metrics are the sceptical reader's tools. Faced with a yield claim, ask: what PR does this assume, and is it plausible (a claim implying a PR of 0.95 is fantasy)? What specific yield does it imply, and does that match the region's known range? Is the figure the year-one output or a lifetime average (degradation, next section, lowers the average)? Is it modelled or measured? A claim that cannot express itself in these terms - that offers only a big watts-peak number or a raw annual figure with no PR, no specific yield, no measurement basis - is a claim asking to be taken on faith.

None of this is the designer's binding calculation - the certified PR, yield model and derating factors belong to the engineers and verified data. But understanding PR and specific yield turns you from a passive recipient of glossy numbers into someone who can interrogate them, spot the implausible, and design toward a high PR rather than merely hoping for one.

One more honesty is missing from a single-year snapshot: PV modules degrade. Their output declines slowly but relentlessly over their working life, typically on the order of half a percent per year, so a module producing 100 percent in year one delivers perhaps 80 to 85 percent by the end of a 25-year warranty. This is why manufacturers offer *performance warranties* that guarantee a minimum output at, say, year 25 - an admission, in the fine print, that the panels you install today will generate less in two decades than they do now.

Degradation reshapes any honest yield or payback figure. A claim quoting year-one output overstates the lifetime average, because the system spends most of its life below its opening figure. A serious energy or financial model uses the *lifetime-average* yield - the area under the slowly-declining curve - not the shiny first-year number. When you read a payback or net-zero claim, ask whether it accounts for degradation; many optimistic figures quietly do not, and the difference compounds over a long life.

Put all of this together and you have a discipline for reading solar numbers sceptically - not cynically, but carefully. Confronted with any yield figure, run it through a short interrogation. Is it nameplate or delivered - a watts-peak rating or actual energy? Is it measured or modelled - metered reality or a simulation? Does it embed a plausible performance ratio, or does it implicitly assume near-perfect conversion? Is it year-one or lifetime-average - does it account for degradation? And is it stated as a share of demand, or as a raw number designed to impress? A figure that answers these cleanly is trustworthy; one that dodges them is marketing.

This scepticism is the through-line of the whole module. Solar-washing (9.1) survives only where numbers go unexamined; the honest go/no-go (9.2) depends on realistic yields, not hopeful ones; and the end-of-life question (9.4) begins the moment degradation makes a module not worth keeping. The designer who reads solar numbers this way designs better systems, makes honest claims, and is vindicated by the meter rather than embarrassed by it. As always, the binding yield, PR, degradation and warranty figures are the manufacturers', the engineers' and the verified data's - the designer's job is to demand them, understand them, and never mistake the label for the result.