System efficiency

What it covers

Real installations always deliver less power than the panel ratings suggest, because:

• Inverter conversion losses (~3–5 %)
• DC and AC wiring losses (~1–2 %)
• Soiling — dust, pollen, light bird droppings (~2–5 %)
• Mismatch losses across panels in a string (~1–2 %)
• Inverter clipping when DC array exceeds AC output rating

Note: as of April 2026 solarprod models panel temperature derate separately, using the standard silicon coefficient of −0.4 %/°C off the 25 °C STC reference and a Nominal Operating Cell Temperature of 45 °C. This means System efficiency now represents structural losses only (inverter, wiring, soiling, mismatch, clipping) — the heat penalty on a 35 °C summer day or the cold-panel bonus on a clear 0 °C morning is computed automatically from the real ambient temperature fetched alongside the cloud-cover forecast. You don’t enter it.

Multiplied together, structural losses for a typical residential install run ~88–94 % of STC nameplate (so System efficiency around 0.90–0.92 is accurate); the temperature term then derates further on hot days and lifts on cold days.

What to enter

Now that temperature is modelled separately, these recommended values represent structural losses only:

• 0.92 — new install, modern string or micro-inverter, low soiling, well-tuned DC sizing.
• 0.88 — sensible default for a clean, well-installed residential array.
• 0.82 — older install (10+ years), some accumulated soiling, mid-range inverter.
• 0.72–0.78 — significant soiling, partial shading, or known under-performance.

The form’s default of 0.85 is kept for backward compatibility with existing saved layouts — under the new model that’s slightly conservative for a typical install. Bumping it to 0.88–0.92 will match new installs more accurately; existing users who have already calibrated against measured data should leave it alone, since their tuning will already implicitly include some temperature term that’s now double-counted.

The most accurate way to dial this in for your specific roof is still to compare measured production against solarprod on a cool, clear, sunny day in the past — for past dates the calculator pulls real reanalysis weather (no forecast uncertainty) and the temperature derate is now computed from real ambient temps, so any gap between measured and predicted kWh is almost entirely your structural System efficiency.

Pick a recent clear day where your monitoring shows a clean dome-shaped curve with no shading dips, run solarprod for that exact date with the toggle on, and compare totals. If solarprod predicts 28 kWh and you measured 25 kWh, multiply your current efficiency by 25/28 (= 0.89) and re-run — e.g. 0.85 × 0.89 = 0.76. Repeat across two or three different clear days to confirm the figure is stable, then use that as your go-forward setting. Cool days matter because the temperature-derate term is small at low ambient temps, so what you’re measuring on those days is closest to your installation’s structural efficiency (inverter + wiring + soiling + mismatch).