Skip to content

How Solar Benchmark Calculates Your Expected Production

We don't guess. We model your specific system using real weather data and peer-reviewed physics — accurate to within 5–7%.

The Problem

Why Averages Don't Work

Every solar monitoring app shows you what your system produced. The hard part is knowing what it should have produced.

Most tools — including the widely-used PVWatts calculator — estimate expected production using a “Typical Meteorological Year”: a synthetic average built from 15–30 years of weather records. But no actual month matches the average month. If January had 30% more cloud cover than typical, a TMY estimate will overstate expected production by 30% or more.

January: TMY Estimate vs. Actual

-30%

TMY Estimate

Actual

740 kWh

520 kWh

Is this a real problem or just a cloudy month?
TMY can't tell you. We can.

Data Sources

Real Weather, Real Panels

We use hourly weather data from ERA5 — the world's most comprehensive atmospheric reanalysis dataset, widely used in climate science and energy modeling. ERA5 is the gold standard for historical weather data.

Your panel specs come from the California Energy Commission's publicly maintained database of over 30,000 photovoltaic modules — giving us manufacturer-specific characteristics for your exact panel model, not generic assumptions.

ERA5 Reanalysis

Hourly irradiance, temperature, and wind speed for your exact location.

CEC Module Database

30,000+ PV modules with manufacturer-specific electrical and thermal characteristics.

The Physics

Six Steps from Sunlight to Verdict

Each hour of the month runs through our physics pipeline. Here's what happens at each stage.

Step 1

Solar Position

For every hour of the month, we calculate the sun's exact position — altitude and azimuth — relative to your location using astronomical algorithms.

Step 2

Plane-of-Array Irradiance

We convert horizontal irradiance to the light actually hitting your tilted panels using the Perez transposition model — the same model used by PVsyst.

Step 3

Cell Temperature

Solar panels lose efficiency as they heat up. We model cell temperature using ambient temperature, wind speed, and your module's specific thermal characteristics.

Step 4

DC Power Output

Using your module's real electrical characteristics — efficiency curve, temperature coefficients, nameplate rating — we calculate expected DC power output for each hour.

Step 5

System Losses

We apply a calibrated loss chain: inverter efficiency, wiring and mismatch losses, soiling by region and season, and technology-specific degradation rates.

Step 6

Expected vs. Actual

We sum the hourly expected production into a monthly total and compare it to your actual data. The result is a Performance Ratio — the percentage of expected production your system delivered.

Accuracy

Why This Is More Accurate

Side-by-side: traditional estimates vs. Solar Benchmark.

Weather data
TMY
30-year average
Solar Benchmark
Actual weather, that month
Time resolution
TMY
Monthly aggregates
Solar Benchmark
Hourly calculations
Sky model
TMY
Isotropic (uniform sky)
Solar Benchmark
Perez anisotropic (realistic)
Panel model
TMY
Generic assumptions
Solar Benchmark
Your specific module from CEC
Temperature model
TMY
Basic NOCT
Solar Benchmark
Module-specific thermal coefficients
Soiling
TMY
Not modeled
Solar Benchmark
Regional and seasonal calibration
Degradation
TMY
Not modeled
Solar Benchmark
Technology-specific annual rates
Monthly accuracy vs. measured
TMY
~38%
Solar Benchmark
~5–7%

Your Report

What Your Report Tells You

Each monthly health report gives you a complete picture of your solar system's performance — in plain English, not engineering jargon.

Expected production

for the month, given actual weather at your location

Actual production

from your uploaded monitoring data

Performance Ratio

how your system performed relative to physics-based expectations

Health status

a plain-English verdict on whether your system is performing normally

Trend analysis

month-over-month and seasonal patterns

Recommended actions

when warranted, specific next steps to take

If your system is healthy, we tell you. If something needs attention, we tell you that too — with enough specificity to have a productive conversation with your installer.

Solar Health Report

March 2026

Healthy
Performance Ratio94%
Produced847 kWh
Expected901 kWh

6-Month Trend

Oct
Nov
Dec
Jan
Feb
Mar

Performance improving — up from 68% in January

Recommended Actions

No action needed — system performing as expected

Spring cleaning: consider panel wash before peak season

Our Standards

Built on Industry-Standard Methods

These are the same methods and data sources used by professional solar engineers, asset managers, and tools like PVsyst. We've made them accessible to homeowners.

pvlib

Sandia National Laboratories

Open-source photovoltaic modeling library maintained by the PV Performance Modeling Collaborative.

Perez 1990 Model

Industry Standard

The standard method for calculating irradiance on tilted surfaces, used by PVsyst and professional tools.

CEC Module Database

California Energy Commission

The most comprehensive publicly maintained database of PV module characteristics.

ERA5 Reanalysis

Global Reanalysis

The world's most comprehensive atmospheric reanalysis dataset, used globally for energy modeling and climate research.

IEC 61724

International Standard

The international standard metric for solar system performance assessment (Performance Ratio).

Independent Analysis

Solar Benchmark

No installer partnerships, no manufacturer relationships. Our only customer is you.

Your system is producing right now.
Do you know if it's performing?

Get an independent, physics-based health report for your solar system.

Free to join. We'll notify you when your report is ready.