PVOUT vs Peak Sun Hours: Why High PSH Doesn't Mean High Output

Two metrics, two different questions

PSH divides the day's total irradiance on a horizontal surface by 1,000 W/m²: a desert day delivering 7.5 kWh/m² is "7.5 peak sun hours" — a resource number, not an energy forecast.

PVOUT starts from the same irradiance and subtracts what the system loses: cell heating, soiling, inverter and wiring losses, mismatch, downtime. It is an asset number — how much energy the meter recorded. PSH is the theoretical ceiling of the resource; PVOUT is the physical reality of the yield.

The Douglas, Arizona anomaly

In arid zones like Douglas, Arizona, monthly output data shows a pattern that surprises buyers: net generation hits its annual minimum in July, the peak of summer. The annual peak falls in spring instead, March through May: bright months before the heat, dust, and monsoon penalties arrive. Four mechanisms compound in July:

1. Cell temperature losses. Crystalline silicon (c-Si) cells carry a negative temperature coefficient, typically -0.3%/°C to -0.5%/°C. As cell temperature rises above the 25 °C reference of standard test conditions, open-circuit voltage (Voc) falls; short-circuit current (Isc) rises slightly but cannot offset the loss, so maximum power (Pmax) degrades. Run the arithmetic: a cell at 65 °C sits 40 °C above the reference, and -0.3 to -0.5 %/°C translates into a 12–20% power loss against the nameplate.

2. Inverter thermal derating. The DC output passes through string or central inverters. July ambient temperatures push the enclosure's heat dissipation to its limit, and to keep switching devices such as IGBTs (insulated gate bipolar transistors) out of thermal runaway, firmware engages over-temperature protection curves and clips output in the hottest hours — exactly when PSH is highest.

3. The monsoon changes the light. From July through August the American Southwest shifts into the North American Monsoon, and the irradiance composition changes: more diffuse horizontal irradiation (DHI), less direct normal irradiation (DNI). Total irradiance stays high; the low-loss, directly usable share does not.

4. Dust cements itself onto the glass. Convective winds ahead of monsoon rain raise haboobs, severe dust storms that coat module glass in fine particulate. The first rain turns that dust into a mud crust that later rain cannot rinse off, and soiling loss jumps in the weeks when generation should peak.

Why this matters for off-grid sizing

Our work is off-grid solar lighting, and this distinction decides whether a light stays on in the worst month of the year:

• Design to the worst month, not the annual average. A battery bank sized on annual-average PSH under-charges all winter. We size the array so the design month (the shortest, cloudiest month) closes the energy balance on its own.
• Apply a temperature derate in hot climates. The same -0.3 to -0.5 %/°C coefficient that eats July output applies wherever cells run hot. In cold climates the effect reverses: voltage rises as temperature drops, which is why MPPT controllers earn their cost in winter.
• Use PVOUT datasets for yield; PSH for screening. Public PVOUT databases such as PVGIS and the Global Solar Atlas return monthly kWh/kWp for a location, and that is the figure a battery calculation needs. PSH maps answer the first-pass question: is solar plausible here at all.
• Treat soiling as a line item, not a footnote. In dust-prone regions we add a soiling allowance and specify a cleaning interval. On highway and desert projects that allowance is routinely the difference between meeting and missing the autonomy target.

What this page does not cover

This page explains the two resource metrics and why they diverge in hot seasons. It does not cover financial yield forecasting (P50/P90), bifacial gain, tracking, or grid-tied plant design. To size an off-grid lighting system, our energy balance calculator takes monthly irradiance and returns array and battery sizes; the panel sizing calculator runs the array side.