Peak Sun Hours: What the Number Means and How to Get Yours

Every off-grid sizing calculation eventually needs one number: how much sun does the array actually see, per day, at this location, at this tilt? That number is usually called peak sun hours, and it gets misused constantly — quoted from city-average tables that ignore tilt and season, or treated as a single fixed constant for a whole state.

Peak sun hours is not a measurement of weather. It is a unit conversion applied to a real measurement: solar irradiation. Understand the conversion and you can generate a trustworthy number for your own site instead of borrowing someone else’s.

What peak sun hours actually means

Solar radiation reaching a surface is measured as irradiance — instantaneous power per unit area, in watts per square meter (W/m²). It swings from zero before sunrise to a peak somewhere near solar noon and back to zero after sunset, following the sun’s position and the day’s cloud cover.

Integrate irradiance over a period of time and you get irradiation (also called insolation) — energy per unit area, in kilowatt-hours per square meter (kWh/m²). This is what a pyranometer or a satellite-derived dataset actually reports for a given day: not a peak value, but a total.

The PV industry defines a reference irradiance of exactly 1,000 W/m² — the intensity used for Standard Test Conditions, the same reference a module’s nameplate rating is measured against. Because 1,000 W/m² is equal to 1 kW/m², one hour at that reference intensity delivers exactly 1 kWh/m² of irradiation. That equivalence is the entire definition:

1 kWh/m² of daily irradiation = 1 hour at the 1,000 W/m² reference intensity
                               = 1 peak sun hour

A site that receives 5.2 kWh/m² of irradiation over a day has, by definition, received 5.2 peak sun hours — not because the sun was ever literally at 1,000 W/m² for exactly 5.2 hours, but because the day’s real, variable irradiance curve integrates to the same total energy as 5.2 hours at that reference intensity would. Peak sun hours restates a quantity you already have — daily irradiation in kWh/m²/day — as a number of hours, which is a genuinely useful trick: it turns array output estimation into simple multiplication.

Daily DC energy (kWh) ≈ Array DC rating (kW) × Peak sun hours × Derate factor

The derate factor is the combined loss fraction from soiling, shading, wiring, and the rest — see our PV system losses reference page for how that number is built.

Why the worst month, not the annual average, sizes an off-grid system

Peak sun hours vary by month — dramatically, at higher latitudes. A site might average 6.5 peak sun hours in June and 2.5 in December. If a system is sized against the annual average, it will be comfortably oversized for eight months of the year and running a deficit for the other four.

A grid-tied system can absorb that swing, because the grid makes up the difference in the low months and the system simply exports less or more depending on the season. An off-grid system has no such backstop. Every kilowatt-hour it does not generate has to come from stored battery capacity, a generator, or reduced loads — there is no fourth option. Sizing to the annual average guarantees the system underperforms specifically during the months when demand for heating, lighting, and battery-safe charging is often highest.

This is why off-grid sizing convention uses the worst month — the single lowest-production month at the site’s chosen tilt and azimuth — as the design point, not the annual average. A system sized to cover its loads in the worst month has surplus capacity for the rest of the year, which is the correct direction to be wrong in. A system sized to the average is undersized for roughly half the year by definition.

Why this page does not include a city lookup table

Peak sun hours depend on tilt and azimuth, not just location. The same city produces a different peak-sun-hour figure for a roof-pitch, south-facing array than for a flat horizontal surface, and a different figure again for an array facing east or west. A table of numbers with no stated tilt and azimuth is answering a different question than the one being asked, and most of the tables circulating online do not state their assumptions clearly enough to know which question they answered.

We also could not verify third-party city-by-city compilations against NREL’s own primary solar resource data — NREL’s relevant tools were unreachable during this page’s research, and we are not willing to publish numbers we cannot trace to a source we checked ourselves. A wrong number with an authoritative-looking table around it is worse than no number at all, because it looks trustworthy right up until a system runs out of power in January.

The fix is not a better table. It is a five-minute calculation run against your own address, your own roof or mount angle, and your own array orientation — which is also more accurate than any table could be, because it is specific to your actual installation rather than an average across an entire metro area.

Step-by-step: getting your own figures from PVWatts

PVWatts (pvwatts.nrel.gov), NREL’s free PV production calculator, is the standard public tool for this. The general procedure below reflects its well-documented public interface; we could not reach the live tool while writing this page to confirm the current screen layout, so treat field names as close guides rather than an exact walkthrough, and use the closest equivalent if something has moved.

1. Enter your address or coordinates. The tool pulls historical solar resource data for the nearest available grid point.
2. Choose your array type. “Fixed (open rack)” or “Fixed (roof mount)” for a typical DIY system; only choose a tracking option if you are actually building a tracker.
3. Set tilt to your actual mounting angle, not a rule-of-thumb default. A ground-mount rack, a shed roof, and a van roof rarely share the same angle, and tilt materially changes the seasonal curve, especially the worst-month figure.
4. Set azimuth to your array’s actual compass-facing direction, in degrees, where 180° is true south in the northern hemisphere. An array 30° off true south produces measurably different monthly output than one facing due south.
5. Enter your DC system size in kW, and leave the default system-losses value in place unless you have a specific reason to change it — see our PV system losses page for what that default represents.
6. Run the calculation and open the monthly output table. PVWatts reports both an AC energy column and a solar radiation column, given in kWh/m²/day for each month — that solar radiation figure is your peak sun hours for that month, directly, by the definition above.
7. Identify your worst month from that column, not the annual average at the bottom of the table. That worst-month figure is the number to carry into an off-grid sizing calculation.
8. Re-run with a different tilt if you have flexibility in mounting angle. A steeper winter-favoring tilt often raises the worst-month figure measurably, even if it lowers summer output slightly — a trade worth taking for an off-grid system, since the worst month is the constraint.

What this number does not tell you

Peak sun hours describes the resource arriving at the array’s plane — it says nothing about losses inside the system. Combine it with the derate factor from PV system losses to get expected DC array output, and account for inverter or charge controller efficiency separately to get usable energy at the battery or load. Peak sun hours is one input to a sizing calculation, not the calculation itself.