Irradiation Published in MJ/m²/day, Design Tools Wanting Peak Sun Hours
Solar resource data has an awkward split personality. National meteorological services, agronomy datasets and older climate atlases publish global horizontal irradiance as megajoules per square metre per day, because that is the SI form a pyranometer integration naturally produces. Every photovoltaic sizing tool, module datasheet and rule-of-thumb array calculation, on the other hand, works in kilowatt-hours per square metre per day — the number the industry calls peak sun hours. One divide sits between the dataset you can download and the figure your design actually uses.
What the Converted Figure Actually Describes
One Peak Sun Hour Is a Kilowatt-hour per Square Metre
Irradiance Is Power, Irradiation Is Energy
Horizontal Is Not the Plane of Your Array
Turning a Meteorological Table Into an Array Yield Estimate
Resource data usually arrives as twelve monthly averages. The route from that table to a daily generation figure runs through four steps, and only the first one is a unit change.
Divide the published irradiation by 3.6
Type the month's MJ/m²/day value on the left and read kWh/m²/day on the right. The per-square-metre-per-day basis travels through untouched — the tool converts the energy, and keeping the area and the period in your own column heading avoids the classic mix-up with a monthly total.
Adjust from horizontal to your tilt and azimuth
The converted figure is still the horizontal resource. Apply the transposition factor from a design tool or a published tilt table for the latitude before it becomes plane-of-array irradiation, since this correction is usually larger than any rounding in the conversion itself.
Multiply by array size and performance ratio
Daily output in kilowatt-hours is array rating in kWp multiplied by peak sun hours multiplied by the performance ratio. At 5.5 peak sun hours a 6.6 kWp roof with a ratio of 0.80 produces about 29 kWh on an average day of that month.
Swap the fields to check against the original dataset
When a report quotes kWh/m²/day and the source archive is in megajoules, reversing the direction puts the design figure back into the units of the raw record so the two can be matched without re-downloading anything.
City Irradiation in Both Units and What a 1 kWp Array Returns
Annual average global horizontal irradiance for eight locations, converted to peak sun hours and then taken through to the daily output of a 1 kWp array at a performance ratio of 0.80.
| Location | GHI (MJ/m²/day) | GHI (kWh/m²/day) | Peak sun hours | 1 kWp at PR 0.80 (kWh/day) |
|---|---|---|---|---|
| London, United Kingdom | 10.1 | 2.81 | 2.81 | 2.24 |
| Berlin, Germany | 10.8 | 3.00 | 3.00 | 2.40 |
| Tokyo, Japan | 13.0 | 3.61 | 3.61 | 2.89 |
| Singapore | 15.5 | 4.31 | 4.31 | 3.44 |
| Madrid, Spain | 17.3 | 4.81 | 4.81 | 3.84 |
| Cairo, Egypt | 20.2 | 5.61 | 5.61 | 4.49 |
| Phoenix, United States | 20.9 | 5.81 | 5.81 | 4.64 |
| Antofagasta, Chile | 25.2 | 7.00 | 7.00 | 5.60 |
Two things stand out. The peak-sun-hour column is a copy of the kilowatt-hour column, because the two are the same quantity under different names, and Singapore sits below Madrid despite being close to the equator — cloud and humidity beat latitude comfortably. Multiply the last column by 365 and London's 2.24 kWh/day becomes roughly 820 kWh a year per kilowatt-peak on the horizontal, which a sensible tilt then lifts towards the 900–1,000 a real installation reports.
Working Through a Resource Table
A Whole Year of Monthly GHI Rows Without Retyping
The result tracks each keystroke, so twelve monthly values can be entered one after another and read off in kilowatt-hours without clearing or re-selecting anything between rows.
Peak Sun Hours Copied Into the Yield Spreadsheet
Copying gives the bare figure with no unit attached, which is what the kWp × PSH × PR formula in the next column expects to multiply.
Go Back to MJ/m² for a Meteorological Cross-Check
Reversing the direction restates a design figure in the units the weather archive holds, which is the only way to confirm that a modelled dataset and a ground station agree.
Older Langley Data Reaches the Calorie Entry
Historic radiation records are often in langleys, one calorie per square centimetre. Searching the dropdown for the calorie lets those archives be brought onto the same scale as a modern megajoule series.
Irradiation Questions From the Array Layout
My dataset says 20 MJ/m²/day — how many peak sun hours is that?
5.56. The 3.6 comes from the definition of the watt-hour: a watt is a joule per second and an hour is 3,600 seconds, so a kilowatt-hour is 3.6 million joules, which leaves one megajoule worth 1/3.6 of a kilowatt-hour. Because the reference irradiance in the peak-sun-hour idea is exactly 1,000 W/m², the kilowatt-hours per square metre and the hour count come out identical, and no second step is needed. Most inhabited sites land between 3.5 and 6.5 peak sun hours as an annual average, which is 12.6 to 23.4 MJ/m²/day.
What does a peak sun hour physically represent?
It is a bookkeeping device, not a description of the sky. Real irradiance climbs from zero at sunrise to a midday maximum and falls back, and the area under that curve is the day's irradiation. A peak sun hour flattens the curve into a rectangle of the same area at 1,000 W/m². A site with five peak sun hours may have had fourteen hours of daylight, none of them at exactly 1,000 W/m², and a bright cool day with cloud edge enhancement can briefly exceed that reference.
How do I get from an irradiation figure to daily kilowatt-hours for a given array?
Array rating in kWp multiplied by peak sun hours multiplied by the performance ratio. A 6.6 kWp roof at 5.5 peak sun hours with a ratio of 0.80 gives 6.6 × 5.5 × 0.80 = 29.0 kWh for an average day of that month. The rating already assumes 1,000 W/m², which is what makes the multiplication legitimate: peak sun hours are counted against the same reference the module was measured at. The performance ratio then absorbs temperature, soiling, wiring, inverter conversion and downtime.
Why is my site's December figure a fraction of its June one?
Three effects compound. Day length shortens, the sun sits lower so each square metre of ground intercepts a thinner slice of the beam, and the light travels through more atmosphere at that low angle. At London's latitude the swing runs from roughly 5.2 kWh/m²/day in June to about 0.6 in December — a factor near nine. Madrid moves from around 7.3 to 2.1, a factor of about 3.5, and a tropical site barely moves at all. This is why sizing anything off an annual average misleads badly for winter-critical loads.
Can tilted-plane irradiation be derived from a horizontal figure?
Yes, but not with a single multiplier. Transposition models split global horizontal irradiance into its direct and diffuse components, geometrically project the direct beam onto the tilted plane, apply a sky model to the diffuse part and add ground-reflected light using an albedo assumption. That is why serious tools want the diffuse fraction as well as the total. A published tilt table for the latitude is adequate for a first estimate — roughly a 10–25 % annual gain at mid-latitudes for an equator-facing array near latitude tilt — but the monthly split changes shape, with steeper tilts trading summer output for winter.
No comments yet. Be the first to comment!