How the output of a solar PV system is estimated

An MCS solar quotation has to say how much electricity the system is expected to generate in a year, and it has to arrive at that figure by a method set out in MIS 3002. The method is short enough to fit on a postcard:

Annual output (kWh) = kWp × K_k × SF

Three numbers. The size of the array, a figure from a national table, and a shading allowance. That is the entire calculation. Because it is fixed, two installers quoting on the same roof have to produce the same answer, so a householder can compare quotations on price and not on optimism.

kWp: the size of the array

Peak kilowatts is the sum of the nameplate ratings of the modules, and nothing else. Twelve panels rated 440 W each make 5.28 kWp.

That rating is measured under standard test conditions: 1,000 watts of light per square metre, a cell temperature of 25 °C, and a defined spectrum. Those conditions are a laboratory, not a British roof. On a bright June afternoon a panel is well above 25 °C and losing a few per cent to heat; on a January morning it may be cooler than the test but seeing a fraction of the light.

kWp is therefore not a prediction of anything. It is a label that lets panels be compared with each other, and it is the first term in the sum because everything else scales with it.

K_k: what a kilowatt yields where you are

The second term does all the work. K_k is the number of kilowatt-hours a single kWp will produce in a year, at your location, at your roof's pitch and orientation, after allowing for the losses in a real system.

MCS publishes it as a set of tables, one for each of twenty-five postcode zones, giving a value for every whole degree of pitch from flat to vertical and every five degrees of orientation. The data comes from the European Commission's Joint Research Centre, from the Climate-SAF-PVGIS dataset, and MCS multiplies it by 0.8 to account for system losses: inverter conversion, cable resistance, dirt on the glass, mismatch between modules, and the fact that panels run hot.

The zones are the same ones SAP uses, each represented by a city whose weather record stands in for the region: zone 1 is London, 5E is Bristol, 11 is Sheffield, 17 is Inverness, 21 is Belfast. Your postcode area determines your zone, with some areas split by district, so BD is Sheffield's zone but BD23 and BD24 around Skipton are Middlesbrough's.

A few reference points, all for a due south roof at 35 degrees:

The spread from the south coast to Shetland is about a third. That is a real difference, but a well placed roof in Glasgow still does around 85 per cent of what the same roof does in London, and the gap between a good roof and a shaded one in the same street is often larger than the gap between two ends of the country.

Orientation and pitch

Orientation is measured as an angle away from due south, in either direction. Due south is 0. South-east and south-west are both 45. East and west are both 90. Due north is 180. The standard says to round it to the nearest five degrees.

Pitch is degrees from horizontal: a flat roof is 0, a typical British pitched roof is 35 to 45, a wall is 90, and the figure is rounded to the nearest degree.

The penalty for facing the wrong way is gentler than intuition suggests. Taking a London roof at 35 degrees as the baseline:

An east-west roof, with panels on both slopes, is a perfectly reasonable proposition. It generates about four fifths of what a south roof would per kWp, but it spreads generation across the morning and evening instead of concentrating it at noon, which often means more of it is used in the house and less exported. On a house with an export tariff worth a fraction of the import price, that shape can be worth more than the extra kilowatt-hours a south roof would have made.

The optimum pitch in Britain is around 35 to 40 degrees, but the curve near the top is remarkably flat. In London, 30 degrees gives 99 per cent of the best and 20 degrees still gives 96, so it is rarely worth arguing about. An array laid flat on a flat roof is a different matter: it loses about 16 per cent, which is why flat roof systems are usually tilted on frames, though that brings its own constraints on row spacing and ballast weight.

SF: the shade factor

The third term is a multiplier between 0 and 1 representing what shading takes away. A clear horizon is 1.00. An estimated 11 per cent loss gives a shade factor of 0.89.

Where there is genuinely nothing blocking the horizon, MIS 3002 allows the assessment to be skipped and 1.00 used. Where there is anything, the standard points at MGD 005, which sets out how to assess it: stand at the middle of where the array will go, plot what blocks the horizon onto a sunpath chart, and count the segments covered. The chart has 84 segments and each is worth one per cent.

Shading is the term most likely to be wrong on a quotation, because it is the only one that requires somebody to go and look. It is covered in more detail in the article on shading.

What the number means

The estimate is an AC figure, measured after the inverter. It is what a generation meter would count, not what the panels produce on their DC side.

It is a long-run average. A bright year will beat it and a dull one will fall short, by perhaps ten per cent either way. Judging an installation against its estimate after one summer is not a fair test.

It says nothing at all about money. It is a quantity of electricity. What that electricity is worth depends entirely on how much of it is used in the house and how much exported, which is a separate calculation with its own MCS method.

And it carries no allowance for the particular equipment. The 0.8 factor is a single national number covering inverter efficiency, cable losses, soiling and heat, applied identically to every system in the country. A quotation claiming a higher figure because of better panels or a better inverter has departed from the method, and its figure cannot be compared with anyone else's.

Checking a quotation

The five inputs are all things you can establish yourself: capacity, postcode, pitch, orientation, shade factor. Put them through the same table and you should get the same answer, within rounding.

If a quoted figure is materially higher than the method gives, there are only a few possibilities. The shade factor may have been assumed as 1.00 without anyone looking. The pitch or orientation may have been recorded optimistically. Or the figure may not have come from the method at all, in which case it is worth asking what it did come from.