Shading, and why it costs more than it looks

Of the three terms in an MCS solar estimate, two are facts about the roof and one is a judgement. Capacity is arithmetic. The yield figure comes from a table. The shade factor requires somebody to go and look, and it is the term most likely to be missing.

Counting segments

MGD 005 describes the method. Stand as near as possible to the centre and base of where the array will go, which usually means leaning out of an upstairs window, with no need to climb on the roof. Plot everything that blocks the horizon onto a sunpath chart, which is a plot of where the sun travels across the sky over a year, marked out with the hours of the day and the months of the year.

The chart is divided into 84 segments. Each is worth 0.01. Count the segments that an obstruction touches, and the total is the loss.

An estimated loss of 11 per cent gives a shade factor of 1 − 0.11 = 0.89, and the annual estimate is multiplied by it.

MGD 005 is candid about its accuracy: it expects to land within about ten per cent of real annual output on most systems. It is a screening method, not a simulation, and it is designed so that two surveyors assessing the same roof arrive at similar answers instead of at whatever they felt.

Near shade and far shade

The procedure splits at ten metres, and the split matters.

Anything further away than ten metres, a hill, a line of trees across the road, a neighbouring terrace, shades the whole array at once. It blocks a portion of the sky, the segment count measures that portion, and the loss is roughly proportional, so the method handles far shade well.

Anything within ten metres, a chimney, an aerial, a dormer, a satellite dish, is different. It casts a hard shadow that moves across the array during the day, covering different modules at different times. MGD 005 requires each near object to be assessed from the part of the array it affects, not from the middle.

The guidance is blunt about it: near shading has a considerable effect on system performance and should be avoided, and solar PV systems should not be sold where the impact of shade could be severe. Where a proposed location is shaded by numerous objects, the customer should be told that the location may simply not be appropriate.

Why a small shadow costs a lot

The segment count measures how much light is blocked. What is lost can be considerably more, and the reason is electrical, not optical.

Modules in a string are wired in series, so the same current flows through all of them. A shaded module produces less current, and because it is in series it limits the current for the whole string. Bypass diodes inside each module reduce the damage by allowing current to route around a shaded section, but a module typically has only three of them, so the granularity is coarse.

The practical effect is that a chimney shadow falling across the bottom row of a string can cost far more than the bottom row's share of the array. A shadow covering five per cent of the surface can cost fifteen or twenty per cent of the output while it lies there.

This is why the design response to shading is usually about wiring and not about the panels: keeping the shaded modules on their own string so they cannot drag the unshaded ones down, or fitting optimisers or microinverters so that each module works independently.

Optimisers and microinverters

Module-level power electronics genuinely help a shaded array, and MGD 005 gives them no credit at all.

That is not an oversight. The guidance says the benefit varies too much between projects to tabulate, and it is right: what an optimiser recovers depends on the pattern of the shadow, the string layout, and the time of year, in a way no single factor could capture.

So the shade factor is calculated as if they were not there. Where proprietary software has been used to model the array both with and without the obstructions, the ratio between the two may be used instead, and MIS 3002 allows that provided the alternative method can be shown to be at least as good as MGD 005.

The consequence for a buyer is that a quotation may show an unchanged shade factor while also quoting optimisers. That is correct under the standard, though it does mean the estimate is conservative for that system.

Shade moves

Two things change through the year, and both are easy to miss on a single visit.

The sun's height changes. A low December sun casts shadows several times longer than a June one, so an obstruction that clears the roof all summer may cover half the array at midday in winter. The sunpath chart handles this, which is why the assessment uses one and not a photograph.

Trees grow, and deciduous trees have leaves for part of the year and not the rest. An assessment made in February on a bare tree will not describe July. A neighbour's leylandii, unchecked, can turn an unshaded roof into a shaded one inside a decade.

There is no right to light for a solar panel in Britain, so a neighbour is generally free to grow a hedge or build an extension that shades your array. High hedge legislation exists for evergreen hedges over two metres, but it addresses light to a dwelling, not to a solar installation.

Reading it on a quotation

A shade factor of 1.00 is a claim that nothing blocks the horizon at all. On an open roof with a clear southern aspect that may be exactly right, and MIS 3002 explicitly permits the assessment to be skipped in that case.

On a terraced street, in a wooded area, or on a roof with a chimney on it, a shade factor of 1.00 means nobody looked. It is worth asking whether an assessment was done and what it found, because an unexamined shade factor is the easiest way for an estimate to be ten or fifteen per cent optimistic while still appearing to follow the method.