Most of the world’s solar panels are facing the wrong direct

[cubes]

http://qz.com/148078/most-of-the-worlds ... direction/

You’d think it would be easy: the sun is “up,” and, like leaves and basking reptiles, solar panels should face in that general direction. But most installers of solar panels, especially the ones for homes, follow conventional wisdom handed down from architects, which holds that in the northern hemisphere, windows and solar panels should face south.
+

This makes intuitive sense since it would seem to maximize the amount of sunlight a panel will get as the sun tracks from one horizon to the other. But it isn’t true, at least according to a single study of homes in Austin, Texas. The Pecan Street Research Institute found that homeowners who aimed their panels toward the west, instead of the south, generated 2% more electricity over the course of a day.

[adam2]

PV modules should face more or less south, but the exact angle for best results depends on local conditions.

As the sun rises from the east and sets towards the west, then a bit west of south will give good results if the weather in the locality tends to be clearer in the afternoon.
Likewise in a locality with clear morning weather but regular afternoon cloud, then a bit east of south may be better.

Also some people confuse energy produced with achieving the best financial return. In many parts of the USA, electricity is sold at a higher price in the afternoons in summer, due to the widespread use of airconditioning. In such circumstances it is often worth maximising summer afternoon production to get the best return in money, even if total KWH production is slightly reduced.

For off grid applications it depends on for what purpose the electricity is to be used.
For lighting, due south and nearly vertical is often best in order to maximise winter production when lighting needs are greatest and sunlight limited.
For air conditioning, a fairly shallow angle towards the west of south will maximise production at the warmest time of day.

Finally, in the case of very large grid tied systems that cost millions, these are virtually certain to be correctly aligned for best results. There is a lot of money innvolved and those in charge have a clear interest in getting it just right.

Smaller installations are often atop existing buildings and may not be perfectly aligned, but still generaly produce acceptable results.

[biffvernon]

When you look at the graphs that show performance plotted against orientation and tilt it's evident that there's quite a bit of wiggle-room before you lose a lot of power.

There must be quite a bit of variation with season. In the summer there are many hours of the day when a south-facing panel is shaded by itself because the sun is north of east or west. A flat mounted panel may be more useful in summer.

[adam2]

Yes, a flat or nearly flat mounting can be effective, the main problem being that modules mounted flat, or nearly so, are very liable to accumalation of dirt reducing the output.

At steeper angles the self cleaning effect of the rain helps more.

At the equator, flat mounting is often best provided that affordable labour and safe access are available for cleaning.

[RenewableCandy]

Here in the Exotic East (of the UK) we often have misty mornings (Haar) on otherwise-sunny days. So I can well imagine SW would be a better bet than due S. However, on the other side the Pennines they often have rainy afternoons even after a sunny morning, so they're probably better off with theirs facing SE.

[biffvernon]

So really the answer is to have them mounted on solar-tracking mechanisms with a squeegee-walla on standby.

(Or just stick them on whatever roof you've got and accept your losses.)

[adam2]

PV output may be significantly increased by mounting the modules on a tracker, such that they follow the sun.
In practice though the tracking mechanism adds complication and expense and moving parts, thereby removing one of the merits of PV which is that it contains no moving parts.
In many cases fixed mounting at an optimum or reasonably near optimum angle is more worthwhile. Purchasing additional modules to give the required output is often more economic than a reduced number of modules on a tracker. And of course the additional modules will probably have a longer working life than the tracker.

An exception is small installations in the developing world for schools, clinics and the like. In such cases a simple manually operated tracking mount can be worthwhile, the labour cost being effectively zero for manual intervention.