No south face? Almost never a reason to give up

How much is really lost

In our area, an east-west pitched setup yields around 82 to 85 percent of what the same system would deliver facing south. That is the factor in our calculator and studies, taken from PVGIS for this latitude. Losing 15 to 18 percent is not pleasant, but it is far from turning the installation into a bad deal.

What east-west does better than south

South concentrates production into one huge midday peak. East-west spreads it: the east face works hard in the morning and the west stretches into the evening. For a home with early breakfasts and late dinners, that flatter curve matches real consumption better, and every kilowatt hour used directly is worth more than one exported as surplus.

Put plainly: east-west produces less in total but can be used better. The final sum depends on your profile, which is why it is done with your consumption on the table, not with a rule of thumb.

Designs that rescue awkward roofs

When the roof does not make it easy, these are the design tools:

• Splitting across two faces with the right inverter: modern equipment handles two orientations comfortably, each with its own tracking.
• Optimisers or microinverters where partial shading exists: each panel works on its own and a chimney does not drag down the whole row.
• Tilted mounting on flat roofs: on terraces and industrial decks, orientation is chosen with the structure rather than imposed by the building.
• Dropping the bad face: sometimes the best decision is to skip part of the roof and size the other part properly.

When we say no

There are cases where the honest answer is to stop: heavy shading most of the day, very small roofs already crowded with skylights and chimneys, or steep pure-north faces. They are few, but they exist, and we would rather say so in the study than install something that will not deliver. If your roof is one of the awkward ones, the serious answer is not an opinion over the phone: it is measuring it.

When there is only one slope: the roof that faces only east or only west

The textbook east-west case is a house with two pitches and half the panels on each side. But the towns of this comarca are full of another case: the terraced house between party walls with a single-pitch roof, and that pitch faces only east or only west. There is no splitting anything here: all the production lands in half the day.

That changes the annual total less than people fear, and it changes the shape of the day a great deal. A west-only roof loads production into the afternoon: it suits whoever gets home mid afternoon, runs the washing machine on arrival and leans on the air conditioning through long summer evenings. An east-only roof is the opposite: strong mornings, ideal for anyone working from home or living a morning life. Honest design starts by asking how the house lives, not with the panel catalogue.

The practical consequence is that on these roofs managing consumption matters more than on a south-facing one: scheduling appliances into the production hours, and weighing up a battery by a different standard, because here it does not just shift surplus, it covers the half of the day when the roof is idle. Sometimes it adds up, sometimes it does not: that depends on real consumption, and it gets calculated, not assumed.

The mistake of mixing orientations on the same string

It is the most repeated technical fault on awkward roofs, and it is invisible to the naked eye: east-facing and west-facing panels wired into the same series string. In series, the current of every panel is capped by whichever one is producing least at that moment, so in the morning the west panels drag down the east ones, and in the afternoon the other way round. The result is a system that never yields what its parts add up to.

The right fix depends on the geometry. With two clean orientations it is enough to give each its own inverter input, each string with its own power tracker. When the roof is broken into many small planes, with chimneys and crossing shadows, optimisers or microinverters come into play, making each panel independent. But they are not a compulsory toll: they add cost and electronics on the roof, which means more potential points of failure. They go in when the geometry justifies it, not as a default.

On existing installations this mistake gives itself away in the graph: odd humps mid morning, production below the local range with no shade to explain it. It is one of the first things to check when reviewing someone else's work, because rewiring the strings costs little compared with what it recovers.

Frequently asked questions