What an installer checks before accepting a roof pitch is not the pitch itself. The pitch is fixed by the building, so what is actually being tested is whether a design can survive on the plane that already exists: whether the tilt the yield figures assume can be built there, whether there is room for the rows to sit without one shading the next, and whether the mounting that comes out of it is something a competent person will sign off. Where those three hold, the pitch gets accepted and the design moves forward. Where they do not, the numbers change before anything else does.
I want you standing in the driveway with me for this, because everything that matters is visible from there and almost none of it is written on the page in your hand.
The quote is open on the bakkie's bonnet and the wind keeps lifting the corner of the yield table. Look up instead. Your roof runs north at a decent pitch, concrete tile, the ridge line cutting across a hard sky. Next door the corrugated sheet roof sits almost flat, shallow enough to walk, and their panels lie down on it as if resting. Across the road there is a flat concrete roof with two rows angled up on a frame, and between those rows a gap you could fall into.
Three roofs, three different tilts, and all three of those households will be handed a yield figure that looks about the same. That is the moment I started reading quotes differently. Not the price page. The tilt line.
The same word means two different things
The yield table has a line on it: Tilt 22 degrees. Your roof is about 22 degrees. It matches, and most people stop there.
Then you read the same line on the neighbour's quote. Also 22. His roof is ten degrees at most. One of those two numbers is the roof. The other is somebody's decision, and it is doing a lot of work in the model.
Here is the distinction that reorganises the whole conversation. On a pitched roof, the panels follow the plane, so the roof pitch sets the tilt. The designer has no say. On a flat concrete roof, the roof has no meaningful pitch at all, maybe a degree or two for drainage, and the tilt is whatever the racking is set to. Same word, same position on the page, two completely different kinds of fact. One of them is a constraint. The other is a choice, and choices can be defended or not.
So the first question out of my mouth is never "what tilt did you use?" It is "is that tilt my roof, or is it your racking?" If the person across the table has to think about the answer, that tells you something too.
Flat concrete gives the most freedom and takes it back in spacing
A flat concrete roof is the only common roof type where the designer gets to pick the angle. That freedom is real, and it is paid for in row spacing.
Lean a row of panels up and its shadow gets longer. The row behind has to move back, or sit lower, or not be there at all. On a big flat roof over a double garage you can absorb that. On a small one you cannot, and the honest outcome is fewer panels than the drawing implies. A layout that shows twelve on a flat roof and does not show the gaps between the rows is not a layout yet. It is a wish.
Watch the parapet as well. On a north-facing flat roof it is the north parapet that throws shade onto the array, and in midwinter, when the sun sits at its lowest and the shadow is at its longest, that is exactly when the household needs the production most. Morning and late afternoon the shadow swings sideways instead. None of this is exotic. It is just invisible from the street.
What a steep roof does to the maths
A steep roof does two things well. It sheds rain, which keeps dust moving in the summer-rainfall parts of the country, and it lifts the array clear of the street, the trees and the worst of the ground-level shading.
What it does badly is force the tilt. If the site sits at roughly 26 degrees south and the roof runs at 40, the array is leaning towards the winter sun whether anyone wanted that or not. That is not automatically a mistake. Winter is when a geyser works hardest, when the days are short, and when a household's consumption sits closest to what the roof can make. A winter-weighted tilt can be entirely sensible.
If you are buying annual kilowatt-hours, though, a steep plane costs you December. The question is which one you are buying, and the roof has already answered it. I have watched people argue with a 38-degree roof for a week and then accept a number that quietly assumed 25.
Two planes under one roof
Plenty of South African houses have a north face and a west face under the same roof, or a pitched section next to a flat one. One tilt line cannot describe that.
If the quote carries a single tilt, a single azimuth and a single yield figure, ask which part of the roof that number belongs to. Then ask whether the two planes are on the same string, because a steep north roof and a shallow west roof sharing one string pull each other down towards somewhere neither of them wanted to be. That is an electrical consequence of a structural fact, and it is exactly the kind of thing a reviewer should be catching, not the installer explaining afterwards.
Shallow sheet, steep tile, and what the surface decides
A tile roof's pitch was decided long before anyone thought about solar, and it is not going to change for you. The plane is also broken by hips, valleys and a chimney flue, so the usable rectangle is smaller than the outline suggests. When the drawing fits panels edge to edge across the whole north face, that is worth questioning.
A sheeted roof usually runs shallow. Five to fifteen degrees is common. The array ends up close to the surface, which means it stays dirty, and in a Highveld winter with no rain for weeks a shallow array drifts away from a model that assumed ordinary soiling losses. If the sheeting turns out to be one of the older materials, the conversation stops being about tilt altogether, because that roof is not something anyone can work on without the right specialist, and the proposal has to be rebuilt around that fact.
The questions I ask, in the order I ask them
- Is the tilt in this table my roof or your racking?
- How many rows, what is the gap between them, and what does that gap look like in June?
- Which planes are you using, and are they on one string or separate ones?
- What soiling loss did you assume, and does that match how often I will realistically clean this roof?
- Does the yield figure change if the array sits on my actual plane rather than a clean north-facing one at an ideal angle?
- What has been signed off, by whom, and does that sign-off cover the mounting as well as the wiring?
- If the design changes after the SSEG application goes to the municipality, who corrects the paperwork?
- What changes on this quote if any of the answers above turn out to be wrong?
On the yield estimate specifically: a design built on your real roof plane and a design that quietly assumes an ideal one can both produce plausible-looking numbers. What usually gives it away is the specific yield, the annual kilowatt-hours divided by the installed kilowatts. From what I have seen, a clean, north-facing, well-tilted Highveld array tends to land in the region of 1,600 to 1,800 of those units per kilowatt. A shallow, slightly east-facing plane that comes back with the same figure has something in the model that the roof cannot cash. It is not proof of anything. It is a reason to ask.
The part that isn't yours to decide
The pitch, the mounting hardware and the waterproofing under it are one conversation, and it belongs to whoever signs off the design. Tilt and mounting choices feed straight into wind loading, which is one reason a professional sign-off is required before anything goes onto the roof at all. SANS 10142-1 governs the wiring, including the DC side of the array, but the sign-off that matters at this stage is about the structure and how the array is held down.
That is not my field and I do not pretend otherwise. My role, and yours, is to make the questions specific enough that the answers can actually be checked. You are not specifying the fix. You are finding out what the design assumes and whether anyone has stood on the roof and confirmed it.
One piece of paperwork worth chasing before anything is ordered: the tilt, the azimuth and the expected annual production usually travel into the SSEG application. If the design shifts after that application is submitted, the application is describing a system you will not be getting. Ask who updates it, and get a name.
Back in the driveway, the wind is still lifting that page. You are not reading the price any more. You are looking for one line, in a table nobody highlights, where the designer has written down what angle they think your roof is. The pitch is not the problem. The pitch is the constraint. The problem is a yield table that has never met it.
