Roof orientation and tilt: what actually moves C&I yield

Every roof-by-roof walk-down eventually lands on the same question: does this building face the right way, and at what angle do we rack it. For a single site that's a quick call. Across a forty-site portfolio, orientation and tilt decisions are what separate a shortlist that holds up through PPA bidding from one that gets re-cut three times.

Azimuth: why south-facing isn't the only answer anymore

Azimuth is the compass direction a panel faces, measured in degrees from true north. A roof plane running due south gives the highest annual yield per kWp in the northern hemisphere, and for a long time that was the whole conversation. On flat C&I roofs it rarely is, because the roof itself doesn't dictate azimuth. Racking does.

Most flat industrial roofs use ballasted racking, and the installer can point that racking almost anywhere the structure and setback rules allow. The azimuth that ends up mattering is the one the racking plan sets, driven by array layout, racking type and whatever rooftop equipment is already up there. A roof with a slight southwest skew and no parapet shading will often beat a textbook south-facing roof with three rooftop units in the way.

Where azimuth still matters structurally is pitched roofs on distribution centers, manufacturing buildings with sawtooth monitors, and anything with fixed-tilt framing already built in. There, the roof plane sets the azimuth and you're working with what you've got.

Flat roof tilt angle: the self-shading trade-off

On flat roofs, tilt is a racking choice, and it's the one with the most interacting variables. A steeper tilt (15-20 degrees) captures more sun per panel but needs wider row spacing to avoid self-shading, which means fewer rows fit and usable area drops. A low tilt (5-10 degrees) sacrifices some per-panel output but lets you pack rows tighter, often netting more installable kWp on a roof with plenty of open area and no penetrations to dodge.

This is the ground coverage ratio trade-off every EPC runs through at detailed design, and it's exactly why a square-metreage number from a desk screen is close to useless on its own. Two roofs with identical usable area can land ten or fifteen percent apart on installable kWp depending on how much of that area gets eaten by shadow setback at the tilt the racking plan uses.

East-west arrays: more kWp, flatter production curve

East-west racking has become the default for a lot of flat-roof C&I work. It's a layout choice, distinct from the azimuth question above, and it changes how rows sit on the roof rather than which way the roof faces. Instead of south-facing rows spaced apart to avoid shading each other, east-west arrays butt rows together in a sawtooth, with half the panels facing east and half west at a low tilt, usually 5-10 degrees.

The per-panel yield on an east-west array is lower than a south-facing array at optimal tilt. But because rows can sit almost touching, GCR goes up substantially, and a given roof fits more panels. On roofs with generous open area the installable kWp often comes out higher east-west than south-facing, even with the yield penalty per panel. The production curve also flattens across the day instead of peaking at solar noon, which can matter for sites with a daytime load shape or an offtake agreement that values spread generation.

This kind of call is hard to make from a satellite image and a protractor, and driving to forty roofs to make it by eye doesn't scale either. A portfolio rating needs to score obstruction, usable area and orientation together, site by site, and hand back a kWp figure per roof rather than a raw area total. Rooftop Solar Suitability runs that pass across a whole portfolio from one VHR capture, ranking roofs by the install-ready design they'd actually support.

If your shortlist is still a spreadsheet of rooftop square metres, it's worth running the portfolio through that kind of pass before the next bid cycle.