Why orientation choice matters more on a tight roof
The math behind portrait versus landscape
Every roof plane is a rectangle, and every panel is a smaller rectangle you are tiling across it. This calculator counts how many whole panels fit along each edge with a 1-inch gap between them, then multiplies columns by rows. Portrait orientation stands the panel's long side up the slope, fitting more columns across a narrow roof but fewer rows up a short one, and landscape flips that trade-off. Packing panels edge to edge also raises the shading risk, since a panel in one row can shade the row behind it when the sun sits low. The shading loss estimator works out how much output that inter-row shading costs.
Why no simple ratio picks the winner
Because only whole panels count, the winner is set by how much leftover strip each orientation wastes along the two edges, not by the roof's overall shape. A common rule says portrait wins when the slope distance is more than 1.5 times the eave length, and landscape wins when the eave runs more than twice the slope distance, but the numbers do not honor it. With 44.6 by 67.8 inch panels, a plane 20 feet up the slope and 12 feet along the eave passes the portrait test yet fits 10 panels in landscape against 9 in portrait. A plane 12 feet up the slope and 27 feet along the eave passes the landscape test yet fits 14 in portrait against 12 in landscape. This tool assumes a simple rectangle, so subtract setbacks first with the usable roof area calculator, then compare both orientations on the real remaining dimensions.
Orientation changes output only when shade is involved
Under even sunlight, a panel at the same tilt and direction produces the same energy in portrait or landscape, so panel count is the number that moves your system size. Differences show up when shade covers part of the array. Bypass diodes split each module into blocks of cells, and which blocks a shadow crosses depends on the orientation. Half-cut-cell modules tend to lose less to row-to-row shade in portrait, which is why that orientation is preferred for them. Shade from a chimney or tree can favor the other orientation, so look at where your shadows come from before assuming either one is better. NLR's PV shading documentation explains how shade patterns interact with module layout.
Why mixing orientations is a shading question
On a single roof plane, portrait and landscape panels see the same sunlight, so a mixed layout works electrically. The risk is shade. NLR's test of a partially shaded grid-tied string found that in one configuration a shadow cut power by more than 30 times its own size, and a shadow crossing a portrait panel and a landscape panel in the same string can cost more than the same shadow on matching panels. Keep each orientation on its own string or MPPT input where you can, or use module-level electronics. Panels on roof planes with different tilt or direction always need separate strings, a case the multi-roof plane array splitter helps plan.
Panel model choice changes the math too
Panel sizes vary more than most people expect. A 60-cell panel is about 39 by 66 inches, roughly 18 square feet, while 72-cell, 144-half-cell, and many current 400-watt-class panels run around 21 square feet, and the 44.6 by 67.8 inch default in this tool matches the larger style. On a tight roof plane, a smaller panel can fit more units but fewer total watts, a trade-off the panel wattage vs count tradeoff tool weighs. The roof square count to panel count calculator approaches the same fitting problem from actual roofing measurements instead.