Calculate panel count for your roof
Solves the layout as a discrete packing problem — portrait, landscape and mixed bands — not as an area quotient. The quotient systematically overestimates because modules cannot be divided; in the reference case by 11 %. For flat roofs the calculator couples racking tilt, row spacing and module count: the shallower tilt fits more often and clearly beats the yield-optimal angle.
The formulas behind the calculator
Every number above can be recomputed: the full calculation path, all assumptions and the data source with retrieval date — plus cross-validation against independent references. Disclosed, not claimed.
An estimate based on the stated assumptions. The final design must be checked by a qualified professional against the rules that apply where you are.
| Step | Formula | Value | Provenance |
|---|---|---|---|
| Usable width | W_roof - 2 * margin_side | 6 m | exact |
| Usable height | H_roof - margin_ridge - margin_eaves | 3.7 m | exact |
| Best uniform layout | max(cols_p * rows_p ; cols_l * rows_l) | 10 | exact |
| Module count | max(uniform ; best band split) | 10 | exact |
| System size | n * P_module / 1000 | 4.4 kWp | exact |
- Formula
Pitched: columns = ⌊W_usable / (w + gap)⌋ · rows = ⌊H_usable / (h + row gap)⌋ · Flat: rows = ⌊(H_usable − L·cos β) / pitch⌋ + 1 with pitch from calculator 26- Valid for
- Rectangular, obstruction-free area with uniform setbacks. Flat-roof mode: parallel rows of equal height on level ground, racking tilt 5 to 45°, yield curve from the hourly TMY simulation of the site.
- Not covered
- Roof windows, dormers, chimneys, non-rectangular areas and self-shading from the row in front — that is the row spacing calculator. In flat-roof mode additionally: soiling, snow, east-west racking and the slightly better cooling of steeper rows.
Frequently asked questions
How many solar panels fit on my roof?
The calculator treats the layout as a discrete packing problem: columns = ⌊W_usable / (w + gap)⌋ and rows = ⌊H_usable / (h + row gap)⌋, evaluated in portrait, landscape and mixed horizontal bands optimised by dynamic programming. In the reference case — a usable area of 6.00 × 3.70 m and a 1.762 × 1.134 m module — portrait fits 10 modules while landscape fits only 9.
Why can't I just divide the roof area by the module area?
Because the area quotient systematically overestimates: modules are rectangular, indivisible objects, and the remainder at every edge is lost. In the reference case the quotient suggests 11.1 modules where only 10 actually fit — 10 % too high, and the absolute error grows with larger roofs. The calculator reports the quotient figure alongside as a comparison.
Does the calculator handle roof windows, dormers and chimneys?
No — it assumes a rectangular, obstruction-free area with uniform setbacks and warns when no obstructions have been entered. The setbacks depend on building and fire codes, which is why they are inputs. Self-shading from the row in front is also out of scope — that is what the row spacing calculator computes.
What does the mixed portrait/landscape layout achieve?
After the full rows of one orientation, a leftover strip often remains that still fits the other orientation — exactly this strip is what the band decomposition searches (dynamic programming over the usable height in 1 cm steps). If the mix beats the best uniform grid, the calculator says so explicitly. In the reference case (6.00 × 3.70 m usable) portrait wins with 10 modules; on other dimensions this flips quickly.
Which setbacks do I have to keep — and why are they only an assumption here?
Building and fire codes govern this, not physics: it differs by jurisdiction, building class and boundary distance. The calculator prefills 0.3 m per edge and marks it as an assumption with a range — the binding number comes from your building code or installer. That is why the warning stays until you enter your own values.
Which module size does the calculator assume?
The prefill is the 2026 market standard: 1.134 × 1.762 m at 440 Wp (108 half-cells). Deviating formats persist — glass-glass, full-black special sizes, 132-half-cell modules over 1.9 m long. One centimetre of module width can cost a whole column: at 6 m usable width, five 1.134 m modules fit, but only four at 1.21 m. So enter the size from your module datasheet.
How many kW fit on 650, 1,300 or 1,600 sq ft of roof?
The calculator answers that for the actual shape, not for the area — and that is where the difference lies. With a standard module (1.134 × 1.762 m / 44.6 × 69.4 in, 440 W) and 0.3 m setbacks all round: 60 m² (646 sq ft) as 12 × 5 m gives 21 modules and 9.24 kW, 120 m² (1,292 sq ft) as 12 × 10 m gives 48 modules and 21.12 kW, 150 m² (1,615 sq ft) as 15 × 10 m gives 64 modules and 28.16 kW. The area quotient promises 25.1, 53.6 and 67.7 modules instead — 20 %, 12 % and 6 % too high. The error is largest on small and awkwardly shaped roofs, where the unusable edge strip weighs relatively more.
Is the rule of thumb of 15 sq ft per panel still right?
No — it describes a panel generation that is no longer sold. A current 440 W module measures 1.134 × 1.762 m, which is 1.998 m² or 21.5 sq ft, not 15. Using 15 sq ft overestimates the module count by more than 40 % before any packing losses are counted. The second common rule, roughly 1 kW per 70 sq ft of total roof, comes closer to the calculated result: the examples above work out at 70 sq ft per kW on a small roof, 61 on a medium one and 57 on a large one. Rules of thumb cannot capture that spread, because it comes from the shape and not from the area.
How many panels fit on a flat roof — and at what racking tilt?
On a flat roof the tilt is not a free choice: steeper means more yield per module, but a longer shadow and therefore fewer rows. Example, 20 × 15 m in Berlin, shade-free from 9 am to 3 pm on 21 December: at 5° tilt 70 modules (30.8 kW) fit, at 40° — the best angle per module — only 32. The shallow layout delivers about 29,900 kWh a year, the steep one 15,800 kWh: 90 % more, even though each individual module is worse placed. The calculator therefore varies tilt and row spacing together and shows both curves. Soiling and snow are not part of the calculation — very shallow racking performs worse in practice than it does here.
Why does the flat-roof mode ask for a design time?
Because that question drives the layout more than anything else. On the same 20 × 15 m roof in Berlin, 96 modules fit if the array only has to be shade-free at noon on 21 December — and only 70 if it has to be shade-free from 9 am to 3 pm. That is a 37 % difference from a single design decision. Both answers are correct; the only wrong move is to answer the question silently. The row spacing itself comes from the same formula as the row spacing calculator, so both pages give the same number.
How many panels fit on a garage or carport roof?
For a 6 × 3 m garage roof with 0.3 m setbacks the calculator finds 6 modules and 2.64 kW, filling 93 % of the usable area. The area quotient promises 6.5 modules — here it is almost right for once, because the dimensions happen to suit the module grid. That is exactly the point: whether the quotient is 8 % or 26 % off depends on the shape and cannot be estimated, only computed.