Testing Irregular or Constrained Sites
Textbook rectangular sites are the easiest to plan around, but at concept stage they're often not what you're actually working with. Corner blocks, angled boundaries, narrow frontages, sites that widen or narrow partway back, easements cutting across part of the land — irregular and constrained sites come up constantly in real feasibility work.
The good news is that a layout-based approach handles this reasonably well, provided the actual site geometry is tested rather than assumed. Here's a walkthrough of how that testing process plays out.
Starting point: draw the actual boundary
Rather than approximating an irregular site as a rectangle of similar area, the first step is drawing (or entering) the site's actual boundary shape. This matters because, as covered in why site area alone doesn't predict parking yield, two sites of the same area can produce very different results once their real shape is accounted for. An angled boundary, for example, often leaves an unusable triangular strip at one end of a bay row — space that a simple area calculation would have counted as usable, but that a layout drawn against the actual boundary correctly excludes.
Place an access point, then question its location
On a constrained site, the access point often does more than let vehicles in — it can determine whether an efficient circulation layout is possible at all. A frontage-constrained corner site, for instance, might only have one or two realistic locations for an access point, and the difference between them can be significant. This is worth testing directly rather than assuming: place the access point where it seems most natural first, generate a layout, then try an alternative location and compare. Sometimes moving an access point toward a corner — rather than centring it on the frontage — opens up a circulation path that uses the site's constrained shape more efficiently. Other times a central access point genuinely is the better option. The only way to know is to test both against the real geometry.
Try more than one parking angle
Constrained sites often respond differently to angle changes than a generous rectangular site would. A narrower site might not have room for the aisle width a 90° layout needs, while a shallower angle — 60° or 45° — could fit more comfortably within the available width, even though it typically uses more linear frontage per bay. See choosing a parking bay angle for the general trade-off; on a constrained site, it's often more pronounced, so it's worth testing rather than defaulting to a standard angle.
Watch what happens at the awkward parts of the site
On an irregular site, the interesting result usually isn't in the main body of the layout — it's at the edges, corners, and narrow points where the boundary constrains things. This is where a slightly different access point location changes whether an extra bay or partial row fits, a different angle handles a narrow section better or worse, and circulation that works fine in the main part of the site turns out to be too tight where the site narrows. These are exactly the details a rough area-based estimate can't surface, because it isn't testing the site's actual shape at all.
Compare scenarios before settling on one
Because testing an irregular site is quick, it's worth generating a handful of scenarios — different access locations, different angles — rather than settling on the first layout that looks reasonable. On a constrained site in particular, the difference between an untested first attempt and a layout that's been compared against two or three alternatives can be meaningful.