Why Does a Title 24 Design Not Comply?
Nearly always one of three things: too much glazing for the prescriptive limits, equipment below the required efficiency, or a control that was never specified. None of them means the design is wrong — they mean the compliance route or one input has to change.
Glazing is the most common single cause
The prescriptive path sets limits on glazing area and performance, and a design that is architecturally excellent can exceed them comfortably. A heavily glazed elevation, a corner window, a wall of sliding doors to a garden — all normal, all capable of failing a prescriptive check.
The answer is usually not less glass. It is the performance path, which models the whole building and lets a generous glazed area be offset by better performance elsewhere. That has to be modelled rather than asserted, but it is the normal resolution and it usually works.
Equipment below the minimum
Every piece of regulated equipment has a minimum efficiency, and a unit chosen on price or availability frequently sits under it. This is the failure that most often appears late, because equipment is selected after the design is settled and sometimes substituted after the calculations were done.
It is also the cheapest to fix at specification stage and the most expensive to fix on site. A model number checked before an order is placed avoids the entire problem.
A control that was never drawn
Controls are where a design fails without anyone having designed anything badly. The equipment is right, the envelope is right, and a required occupancy sensor, daylighting control, economizer or shutoff is simply absent from the drawings because nobody identified it as a requirement.
On commercial projects this is very common, and it compounds: a missing control can also mean a missing acceptance test, which is a scheduling problem at the end of the job rather than a drawing revision.
Two things that are not really failures
A missing input. Software cannot assume an unstated assembly, so an incomplete drawing set produces a non-compliant result that is really a documentation gap. Naming the construction of each assembly and the use of each space resolves it without changing anything.
And the wrong compliance route. A project assessed prescriptively when it was always going to need performance has not failed; it has been assessed against the wrong test.
Why finding out early is the whole game
Non-compliance discovered during design is a specification conversation. The same non-compliance discovered at plan check is a resubmittal, and discovered on site it is installed plant. The result is identical in each case — what differs by an order of magnitude is what it costs to resolve.
That is the argument for running the calculations while the drawings are still moving rather than treating them as a permit formality at the end.