What Is a Constructability Review?

A perfectly coordinated set of drawings can still describe a building that is painful to construct. The steel can be detailed correctly and still require a crane that cannot reach the far corner of the site. A wall assembly can be specified correctly and still demand a tolerance no mason will hold. A mechanical room can be dimensioned correctly and still have no path to get the chiller into it after the structure is up.

None of those are documentation errors, so drawing review does not catch them. They are consequences of designing without construction knowledge in the room, and the only way to find them is to walk the design through as a builder would: what goes in first, what does that block, what has to be in place before it, and what happens when it is not.

The leverage curve is steep. At schematic design a constructability finding changes a concept sketch. At design development it changes a system selection. At fifty per cent construction documents it changes details and coordination. At ninety per cent it changes a document that is about to be bid, and after award it is a change order.

This is the argument for engaging a contractor during design — construction manager at risk, design-build, or a preconstruction services agreement. The value is not the estimate. It is having someone in the room during design whose instinct is to ask how a thing gets installed and who has watched the answer go wrong.

The recurring categories are predictable across project types. Access and sequence: equipment with no installation path, work that requires a lift where a lift cannot stand, a sequence that traps one trade behind another. Tolerances: interfaces between a cast-in-place element and a factory-fabricated one where the two tolerance regimes cannot both be satisfied.

Then availability — a specified product with a lead time longer than the schedule allows, or a labour skill scarce in that market — and temporary works, where the design assumes shoring, dewatering or bracing that nobody has drawn or priced. Most findings are unremarkable individually. Their value is that each one is a change order that did not happen.

A constructability review is only as good as the experience of the reviewer and the willingness of the design team to act on it. Reviews performed too late, or by someone reading drawings rather than imagining the sequence, produce lists of observations that everybody acknowledges and nobody incorporates.

The reviews that work are structured around the build sequence rather than around the drawing set, are performed by people who have installed the systems in question, and end with findings assigned to someone with a date. The measure of a good review is not how many comments it generated — it is how many of them changed a drawing.