The frame and the envelope under one subcontract

On paper, splitting the steel package from the cladding package looks like good procurement. Two trades, two specialisms, two competitive tenders. In practice it creates one interface, the line where the cladding fixes to the frame, and that line is where most envelope disputes on a steel building begin.

The interface is commercial, not just physical

The reason the split causes trouble is that the interface is not only a joint between two materials. It is a joint between two contracts.

The cladding contractor's system assumes a tolerance for the steel it fixes to. The steel contractor works to the tolerance in their own specification. Both can be comfortably within their own stated tolerances and still not fit each other, because nobody owned the sum of the two.

When the roof then leaks, or a panel line runs visibly out of true across a long elevation, each party can point at the other's compliance certificate entirely honestly. Nobody is lying. The gap was in the procurement, not in the workmanship, and it was there from the day the packages were split.

Tolerance is the part that gets missed

It is worth being concrete about what accumulates.

Steel is fabricated to a tolerance and erected to another. Cladding rails have their own. Panel systems have a fixing tolerance and, separately, a limit on how much out-of-plane deviation they can take up before the fixings start to strain or the line becomes visible. Each of those allowances is individually reasonable. Added together across a long elevation, they are not automatically compatible.

Someone has to check that the sum works before anything is fabricated. Where the packages are separate, that person often does not exist. Both contractors are checking their own half against their own specification, and each is correct.

It shows up in the programme too

The second cost of the split is sequencing.

Cladding cannot start until enough frame is up and surveyed. The frame contractor's programme is not usually written around the cladding contractor's mobilisation, and there is no particular reason it would be. The survey itself takes time, and if it finds anything, the resolution sits between two parties who each have their own float to protect.

Where the two are separate, that gap is nobody's float to manage. Where they are one subcontract, it is an internal sequencing problem rather than a claim, which is a materially different thing to deal with when the programme is already tight.

What one subcontract actually changes

Roofing and cladding systems are named in Masas's ISO-certified scope alongside structural steel and pre-engineered buildings, so the frame and the envelope can come under one subcontract. The ISO 9001, ISO 14001 and ISO 45001 certificates all name construction, installation and erection within that scope, not fabrication alone, so one subcontractor carries the package from shop drawings through to handover.

The practical effect is that the tolerance question is settled inside one organisation before fabrication rather than between two after erection, and the sequencing question is a planning decision rather than a negotiation.

Be clear about what it does not change

None of this makes the combined route automatically cheaper on the tender return. It usually is not, and a client comparing two numbers on a spreadsheet will see that immediately. Two specialists bidding two packages competitively will often produce a lower total than one subcontractor bidding both.

What the combined route changes is where the risk sits after award. If the envelope fails, there is one party to call, and no argument available to them about whose tolerance was at fault. You are buying a defined point of responsibility, and it is reasonable to ask what that is worth on a particular building rather than assuming it is always worth the same.

When to make the trade, and when not to

The trade is worth making on buildings where the envelope is doing real work. A long roof run, a large clear span, an insulated store, a cold store, anything where a leak is expensive rather than merely annoying, and anything where the elevation is long enough that a visible line problem cannot be hidden.

It is also worth it where the programme has no slack, because the sequencing gap between two packages is one of the easier ways to lose two weeks that nobody planned to lose.

On a small, simple, forgiving building, two packages may well be the right call, and it is fair to say so. A short roof with a steep pitch and a modest elevation does not accumulate enough tolerance or enough movement for the interface to be the main risk, and the tender saving is real money.

The question to ask is not which route is better in general. It is how much this particular envelope has to absorb, and what a failure at that interface would cost to resolve once the building is up.