When the client asked for a column-free sports hall, steel was the assumed answer - long spans and steel portals go together by reflex. But assessed on a whole-life basis rather than a first-cost one, glulam won, and the decision is a useful example of how the framing material is no longer a foregone conclusion for long spans.
Why timber won
The case for glulam was not sentimental; it was a balance of three concrete advantages:
- Embodied carbon roughly half that of the steel option. As carbon starts to carry real weight in decisions - through policy, client targets, or genuine conviction - a structural material that halves the embodied figure is a serious argument, not a footnote.
- Faster erection with lighter, craneable components. Glulam's strength-to-weight ratio means the members could be lifted with smaller cranes and assembled quickly, which fed back into programme and site cost.
- A warm, exposed soffit the client wanted anyway. The structure doubled as the finished ceiling, removing a separate lining and giving the architecture something steel would have needed to hide or clad.
When one option is lower-carbon, quicker to erect, and the finish the client was going to pay for regardless, the whole-life sum can tip decisively even if the bare frame tonnage looks comparable.
The hard parts
Glulam at long span is not a free lunch, and two issues needed real engineering.
The first was the moment connection at the apex of the portal. A long-span portal frame has to resist a substantial moment where the rafters meet, and developing that in timber is harder than in steel, where you simply weld or bolt a stiff joint. The solution here was a glued-in rod connection, which was prototyped and load-tested before sign-off - a reminder that timber moment connections often warrant physical verification rather than calculation alone.
The second, and the real negotiation, was fire:
Fire was the real negotiation: a charring-rate calculation plus sprinklers satisfied the authority without intumescent coatings.
Timber's fire behaviour is, counter-intuitively, one of its strengths in large sections. It chars at a predictable rate, and the char layer insulates the sound timber beneath, so a large glulam member retains structural capacity for a calculable period. Designing to that charring rate, supported by sprinklers, met the fire requirement without coatings - keeping the exposed timber the client wanted.
What the example shows
The lesson is not "timber always wins" - it does not. It is that the material decision for a long-span roof deserves a genuine whole-life comparison rather than a default:
- Compare on whole-life terms - embodied carbon, erection, and the finish - not just frame first-cost. The answer can flip.
- Price the hard details honestly. Timber's moment connections and fire strategy are real work; a fair comparison includes them, including prototyping where the connection warrants it.
- Let the brief into the structural choice. Here the client's wish for an exposed warm soffit was not a constraint on the structure - it was part of what made the structure the right one.
The related precast-vs-in-situ framework applies the same whole-life, decision-led thinking to concrete, and the form-follows-force article covers letting an exposed structure become the architecture.