The client had signed tenants before they had a slab. That single fact shaped every decision on this project: the programme was not a target, it was a contractual obligation with penalties attached to specific bay numbers.
Why design–build won here
A traditional tender would have put six months between concept approval and first pour. We took the design risk instead, ran authority approvals in parallel with earthworks, and started piling while the envelope package was still in detailed design.
The compression is paid for elsewhere. It means committing to structural grids before the tenant fit-out requirements are fully known, which in turn means designing in tolerance — wider bays, heavier slab specification in zones we suspected would take racking, and spare conduit capacity nobody had asked for.
The earthworks problem
The site fell 11 metres across its diagonal. The obvious answer — balance cut and fill on site — collided with a wadi easement that could not be built over and could not be filled.
We re-graded the masterplan into three terraces instead of one platform. It cost the client 4% of yard area and saved roughly seven weeks against trucking spoil off site, plus the haul-road maintenance that would have come with it.
Terracing a logistics yard is heresy to a yard planner and gospel to an earthworks engineer. The compromise was three levels instead of five.
Cold chain, commissioned twice
The refrigerated block is the part that keeps people awake. We commissioned it dry, signed it off, then re-commissioned it under load with the client’s own pallets in place — because a cold room that holds temperature empty tells you almost nothing about one holding 4,000 pallets of stock.
The second pass found a defrost cycle timing fault that would have cost the tenant product within a fortnight of go-live.
What we would do differently
Two things. We would fix the dock-door schedule earlier — late changes to door positions rippled into the apron pours and cost us a fortnight of rework. And we would resource the as-built documentation from month one, rather than scrambling to assemble it in the final quarter while the site team was already demobilizing.