Most capital works projects are judged, right up until the final weeks, on construction progress. Concrete poured, pipework installed, equipment on site — the metrics that fill a monthly report all point one way. Then commissioning starts, and the project that looked 95% complete suddenly stalls for another six to ten weeks while nobody quite knows why.
This isn't bad luck. It's a predictable outcome of how most capital projects are planned, resourced and reported. Commissioning is treated as the final line item on a construction program rather than as its own discipline with its own risks, its own resourcing needs and its own failure modes. By the time it starts, most of the budget, attention and senior oversight has already moved on to the next job.
Commissioning fails for the same few reasons, every time
It's planned as a duration, not a process
A commissioning allowance often shows up on a program as a block of weeks with no detail behind it — no sequence of hold points, no defined test criteria, no named responsibility for sign-off. When the block runs out and the plant still isn't performing to spec, there's no structure to fall back on, just pressure to declare it finished.
The people who understand the asset weren't involved early enough
Operations and maintenance staff know exactly how a plant needs to behave day to day. If they're introduced at commissioning instead of during design and construction planning, you find out about their real requirements at the worst possible time — when changing anything is expensive and the pressure to hand over is highest.
Interface risk between packages gets missed
On any project with multiple contractors or equipment packages — mechanical, electrical, controls, process — commissioning is where every interface between those packages gets tested simultaneously for the first time. If nobody has mapped those interfaces and agreed who's responsible for what at each boundary, commissioning becomes a live debugging exercise, and disputes over which contractor caused a fault eat weeks.
Documentation catches up after the fact instead of running alongside
Factory acceptance testing, inspection and test plans, as-built records — when these are treated as commissioning administration rather than a continuous thread from procurement onward, someone ends up reconstructing test evidence under time pressure, which is exactly when errors and gaps get missed.
There's no clear owner
Construction has a superintendent or site manager. Design has a lead consultant. Commissioning frequently has nobody with the authority, the technical depth and the time to own it end to end — so it gets shared informally between whoever's still on site, and accountability disappears with it.
"Construction has a superintendent. Design has a lead consultant. Commissioning often has nobody — and the project that looked 95% done stalls for weeks finding out why."
What good commissioning management actually looks like
The projects that commission cleanly share a few habits, and none of them are expensive.
- Commissioning planning starts at design, not at practical completion. The commissioning strategy — test sequence, hold points, acceptance criteria, who signs off what — should exist as a live document from early design development, updated as the project moves through procurement and construction. It should never be written for the first time once construction is finishing.
- Factory acceptance testing is treated as a genuine gate, not a formality. Catching a performance shortfall on an evaporator, a pump skid, or a control system at the factory, before it ships, is a fraction of the cost and disruption of catching the same fault on site during commissioning with the plant partially live. FAT needs real technical scrutiny and real authority to reject, not a witnessed box-tick.
- Operations input is built into design review, not bolted on at handover. The people who will run the asset should have a seat at design and constructability reviews. Their feedback on maintainability, access, and operational behaviour is cheapest to act on before anything is built.
- Interface responsibility is mapped, named and agreed before commissioning starts. Every boundary between packages — mechanical to electrical, electrical to controls, new works to existing plant — needs a named responsible party and an agreed test method, documented before the first system is energised.
- One person owns commissioning, with real authority. Not a committee, not "whoever's available that week." Someone senior enough to make sequencing calls, hold a contractor to an acceptance criterion, and escalate early when something isn't going to be ready — because early escalation is what protects a handover date; late escalation just moves the argument closer to it.
The cost of getting this wrong
Poor commissioning management doesn't usually show up as a single dramatic failure. It shows up as slipped handover dates, defect lists that keep growing after "practical completion," disputes over who's responsible for a fault at a package boundary, and operations teams left to work around problems that should have been resolved before they took control of the asset. For asset owners, that's the point where a capital project's real cost — reputational as much as financial — gets locked in.
None of this requires more bureaucracy. It requires commissioning to be planned, resourced and owned with the same seriousness as construction, from the start of the project rather than the end of it.
Give commissioning a named owner before it becomes a crisis.
S3NTEC leads commissioning management on capital works projects across water, industrial and civil infrastructure — from FAT through to operational handover.
Talk to Us →