Plant Shutdowns & Turnarounds

Turnaround cost is dominated by lost production rather than by the contractor. For a pharmaceutical or biotech plant, a day of lost output frequently exceeds the entire mechanical cost of the work being done, which is why compressing the window is worth paying for. On the contractor side, expect crew rates of $105 to $185 per hour per person depending on region and trade, with shift premiums of 1.5 to 2 times for nights and weekends, and mobilisation as a separate fixed line. A modest turnaround with a crew of six over five days lands in the $40,000 to $90,000 range for labour before materials. The controllable variable is planning, since work packaged and pre-fabricated in advance runs at a fraction of the duration of work scoped on the day.

Most process plants run an annual or biennial turnaround, with the interval set by the equipment that most constrains it rather than by a general rule. Pressure vessels under a mechanical integrity programme may require internal inspection on a five year cycle. Compendial water and clean steam systems are usually opened for passivation or derouging on a one to three year basis, driven by inspection evidence rather than the calendar. Heat exchanger bundles are pulled on fouling rate. Rotating equipment follows condition monitoring. The practical approach is to establish the shortest genuinely required interval among those, then batch every other outage-requiring task into that same window, because a second shutdown almost always costs more than doing extra work during the first.

Overrun is the defining risk of turnaround work and it compounds. Production restart slips, which on a regulated site can cascade into batch scheduling, stability commitments and supply obligations. Crews frequently have follow-on commitments elsewhere, so extending is not simply a matter of paying more. And on compendial systems the qualification tail sits after the mechanical work, so a mechanical overrun delays sanitisation and sampling, which are calendar-driven and cannot be compressed at all. The usual causes are discovery work found once equipment is opened, materials not staged in advance, and permits or isolations not ready when crews arrive. Contingency should be planned as scope that can be dropped rather than as extra days.

Scope freeze is the point in turnaround planning after which no new work may be added to the outage without formal approval, typically set six to twelve weeks before the window opens. It exists because every added job needs materials procured, permits written, isolations planned, crews sized and a place in the sequence, and work added after freeze has none of those. In practice a freeze is enforced with a small exception process for genuine emergent findings, since equipment opened during the turnaround does reveal problems. Plants that do not enforce a freeze consistently overrun, because the schedule is built on a scope that keeps changing underneath it. The freeze is a planning discipline rather than a contractual formality.

Four approaches reduce or avoid the outage. Installing isolation valves and capped future connections during an earlier planned shutdown lets later tie-ins be made while the plant runs, and this is by far the cheapest option if a second phase is even remotely likely. Temporary bypass piping keeps production going while a section is isolated, though on product-contact systems the bypass must itself be cleanable. Sectional isolation, where only one loop leg or one train is taken down while the rest runs, works where the system was designed with sectional valving. And phased turnarounds take one area at a time across several shorter windows, which suits multi-train plants better than single-train ones.

Isolation authority almost always sits with the plant rather than the contractor, because the plant owns the process safety system and the lockout programme, and a contractor cannot verify that a valve is not passing on a system they do not operate. The contractor is responsible for working within that isolation, for their own permit-to-work compliance, and for not breaking the boundary. In practice the failure mode is ambiguity: a boundary assumed isolated by one party and considered live by the other. Agreeing in writing who applies locks, who verifies isolation by physical proof rather than valve position, and who authorises breaking containment is the single most important pre-turnaround document after the schedule.

Duration follows scope and preparation rather than plant size. A focused outage covering a handful of tie-ins and one or two equipment items typically runs three to seven days. A general turnaround including vessel entries, heat exchanger bundle pulls, instrument work and piping modifications commonly runs two to four weeks. A major turnaround with significant capital work can extend to six weeks or more. On a regulated site, add the qualification tail after mechanical completion: passivation, sanitisation and the sampling campaign on a compendial water system routinely add one to three weeks before production can restart, and that time is set by laboratory incubation rather than by crew size.

Diagnose the constraint before adding people, because adding crews to a congested space slows work rather than accelerating it. Check whether the delay is materials, permits, isolations, discovery work or genuine productivity. Materials and permits are planning failures and are usually recoverable by expediting. Isolation delays mean the boundary was not ready and need plant rather than contractor action. Discovery work is scope growth and should go through the exception process rather than being absorbed silently, since absorbing it is what hides the overrun until it is too late to react. Genuine productivity shortfalls usually indicate too many trades in one area, and resequencing beats resourcing. Protect the qualification tail last, because that time cannot be recovered.

Before the window opens: scope frozen with an exception process agreed; every work package written with materials identified and staged on site; long-lead items received and inspected; permits drafted and isolation plans agreed with named authority; contractor safety prequalification and site induction complete; lifting and rigging plans reviewed for anything non-routine; waste handling and disposal arranged including any spent chemistry; and the sequence built with trade congestion considered rather than assumed away. Critically for regulated sites: the qualification plan written and the validation resource booked, since discovering at mechanical completion that nobody scheduled the sampling campaign is a common and entirely avoidable way to lose two weeks.

Five causes recur. Scope creep after freeze, where emergent work is absorbed without extending the schedule until the overrun becomes unavoidable. Materials not staged, so crews wait rather than work. Isolation and permits not ready on day one, losing the first shift of a window that has no slack. Trade congestion, where too many crews are sequenced into one area and interfere with each other. And the qualification tail being omitted from the plan entirely, so mechanical completion is treated as the finish line when the plant cannot actually restart until sanitisation and sampling complete. Only the fourth of those is a delivery problem; the rest are planning failures that were decided weeks before the outage began.

Screen on planning capability rather than crew size, because turnarounds are won and lost before anyone arrives. Ask how they build work packages and at what point before the window those are complete. Ask how they stage materials and who verifies receipt. Ask how they handle emergent discovery work and what the exception process looks like. Ask how they sequence trades to avoid congestion, and whether they will provide a resource-loaded schedule rather than a bar chart. For regulated plants, ask specifically who plans the qualification tail and whether validation resource is booked alongside the mechanical scope. Paul Industries delivers turnaround planning, mechanical execution, passivation and validation under one contract nationwide.