Turnaround Scope Freeze and the Cost of Late Additions

The turnaround scope freeze is the most consequential date in refinery maintenance planning, and at Louisiana giants like Marathon Garyville and ExxonMobil Baton Rouge it lands surprisingly early: best practice sets the freeze 18 to 24 months before the first unit comes down. Everything about how a turnaround performs against its budget and schedule traces back to how well that freeze held. For vendors selling into Louisiana turnarounds, the freeze date, not the event date, is the real commercial deadline.

What a turnaround scope freeze is and why best practice sets it 18 to 24 months out

At Louisiana refineries and chemical plants, a turnaround is the planned shutdown, inspection, and overhaul of a process unit or group of units, and its scope is the approved list of work orders the event will execute. The scope freeze is the governance milestone after which that list closes: new work is admitted only through a formal exception process, usually requiring turnaround leadership or site executive signoff with a documented reason the work cannot wait for the next cycle.

The 18 to 24 month convention exists because the longest procurement and planning chains in the event need that much runway. Alloy heat exchanger bundles, specialty valves, furnace tubes, and large fabrications carry lead times measured in quarters, not weeks. Detailed job planning, the conversion of a one line work request into an estimated, sequenced, materials attached work package, takes months across hundreds of packages. Contractor bidding on a defined scope, and craft labor commitments against a defined peak, both require the scope to stop moving. Freeze the list 18 to 24 months out and all of those chains can run at normal cost. Every week the list stays open past that point transfers cost from the plan to the event. The procurement side of this same clock is covered in the pillar on why 2027 turnaround procurement starts in 2026.

Why late scope additions drive most turnaround cost overruns and schedule extensions

Across Gulf Coast turnarounds, the mechanics of a late addition explain why it costs multiples of the identical work planned early. Three mechanisms compound.

• Materials lead time. Work added six months before execution needs materials that normally take twelve. The options are expediting premiums, air freight, substitution of available but costlier metallurgy, or accepting that the material arrives mid event. Every option is a premium over catalog price and normal logistics.
• Crew density. The labor curve was built to a planned peak. Late work either extends the schedule or adds workers to units already at planned density, and congested units execute at degraded productivity, so a late workhour buys less than a planned one. Recruiting incremental qualified craft close to an event, in a market where neighboring facilities are drawing on the same labor pool, adds its own premium.
• Schedule logic. A turnaround schedule is a network of dependencies: blinds, scaffold, permits, lifts, and inspections sequenced across shared cranes and shared access. Inserting work rewires that network. The insertion often idles other crews while it resolves, and if the added work touches the critical path, the whole event extends, with every day of extension carrying the full daily cost of the standing army plus the margin loss of the unit staying down.

Some late scope is unavoidable: opened equipment reveals real conditions, and inspection findings cannot be deferred by governance. The distinction operators draw is between discovered work, which planning can shrink but not eliminate, and discretionary late additions, which governance exists to stop. Tools that shrink the discovered category, like laser scanning and model based planning, are examined in the companion piece on digital twins and 3D documentation in Louisiana turnarounds.

How the largest Louisiana refineries run scope challenge processes

Marathon Garyville, at 597,000 bpcd per EIA the largest refinery in Louisiana and the fourth largest in the United States, and ExxonMobil Baton Rouge, at 522,500 bpcd the sixth largest, illustrate why scope governance at this scale is a standing process rather than a single meeting. At facilities this large, a major event's work list assembles from thousands of candidate items contributed by operations, inspection, reliability, and capital projects, and the raw list always exceeds what the window can absorb.

The scope challenge is the filter. Speaking generally, large operators convene structured reviews in which every proposed work order must defend itself against criteria along the lines of: what failure or regulatory requirement drives this work now; what happens if it waits for the next cycle; can it be executed with the unit online instead. Items that survive get estimated and planned; items that do not are deferred, converted to routine maintenance, or dropped. The process runs in passes as the freeze approaches, and its discipline is a major determinant of whether the freeze holds, because a weak challenge lets marginal work in early and forces painful exceptions later. Cycle position shapes the intensity too: the forward looking scope questions at Garyville, where 2027 project startups create tie in and integration windows, are laid out in the Garyville 2027 outlook, and operator level procurement context sits on the Marathon Petroleum sell to page.

The vendor takeaway on timing

For vendors watching Louisiana facilities, the freeze reframes the calendar: the commercial window for shaping what a turnaround buys closes 18 to 24 months before the unit ever comes down. Vendors who engage while scope is still forming, when a reliability engineer is deciding whether a repair makes this cycle's list, tend to find their solutions written into work packages and their lead times reflected in the plan. Vendors who first appear after the freeze encounter a closed list, committed materials budgets, and buyers whose incentive is to say no to anything new. The exception traffic that does occur after freeze flows almost entirely to suppliers already qualified and already known, because expedited work is exactly when a buyer reaches for the vendor that requires no introduction.

The current cycle makes the arithmetic immediate. Kpler in February 2026 framed a heavier US maintenance cycle building through the second half of 2026 into 2027 on 4 to 5 year cycles following the 2021 to 2023 wave, and Industrial Info called 2026 a busy US refinery maintenance year with 1.1 billion dollars of Q1 2026 kickoffs. Applied against an 18 to 24 month freeze convention, a 2027 or early 2028 Louisiana event is freezing its scope during 2026. The influence window for the heaviest stretch of the coming cycle is, in other words, open now and closing facility by facility.

This is an observation about how the market allocates work, not a script. But the pattern across the Louisiana corridors is consistent: the vendors who treat the freeze date as the deadline, and who track which facilities are 18 to 24 months from their next window, are engaged during the months when scope decisions are actually being made. ExecGraph tracks the planning, reliability, and procurement roles who make those decisions at every major Louisiana facility.

ExecGraph maps the verified buying center at every Louisiana facility named above. See how ExecGraph works at /pricing.