Home · Blog · Predictive Maintenance vs Fixed Interval Cycles: How Louisiana Refineries Schedule Turnarounds
Louisiana Turnarounds9 min read

Predictive Maintenance vs Fixed Interval Cycles: How Louisiana Refineries Schedule Turnarounds

How Louisiana refineries blend predictive maintenance with fixed interval turnaround cycles, and why windows at CITGO and Shell Norco moved on the calendar.

Published July 3, 2026

Predictive maintenance is reshaping how Louisiana refineries schedule turnarounds. For decades, facilities including CITGO Lake Charles, Shell Norco, ExxonMobil Baton Rouge, and Marathon Garyville planned major maintenance on fixed interval cycles, with each process unit coming down every 4 to 6 years largely regardless of its measured condition. That model is giving way to a hybrid approach in which vibration analysis, wall thickness monitoring, and process analytics inform whether a unit can keep running or needs to come down early. For vendors selling into Louisiana turnarounds, the shift matters because it changes when windows get set, why they move, and when procurement actually begins.

Direct answer: Yes, Louisiana refineries are moving toward condition based turnaround scheduling, but the fixed 4 to 6 year cycle remains the planning skeleton. Vibration analysis, wall thickness monitoring, and process analytics let operators extend runs when equipment health and margins support it, and pull events forward when they do not. That flexibility is why published windows move.

The shift from fixed 4 to 6 year cycles toward predictive maintenance and condition based scheduling

Louisiana refineries along the Mississippi River corridor and in Lake Charles historically scheduled turnarounds on the calendar, not on the condition of the equipment. A turnaround is a planned, facility scale maintenance event in which a process unit or group of units is shut down, opened, inspected, repaired, and returned to service, typically over several weeks. Under the classic fixed interval model, a crude unit or fluid catalytic cracker went into the plan for a date roughly 4 to 6 years after its last event, and the planning machine worked backward from that date.

Condition based scheduling changes the input. Instead of asking how long it has been since the last event, reliability teams ask what the equipment is telling them. Three data streams carry most of the weight:

  • Vibration analysis is the continuous or periodic measurement of vibration signatures on rotating equipment such as pumps, compressors, and turbines. Changes in those signatures reveal bearing wear, imbalance, misalignment, and looseness months before functional failure, which lets planners decide whether a machine can safely run to the next window.
  • Wall thickness monitoring uses ultrasonic testing and, increasingly, permanently installed sensors to track metal loss on piping, pressure vessels, and exchanger shells. Trending corrosion rates convert into remaining life calculations that either support a run extension or force a unit onto the schedule earlier than its nominal cycle.
  • Process analytics covers unit performance data: catalyst activity decline, exchanger train fouling, furnace tube skin temperatures, and pressure drop trends. When a unit approaches end of run conditions economically rather than mechanically, the same math applies in reverse.

None of this eliminates the cycle. Inspection codes, catalyst life, and mechanical integrity programs still anchor major events to a multi year rhythm. What condition data adds is a defensible reason to flex that rhythm in either direction, and Louisiana operators have been exercising exactly that flexibility.

How predictive programs extend some run times and pull others forward

CITGO Lake Charles and Shell Norco offer the clearest recent Louisiana examples of calendar flexibility in both directions. CITGO ran a multi unit turnaround at Lake Charles from late February to late April 2024 covering the roughly 50,000 bpd FCC-A, two reformers, a naphtha hydrotreater, and Crude Topper C, one of the site's four crude trains, per Bloomberg and trade reports. Trade reporting then indicates CITGO deferred further Lake Charles turnaround work into 2026 during a strong margin period. That is a run extension decision: equipment condition and economics together supported pushing work to the right. The CITGO Lake Charles turnaround schedule page tracks how that deferred work resolves, with ExecGraph tracking listing a possible Q4 2026 event, labeled possible rather than confirmed.

Shell Norco moved the other way. The refinery's 2025 turnaround began August 15, 2025 and ran roughly 50 days into October, covering the RCCU, the residual catalytic cracking unit that upgrades heavy feed into lighter products, along with the 14,800 bpd alkylation unit, the 40,000 bpd naphtha hydrotreater, and the GO-1 ethylene unit, per Reuters as carried by trade press. Trade reporting indicates that start was pulled forward from a mid September plan to August 15. A window that moves several weeks earlier compresses every vendor deadline attached to it: mobilization, material staging, and final scope confirmation all shift with the start date.

These two cases illustrate the operator side of a question ExecGraph covers from the outside in elsewhere. The companion piece on how refinery turnaround timing is predicted deals with reading external signals to anticipate events. This article is about the internal scheduling philosophy that generates those signals in the first place. Vendors who understand both sides tend to interpret a moving window as information rather than noise.

Why condition based scheduling changes the procurement timeline

Louisiana turnaround procurement was already front loaded under the fixed interval model, and condition based scheduling stretches it further in both directions. When a facility like CITGO Lake Charles defers work into a later year, the long lead procurement that was already in motion does not simply pause. Exchanger bundles, alloy piping, and valve orders placed against the original window either get stored, rescheduled, or absorbed into interim maintenance. Contractors who were holding craft capacity for the original window release it, and the facility re enters the queue for that capacity when the new date firms up.

When a window pulls forward, as at Shell Norco in 2025, the compression lands hardest on whatever had not yet been bought. A start date that moves from mid September to August 15 removes about a month from every remaining procurement lane at once. The vendors who absorb that compression well are generally the ones who were already engaged with the planning team before the move, because they hear about the shift when it is being considered rather than when it is announced.

The practical consequence for the Louisiana market is that procurement timing now correlates with data milestones as much as with calendar dates. Inspection campaigns, run length reviews, and margin conditions all feed the decision, and the operators running the most mature predictive programs make those decisions on a rolling basis. The pattern that separates consistently successful vendors is engagement keyed to the facility's planning cadence rather than to a published date that may move twice before execution.

What condition based scheduling means for scope definition at Louisiana facilities

At Louisiana facilities from Norco to Lake Charles, condition data does not just move the date of a turnaround. It reshapes what is in it. A scope freeze is the milestone after which no new work is added to a turnaround without a formal exception, and best practice places it 18 to 24 months before execution. Predictive programs interact with that milestone in two ways. First, condition monitoring shrinks discovery scope: when wall thickness trends are known in advance, fewer surprises emerge once equipment is opened, and the work list is more accurate at freeze. Second, a late run of inspection data can force scope in after the freeze, which is exactly the high cost scenario examined in the companion piece on scope freeze and the cost of late additions.

For vendors, the observable effect is that scope conversations at condition driven facilities happen earlier and reference data rather than habit. A reliability engineer who can show three years of thickness readings on an exchanger shell does not need a turnaround to confirm the repair is required, and the sourcing conversation for that repair can start whenever the data crosses a threshold. Vendors positioned with reliability and inspection teams, not just turnaround procurement, see that work first. The facility level context in the Shell Norco 2027 outlook shows how one completed event plus independent unit cycles frame the next set of windows.

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

Find the decision makers at every facility mentioned above

ExecGraph maps 48,075 verified decision makers at 1,331 Gulf Coast operators in 11 markets, organized by department, seniority, and purchasing authority.

Book a 1 hour walkthrough

60 minute walkthrough. We'll map the decision chain at the facilities in this post.