The Marathon Garyville Expansion: How Capital Projects Create Parallel Turnaround Opportunities in 2027

The Marathon Garyville expansion program disclosed with Marathon Petroleum's fourth quarter 2025 earnings puts two capital projects on the clock at the largest refinery in Louisiana, both targeted online by the end of 2027. Garyville, at 597,000 barrels per calendar day per EIA, is the fourth largest refinery in the United States, and capital work at a plant of that scale never stays contained to the capital budget. Tie ins, integration outages, and commissioning create work that flows through maintenance and turnaround channels in parallel · which is why the 2027 window at Garyville is worth reading as two overlapping opportunities rather than one.

What capital projects make up the Marathon Garyville expansion?

Marathon Petroleum's Q4 2025 earnings guidance, released February 3, 2026, described two Garyville projects. The first is a feedstock optimization project that raises crude rates by about 30,000 barrels per day, carrying about 110 million dollars of 2026 capital spending plus about 185 million dollars in 2027. The second is a product export flexibility project adding about 10,000 barrels per day of export grade premium gasoline flexibility · mainly compression and reliability upgrades rather than new process equipment · at about 50 million dollars in 2026. Both projects are targeted online by year end 2027, with targeted returns above 25%.

The character of these projects matters for vendors. Neither is a new process unit in the classic sense. The feedstock optimization work debottlenecks the crude side of an existing plant, and the export flexibility work is compression and reliability investment that widens what the refinery can ship. Debottleneck and flexibility projects are executed inside and around operating units, which is exactly what makes them interesting from a maintenance market perspective: their construction, tie in, and commissioning phases have to negotiate with the plant's operating and outage calendar rather than proceeding on a greenfield schedule.

What is an integration outage, and why does this work require one?

Capital modifications to an operating refinery generally reach a point where the new equipment has to be physically connected to live systems · piping tie ins, electrical and instrument connections, and control system integration. A tie in is the physical connection of new piping or equipment into existing systems, and on hydrocarbon or high energy systems it usually cannot be made while the circuit is in service. Operators handle this through integration outages: planned windows, either standalone or piggybacked on other maintenance, during which the affected circuits are cleared, the connections are made, and the systems are returned to service.

Commissioning follows the same logic. Before startup, new and modified systems move through mechanical completion · the verified point at which construction is finished and every component is installed per design · then through instrument and safety system testing, loop checks, functional testing of interlocks and shutdown systems, flushing and drying, and finally the staged introduction of feed. Instrumentation technicians, safety system specialists, inspection services, scaffolding, insulation, and specialty cleaning all bill hours in this phase, and much of it lands in the same craft categories a turnaround consumes. That is how commissioning generally works across the industry; how Garyville sequences its specific windows is not publicly disclosed, and any claim about which units come down when in 2027 is inference from the year end 2027 online targets, not a schedule.

How the work is procured: capital projects group vs plant maintenance

The procurement paths for this work overlap but do not merge. Capital projects at a refiner like Marathon Petroleum are typically managed by a projects organization with its own engineering contractors, its own procurement staff, and its own vendor lists, running on the project's schedule. Turnaround and maintenance work runs through the plant's maintenance and turnaround organization on the outage calendar. An integration outage sits at the seam: the projects group owns the new equipment and its commissioning, while the plant organization owns the outage window, the permits, and much of the supporting craft work.

For vendors, the observation that matters is that the same physical event can have two doors. A scaffolding, insulation, valve, or instrumentation vendor positioned only with plant maintenance can miss project driven scope awarded through the capital side, and vice versa. Vendors who track both organizations tend to see the full opportunity, and because capital project procurement commits earlier than outage procurement, the engagement clock starts sooner than turnaround habit suggests · the general timing argument is laid out in why 2027 turnaround procurement starts in 2026.

The physical context sharpens the point. Garyville's crude side runs two trains, Crude 10A and Crude 10B, at roughly 153,000 barrels per day each. A feedstock optimization project that raises crude rates by about 30,000 barrels per day is, by nature, work in and around that existing crude infrastructure · heat exchange, hydraulics, metering, and the supporting systems that set a train's effective capacity. Work of that character is precisely the kind that competes for the same outage windows, the same inspection resources, and the same craft labor that turnaround scope consumes, which is why the two procurement channels inevitably meet at the schedule.

The 2009 GME precedent: expansions seed future turnaround cycles

Garyville has run this pattern before at much larger scale. The 3.9 billion dollar Garyville Major Expansion added about 180,000 barrels per day and started up in late 2009. The first turnaround of the GME units ran in Q1 2017 per Industrial Info, including a hydrocracker upgrade to 121,000 barrels per day. The sequence is the point: a capital expansion creates new installed equipment, that equipment enters the inspection and catalyst cycle on startup, and several years later it generates its own turnaround demand. The 2027 projects are far smaller than GME, but the same logic applies · capacity added in 2027 becomes maintenance scope in the cycles that follow, on top of a plant that set monthly crude throughput records in Q4 2025 per company results and carries its own cycle exposure into the coming seasons.

Where vendors fit in 2027

The Garyville turnaround picture itself · cycle history, 2027 window inference, and buying center detail · is covered in the Marathon Garyville 2027 turnaround outlook, which remains the hub for that question. This article's narrower observation is that the capital program adds a second, partially independent stream of 2027 work: tie in and integration outage support, commissioning services, compression and reliability scope, and the craft categories that both streams consume. That stacking effect at one site mirrors the broader regional pattern of deferred and cycle due work concentrating into 2027, described in the Gulf Coast utilization reset analysis. Vendors who map both the projects organization and the plant organization at Garyville before the windows firm up tend to be on the right lists when the work is awarded.

Garyville sits within the River Parishes corridor, the densest stretch of the state's turnaround market. ExecGraph tracks the buying centers behind both capital and maintenance work across every corridor on the Louisiana turnaround hub.