Turnaround vs. new-unit startup: two different clocks in refinery scheduling

In refinery and LNG scheduling, people outside the fence line often use "turnaround" and "startup" like they describe the same kind of event. They don't, and the schedule risk behind each one is different: a turnaround is a planned interruption to something already running, a new-unit startup is the first time a unit runs at all.
What a turnaround actually is
A turnaround, TAR for short, is a scheduled shutdown of an operating unit. Operators take a crude unit or an FCC offline on a known cycle, usually every few years, to do the work that can't happen while the unit is running: catalyst change-out, vessel entry, tie-ins for debottlenecking, pressure relief valve testing. The scope gets locked months ahead through a work list, contractor crews show up for a fixed window, and the unit comes back up close to where it left off. Duration runs days to a few weeks. The main way a TAR blows its schedule is scope growth found once the unit is opened up, corrosion worse than expected, a vessel needing more repair than the work list assumed.
What a new-unit startup is
A new-unit startup is commissioning: the first time a system that has never operated before gets water-flushed, pressure-tested, and brought onto hydrocarbon feed. There's no prior operating run to fall back on. Every instrument loop gets checked out for the first time, every interlock gets proven, and the unit has to hold design rates before anyone calls it finished. This covers a new LNG train, a new hydrocracker, an expansion tying into an existing plant, anything coming onstream for the first time rather than resuming after maintenance.
Why "startup date" means something different depending on which one you're reading
In a turnaround, the date that matters is the resumption date, when the unit rejoins the production slate at normal rates. It's a tight window, measured against the unit's own operating history.
For a new-unit project, "startup date" (also written as onstream date, first production, or COD) sits much further downstream. It comes after mechanical completion, after commissioning, after the unit has run long enough to show sustained output near design capacity. That's the date written into offtake agreements, project finance covenants, and the schedule EPC contractors report against. It's also the date most exposed to slip, because everything upstream of it, procurement, module fabrication, module set, structural steel closing out, has to land before commissioning can even start.
That's the real difference for anyone pricing risk against a calendar date. A turnaround running two weeks long is a known, bounded event against an asset already sitting in the supply balance. A new-unit startup slipping two months resets a financing model, a cargo commitment, or a hedge book built around a date that hasn't happened yet. And the only read most outside parties get on it is whatever the contractor's own schedule update says.
Watching physical progress on-site has an edge over waiting on a status memo. Module placement, structural steel going up bay by bay, piping runs closing out, all of it is visible from overhead months before anyone inside the project issues an update. That's the independent view behind tracking refinery and LNG construction progress against a planned startup date, built for people carrying financial exposure to a date they don't control.
If a startup date somewhere on your book is still ahead of mechanical completion, it's worth watching the build itself rather than the next schedule memo.