Ask a process engineer where the real P&ID is and, if they're honest, they'll point at the marked-up print on their desk. The controlled copy in the document system is close. The one with the pencil on it is true. Every field change, every 'we moved the drain,' every valve that went in a size up because that's what was on the truck, it's on that print, and often only on that print.

At the end of the job those redlines are supposed to roll back into the drawings and become the as-builts: the record of what actually got built, handed to the owner, and inherited by whoever touches the plant next. Supposed to. In practice the project's winding down, the budget's spent, the good people have rolled to the next job, and a stack of marked-up prints gets scanned to PDF and called 'as-built' without anyone reconciling the marks back into the native drawings.

Why as-builts quietly become fiction

Nobody decides to ship bad as-builts. Reconciling redlines is unglamorous work at the exact moment there's no budget and no schedule left for it. The field marks the print. The mark is real. But rolling it back into the P&ID, the line list, and the datasheet is a separate task nobody's assigned, so the scanned redline becomes the record and the native drawings stay at the last issued revision, which is now wrong.

• The marked-up print is the real master; the controlled drawing is a revision or two behind reality.
• 'As-built' ends up meaning 'a scan of the redlines,' not 'the drawings updated to match.'
• A valve swapped in the field for whatever was on the truck never makes it back to the datasheet or the P&ID.
• The next project forks this document set as its base and inherits every redline that never got reconciled.

The cost lands on the next job

As-built drift usually doesn't bite the project that created it. It bites the turnaround two years later, when someone pulls the 'as-built' P&ID to plan a tie-in and it doesn't match the plant: wrong line size, a valve that isn't there, a drain the drawing never got. Or it bites your own team, when you fork last year's job as the base for this year's and quietly carry forward every mark that never made it home. The redline was accurate the day it was drawn. Two years of not-reconciling is what makes it dangerous.

The marked-up print on the desk is the most accurate drawing on the project, and the only one that isn't backed up, version-controlled, or in the set you'll hand the owner.

How Kord gets the redlines home

• Field markups are captured against the actual drawing version, pinned to the line, the region, the cell, instead of living only on a print on someone's desk.
• Each redline becomes a change you can review and roll into the native document, so 'as-built' means the drawing was updated, not that the marks were scanned.
• The consistency check catches the field valve swap that made it into the P&ID redline but not the datasheet, before the set gets called final.
• The as-built set releases as a versioned snapshot tied to the redlines it came from, so the record is traceable to the field marks it's supposed to represent.
• When the next project forks this set as its base, it forks a reconciled set, not last year's redlines nobody closed out.

The marked-up print exists because the field is where the design meets reality, and reality wins. That's not the problem. The problem is that the truest drawing on the project spends its life as one pencil copy that never gets reconciled back into the record. Get the redlines home while the people who drew them are still on the job, and the as-builts are worth the paper they're printed on, and the next fork starts from something true.