A tie-in point is the place where new piping, equipment, or instrumentation connects into a plant that is already operating. It is where a revamp stops being a paper exercise and starts touching the live process, and on a brownfield document set it is the single row most likely to be numbered wrong, drawn on one sheet and forgotten on the other, or discovered in the field instead of on the drawing.
This is a reference for the engineer assembling or checking a revamp package, covering what a tie-in point is, how it is marked on the P&ID, what belongs in the tie-in register, the difference between a hot tap and a cold cut-in as a scheduling decision, and where tie-ins tend to fall apart across two drawings instead of one.
Key takeaways
- A tie-in point connects new scope to the existing plant, and it belongs to two drawings at once, the existing P&ID and the new-work P&ID.
- On the drawing it reads as a tie-in flag or TP number at the connection point, a revision cloud or triangle around the new scope, and a match line back to the existing sheet.
- The tie-in register carries a row per tie-in, source drawings on both sides, connection type, isolation method, and shutdown dependency.
- Hot tap and cold cut-in are a scheduling and isolation decision. Hot avoids a shutdown. Cold ties the connection to one.
- Most tie-in failures are document-set failures, numbered on one sheet and not the other, or carried in the register but never clouded on the drawing.
What is a tie-in point
A tie-in point is the boundary between what already exists in the plant and what a revamp is adding to it. It could be a new line joining an existing header, a new branch off an in-service vessel nozzle, or a new instrument landing on an existing junction box or marshalling cabinet. Whatever form it takes, the defining feature is the same, one side of the connection is already built and running, and the other side is new.
That split is what makes a tie-in the highest-risk row in a revamp document set. Every other item in a new-work package lives entirely inside the new scope, described on one drawing, built once, and commissioned once. A tie-in point straddles the boundary. It has to be correctly identified on the existing drawing, correctly connected on the new drawing, and physically executed at a specific location in a plant that does not stop running while the engineering happens. It is also, almost always, what the shutdown window itself gets scheduled around, since the tie-in dictates when the existing line has to come down, for how long, and in what sequence relative to every other tie-in competing for the same outage.
How a tie-in reads on the drawing
A tie-in point is marked with a distinct flag or symbol at the exact location of the connection, carrying a unique identifier, commonly a tie-in number such as TP-014. That number is what ties the drawing mark to the register row, and it has to be unique across the whole revamp, not just the sheet it sits on.
Around the new scope, drafters typically enclose the added work in a revision cloud or triangle, the same convention used for any revision, to separate what is being added from what already exists on the sheet. At the tie-in itself, this cloud usually terminates right at the connection point, since everything on one side of it is new and everything on the other side is existing plant.
Where the tie-in connects to a line or system that is drawn on a different sheet, the drawing carries a match line or continuation reference, a note pointing to the sheet number the connection continues onto. This is what links the new-work drawing to the existing drawing it is physically joining. The convention for exactly how a tie-in flag, cloud, or match line is drawn varies by company and by drafting standard, but the three elements, a numbered flag at the connection, a boundary around the new scope, and a reference to the other sheet, are what to look for regardless of house style.
The tie-in register
The register is where every tie-in in the revamp gets tracked as a single row, independent of which drawing it happens to be drawn on. It is the document a construction planner, an isolation planner, and a commissioning engineer can all work from without opening the full drawing set.
| Column | What it carries | Why it bites |
|---|---|---|
| Tie-in number | The unique identifier for the connection, e.g. TP-014, distinct from any line or equipment tag | If a number is reused or skipped across revisions, two people end up pointing at different physical connections |
| Existing line or equipment tag | The tag on the existing, in-service side, e.g. line 6"-CS-102-A1A or vessel V-201 | This side has to describe the plant as it actually is, not as it was designed |
| New line tag | The tag on the new-work side, e.g. line 4"-CS-3401-A1A | Ties the register row to the new P&ID and the isometric that carries the detail |
| Size | Nominal pipe size at the connection | A mismatch here is cheap to catch on paper and expensive to catch in the field |
| Rating / spec | Piping class or spec at the connection | Confirms the new work is compatible with what it is joining, at the level a register needs |
| Connection type | Flanged, welded, hot tap fitting, and so on | Sets what physically happens at the tie-in, and by extension what the isolation plan looks like |
| Hot or cold | Whether the tie-in is made with the existing line in service or isolated | Determines whether a shutdown window is needed at all |
| Isolation method | Line stop, blind, valve isolation, full depressurization | This is what the shutdown schedule is actually built around |
| Shutdown required (Y/N) | Whether the line, unit, or plant has to come down for this tie-in | Rolls up into the outage duration the whole project is scheduled against |
| Drawing reference, existing side | The existing P&ID sheet and revision the tie-in appears on | Goes stale silently if that drawing revises after the register is built |
| Drawing reference, new side | The new-work P&ID sheet and revision | Same risk, on the other document |
| Status | Planned, cleared for construction, made, verified | Tells a reviewer whether the row describes intent or a completed connection |
The instrument index plays a similar role for new instrument tags, and the line list carries the size and spec detail for the piping side. A tie-in register is really where those two documents meet at a single physical point, which is why its drawing-reference columns matter as much as its technical ones.
Hot tap vs cold cut-in
Once a tie-in is identified, one of the earliest decisions is whether it can be made hot or has to be made cold. This is a scheduling and isolation question first, not a metallurgical or procedural one, and it belongs in the register long before the shutdown plan is finalized.
| Dimension | Hot tap | Cold cut-in |
|---|---|---|
| Line status during the work | Existing line stays in service, under pressure | Existing line is isolated and depressurized before the cut is made |
| Shutdown dependency | Does not require a unit shutdown, and can often be scheduled on its own timeline | Usually requires the line or unit to be down, so it lands inside a shutdown window |
| Isolation method | No isolation of the existing line, since the connection is made through a fitting while flowing | Isolation by block valves, blinds, or a line stop, verified before work starts |
| Scheduling flexibility | High. Frequently proceeds independent of a turnaround | Low. Competes with everything else scheduled inside the same outage window |
| Typical driver for the choice | The line cannot be taken out of service without a wider plant impact | A shutdown window already exists for other work, or isolation is straightforward |
The decision is usually made early enough to shape the whole outage plan. A tie-in confirmed hot early in engineering can come off the critical path of the shutdown entirely. A tie-in that turns out to need a cold cut-in late, after the outage schedule is already fixed, is a scheduling problem discovered at the worst possible time.
Where tie-ins go wrong on the document set
A tie-in point is a single physical connection, but it is described in at least two places, the existing P&ID and the new P&ID, plus the register that is supposed to reconcile both. Most tie-in failures happen in the gap between those documents, not at the connection itself.
Numbered on one sheet, not the other. The tie-in flag and TP number appear on the new-work P&ID, but nobody added the corresponding flag to the existing sheet, or the reverse. A reviewer working from only one drawing has no way to know the tie-in exists at all.
The existing sheet is an as-built that lies. The same drift that shows up across a whole brownfield reconciliation shows up at a single tie-in, where the existing P&ID shows a line size, rating, or routing that was accurate once and has since been changed in the field without a drawing update. A tie-in designed against that line inherits the error, and it is usually found at the point of construction, not on paper.
Clouded in the register, not clouded on the drawing. The tie-in is a live line item in the register, tracked, numbered, assigned an isolation method, but the revision cloud around the new scope was never added to the actual sheet. A discipline reviewer scanning the drawing set visually misses it, because the drawing itself gives no indication that anything changed at that location.
Revision drift between the two sheets. The existing P&ID gets revised for reasons that have nothing to do with the tie-in, and the new-work sheet does not follow. The register's drawing reference for the existing side now points at a superseded revision, and the tie-in as drawn may no longer match the line it is meant to connect to.
A match line without a matching reference. The new sheet shows a match line to the existing sheet at the tie-in location, but the existing sheet's own match line does not reference the new sheet back. The connection is documented in one direction only, which is functionally the same as not being documented at all for whoever opens the existing drawing first.
Keeping both sides of the tie-in in agreement
A tie-in register is only as trustworthy as the agreement between the two drawings it reconciles. A register that lists a tie-in correctly but points at a superseded existing-sheet revision, or a tie-in that is clouded in one drawing but not the other, is not a documentation inconvenience. It is the exact condition that gets a construction crew to the wrong isolation point, or gets a tie-in missed from a shutdown scope until it is too late to add.
That agreement does not hold by itself. It holds because both drawings, and the register between them, get checked against each other on every revision, not just at the point the tie-in was first identified. That is the same discipline that governs the rest of a revamp document set, an MOC process built on documentation that stays true through revisions, rather than one built on a drawing set that was correct on the day it was issued and quietly drifted after.