Scoping a turnaround from inspection history

Most turnaround inspection scopes are the last turnaround's scope with a few additions. Everyone knows this is not how it should work, and everyone knows why it does: pulling the real basis out of the inspection history takes weeks nobody has, so the previous list stands in for it. The cost shows up twice, as equipment that did not need opening and as equipment that did and was not on the list. This paper describes how to build a turnaround inspection scope from the records, what API 510 and API 570 interval logic contributes, how corrosion rates and remaining life rank candidates, where risk-based inspection fits, and how damage mechanism history changes the answer.

Why is the turnaround inspection scope usually wrong?

A scope inherits three kinds of error from the way it is built.

• Inertia. Equipment stays on the list because it was on the last one. A vessel whose last two internals found nothing keeps getting opened.
• Omission. Equipment enters scope when someone remembers it. Piping circuits whose corrosion rate quietly shortened remaining life since the last event are missed because nobody looked at the rate.
• Unjustified additions. Items added late, on a hunch or a rumor, with no record behind them, and therefore no defense when the budget review cuts them.

The common cause is that the scope is built from a list rather than from the records. The records hold, for each piece of equipment, when it is next due, how fast it is corroding, how much life it has, how it ranks by risk, and what mechanisms are active. Scope should be a query over those facts, refreshed for each event, not a document edited from the last one.

That requires the records to be in a state where the query can be run: CMLs with their engineering fields populated, readings with dates, rates computed, and damage mechanisms recorded as data. The rest of this paper assumes you are working toward that state and shows what each part of the history contributes.

What do API 510 and API 570 say about who is due?

Both standards tie the inspection interval to remaining life. For scoping, the consequence is that the question "who is due" becomes "whose remaining life expires before the next opportunity."

API 510 (pressure vessels). The internal or on-stream inspection interval is limited to half the remaining life or ten years, whichever is less. When remaining life is short (under four years), the interval is limited further, to the full remaining life up to a two-year maximum. External inspection has its own interval. Where a risk-based inspection assessment is in place, API 510 allows the RBI result to set the interval instead.

API 570 (piping). Thickness measurement intervals are set by piping class (with Class 1 the most frequent) and by half the remaining life, whichever is shorter. External visual inspection has separate, class-based intervals. RBI is likewise permitted to set intervals.

For a turnaround, the window that matters is not "due before this event." It is "due before the next opportunity to do this work." If the next planned turnaround is five years out, every vessel whose internal falls due in the next five years is a candidate for this one, unless an on-stream inspection can substitute. That single reframing usually changes the scope more than any other step.

The interval calculation should be reproducible: the platform or spreadsheet should show the remaining life, the half-life cap, the ten-year (or class) cap, and which one governed. A due date with no visible derivation cannot be defended in the scope review.

How corrosion rates and remaining life rank the candidates

Remaining life is the engine of the interval, and it is computed per CML.

The inputs, all of which must be on the CML record:

• Current thickness (t), the latest reading.
• Minimum required thickness (tMin), computed by the design code appropriate to the equipment: ASME B31.3 for process piping, ASME Section VIII Division 1 for vessels, API 653 for atmospheric storage tanks. The record should carry the method used and the formula, not just the number.
• Corrosion rate, of which there are three:
  • the short-term rate, from the latest two readings;
  • the long-term rate, from the earliest to the latest reading;
  • the governing rate, the larger of the two (the maximum), because the conservative rate protects the equipment.

Remaining life is (t minus tMin) divided by the governing rate. The interval is half of that, subject to the caps above.

For scoping, two consequences follow.

First, one bad recent reading changes everything. A CML whose long-term rate is modest but whose short-term rate has spiked (a process change, a new corrodent, an inhibitor that stopped working) will have a short remaining life under the governing rate, and the equipment comes into scope regardless of what the last event's list said. Sort your CMLs by remaining life and read the top of the list.

Second, the inputs have to be right. A tMin computed against the wrong code or with the wrong allowable stress produces a wrong remaining life. Scoping is the point at which stale engineering fields on the CML get found, and it is better to find them in the scope review than on the scaffold.

Where does RBI fit in scoping?

Risk-based inspection under API 580 and API 581 ranks equipment by the probability of failure and the consequence, and it is allowed by API 510 and API 570 to set intervals in place of the half-life rule.

For turnaround scoping, RBI does two things:

• Extends. Equipment with a low probability of failure and a well-understood mechanism can carry a longer interval than the half-life rule gives, and can be removed from scope with a documented basis. This is where inertia gets challenged.
• Shortens. Equipment with high consequence, or with a mechanism whose probability is poorly characterized, moves up the list even where the thickness data alone would not put it there.

A semi-quantitative approach, scoring probability and consequence on a 1 to 5 scale and multiplying, is enough for scoping most units. What matters is that the assessment is current (reflects the present operating conditions), documented (the inputs and the rank are on the record) and approved (an engineer signed the assessment, not a spreadsheet macro). Where any of these is missing, the interval rules govern and the RBI rank is advice rather than basis.

RBI is also the tool for the budget conversation. When the scope must be cut, cutting from the bottom of a risk ranking is a defensible decision. Cutting from the bottom of a list ordered by tag number is not.

How damage mechanisms and CUI history change the scope

Thickness data catches general and localized thinning. It does not catch everything.

Damage mechanisms as records. Cracking (environmental, fatigue, hydrogen-related), embrittlement and other non-thinning mechanisms do not show up as a corrosion rate. If mechanisms are recorded as data against equipment, the scope query can pull in everything with an active cracking mechanism whose last examination is older than the program requires. If they live in a report PDF, they will be missed.

CUI history. Corrosion under insulation is the classic turnaround scoping problem. Suspect areas are flagged at a visual survey. Insulation removal is scheduled for "the next opportunity." The next opportunity is this turnaround, and only a structured CUI record will surface which circuits are carrying open suspect areas. A complete CUI record runs from visual, through suspect area and insulation removal, to NDE UT and advanced NDE, repairs and recoat, with an inspection and test plan and a frequency assessment at the end. Every open item in that chain is a scoping input.

Repair history. Equipment with prior repairs, temporary repairs in particular, comes into scope for verification whether or not the thickness data says so.

Fitness-for-service screening. Where a CML is already below tMin, a Level 1 screening under API 579-1 Part 4 (remaining strength factor) or an ASME B31G remaining-strength assessment on piping tells you whether the equipment can run to the turnaround or needs earlier action. These results are provisional until an engineer verifies them, and the scope should say so.

Turning the scope into work packs and a gate review

A scope is a list of equipment with reasons. A work pack is what an inspector can execute. The conversion is where most of the value is either preserved or lost.

Each work pack item should carry:

1. The reason. Interval due, rate driven, RBI rank, active mechanism, deferred, or repair verification. One or more; never blank.
2. The CMLs and their history. Prior readings, the governing rate, remaining life, and the drawing with points marked so the technician finds the same spot.
3. The methods. Visual internal, UT thickness, PAUT, RT, DFT verification, and the acceptance criteria for each.
4. The qualifications those methods require, so crew dispatch can be checked against the pack.
5. The expected finding. What the history predicts. A large deviation from expectation is the first sign the inputs were wrong.

The scope then goes to a gate review, recorded as a record: attendees, the scope basis presented, items challenged and the decision on each, and the outstanding actions. Items removed for budget should be removed with their risk rank visible, and the removal should be recorded as a deferral, so that the next event's scope query finds them first.

The scope you can defend is the one where every line points at a record. The scope you cannot defend is the one that came from the last event's list.

How intelliSPEC handles it

intelliSPEC's intelliINTEGRITY module holds the records this scoping method queries. Each CML point carries 64 fields, including nominal thickness, tMin and its method, design pressure, material, OD and schedule, allowable stress, joint efficiency, corrosion allowance, grid geometry, the latest thickness and three corrosion rates. tMin routes by equipment type to ASME B31.3, ASME VIII Division 1 or API 653 and returns the design code and formula with the number. The corrosion summary computes short-term, long-term and maximum rates with the maximum governing, remaining life as (t minus tMin) divided by rate, and the API 510 and API 570 interval capped at half the remaining life. RBI scoring follows API 580 and 581 semi-quantitatively (probability by consequence, rank 1 to 5).

Damage mechanisms (thinning, cracking) are first-class records, and the CUI wizard family covers visual, suspect area, insulation removal, NDE UT, advanced NDE, repairs, recoat, ITP and frequency assessment, so open suspect areas are queryable. API 579-1 Part 4 Level 1 screening (RSF as t over tMin) and ASME B31G remaining strength (original and modified, 1991 and 2012 editions) are available, with results provisional until an engineer verifies them and never auto-derated. API 579 Part 5 Level 1 is deliberately not implemented pending licensed-engineer verification.

The scope becomes work in intelliPROJECTS, where the event, activities, crew, daily progress, delay events, gate reviews and cost entries are structured records, and is executed in intelliINSPECT offline with drawing annotation and automatic CML point detection. Terms used here, including CML, tMin and governing rate, are defined in the glossary. The companion paper on turnaround readiness gates covers what happens after the scope is set.