Piping Integrity Management: What API 570 and ASME B31.3 Actually Demand

Piping is the most extensive pressure-retaining asset class on any industrial plant — and the one most commonly managed as a bulk commodity rather than an integrity system. Vessels get individual asset numbers and inspection histories. Piping gets a paint programme and a hope that nothing is corroding behind the insulation.

API 570 exists because piping integrity failures are not random. They are predictable, mechanism-driven, and location-specific. A program built on that principle — circuit-based, risk-ranked, with inspection locations chosen against known degradation mechanisms — will find wall loss before it becomes a release. A program built on calendar intervals and visual inspections of accessible painted surfaces will find it after.

This is a forensic account of what an execution-grade piping integrity management program requires — the governing standards, the program elements, and the six failure modes that most programs never explicitly address.

The Governing Standard Set

The gap most commonly encountered in field reviews is the treatment of API 570 as the complete standard set. It is not. A facility with an API 570 inspection programme but no CUI programme, no deadleg register, and no injection point inspection protocol is running a partial integrity programme — one that will find general wall loss on accessible piping while missing the locations where the highest-consequence degradation is occurring.

Circuit-Based Inspection — The Foundation

API 570 is a circuit-based standard. Piping is divided into inspection circuits — segments of pipe with similar process conditions, materials, and degradation mechanisms. Inspection locations within each circuit are selected against the known or predicted degradation mechanisms for that service, not distributed uniformly across the system.

A programme that applies uniform inspection point spacing across all piping regardless of service, fluid, temperature, or insulation status is not implementing API 570. It is implementing a measurement schedule that happens to use API 570 thickness measurement methods. The distinction determines whether the programme finds wall loss where it is occurring — or measures wall thickness where it is convenient to measure it.

What an Execution-Grade Programme Requires

1. Piping Circuit Register Every piping system divided into numbered inspection circuits. Each circuit documented with fluid service, operating temperature and pressure, material specification, insulation status, design standard (B31.1 or B31.3), minimum required thickness, and applicable degradation mechanisms. A circuit not in the register is not in the programme.
2. Degradation Mechanism Assessment per Circuit Each circuit assessed for all applicable degradation mechanisms: general corrosion, localised corrosion, erosion-corrosion, flow-accelerated corrosion (FAC), stress corrosion cracking (SCC), CUI, deadleg corrosion, and injection point attack. The assessment drives inspection location selection, NDE method selection, and inspection frequency. Without this assessment, inspection locations are chosen by accessibility — not by risk.
3. CUI Programme — API RP 583 Compliant Every insulated circuit assessed for CUI susceptibility based on operating temperature range, insulation type, coating condition, and service history. CUI susceptibility windows are specific: carbon steel is most susceptible between -4°C and 175°C; austenitic stainless steel is susceptible to chloride stress corrosion cracking above 60°C under wet insulation. Circuits in the susceptibility range require a dedicated inspection programme — window removal, visual inspection, and UT thickness measurement at selected locations — on a documented, risk-justified interval.
4. Deadleg Register and Inspection Protocol Every deadleg — stagnant piping branch, capped connection, bypass loop, or out-of-service segment that remains connected to the active system — identified, registered, and assigned an inspection protocol. Deadlegs accumulate product, develop stagnant water layers, and corrode at rates that can be multiples of the active piping corrosion rate. They are systematically excluded from standard circuit-based inspections because they carry no flow. A deadleg register forces them into the programme.
5. Injection Point Inspection Protocol Every chemical injection point, inhibitor injection connection, and process mixing point identified and assigned an enhanced inspection protocol. API 570 requires inspection of the pipe wall for a minimum of 12 inches upstream and 3 pipe diameters downstream of each injection point. Injection points experience localised velocity-driven corrosion, impingement attack, and chemical concentration effects that are not representative of the general circuit corrosion rate. They fail faster than the surrounding piping and require more frequent inspection.
6. Fitness-for-Service Determination Before Return to Service Every piping segment with measured wall loss assessed against API 570 / ASME B31.3 or B31.1 retirement thickness criteria before being returned to service following an inspection that identified degradation. The inspection identified the condition. The FFS determination establishes whether continued operation is acceptable, at what reduced operating pressure if applicable, and what the required reinspection interval is. Returning piping to service without this step means the inspection was performed but the decision it was designed to support was never made.

The Six Failure Modes RISL Encounters in the Field

Failure Mode 01 — CUI Not Recognised as a Separate Programme Requirement

Insulated piping included in the general circuit inspection programme with no CUI-specific inspection protocol. External surfaces visually inspected where insulation is intact — which means no inspection at all, because CUI occurs beneath insulation that shows no external indication of the corrosion underneath. API RP 583 is explicit: CUI cannot be managed by visual inspection of insulation cladding. It requires scheduled window removal at susceptibility locations. Programmes that do not require window removal are not managing CUI — they are assuming it is not present.

Failure Mode 02 — Deadlegs Not Registered

No deadleg register exists. Stagnant branches, capped connections, and out-of-service piping segments connected to active systems are excluded from the inspection programme by default — they appear in no circuit, they carry no flow measurement, and they generate no work orders. In long-service plants, the number of unregistered deadlegs frequently exceeds the number of registered ones. Each one is corroding at an unknown rate with no inspection interval assigned.

Failure Mode 03 — Injection Points Treated as Standard Circuit Locations

Chemical injection points included in the general circuit inspection grid with the same inspection frequency and measurement spacing as the surrounding piping. API 570 requires enhanced inspection at injection points specifically because the degradation mechanism and rate at injection locations are not representative of the general circuit. A standard UT reading taken 600mm from an injection point does not characterise the wall condition at the injection point. The most aggressive degradation is occurring at a location that is not being measured.

Failure Mode 04 — Corrosion Rates Not Calculated

Thickness measurements recorded at each inspection without being compared against previous readings to calculate an actual corrosion rate. The programme generates inspection records. It does not generate integrity data. Without a calculated corrosion rate, remaining life cannot be determined, the next inspection interval cannot be justified, and the fitness-for-service assessment cannot be completed. The measurements exist in a file. They are not being used.

Failure Mode 05 — B31.1 and B31.3 Applied Interchangeably

Power piping (steam, feedwater, blowdown) inspected and evaluated against B31.3 process piping criteria, or process piping evaluated against B31.1. The retirement thickness calculations, allowable stresses, and fitness-for-service criteria differ between the two codes. Applying the wrong code to a system produces a fitness-for-service determination that is not valid for that system — either non-conservatively (accepting wall loss that the correct code would require to be repaired) or conservatively (requiring repair on piping that meets the applicable standard). Both outcomes represent a programme failure.

Failure Mode 06 — Small-Bore Piping Excluded

Piping below a defined nominal bore threshold — typically 50mm or 25mm — excluded from the inspection programme on the basis that small-bore piping failures are low consequence. API 570 does not support this exclusion as a general rule. Small-bore connections to high-energy systems, small-bore instrument connections on process lines, and small-bore drain and vent connections on pressure-containing equipment represent disproportionate release risk relative to their size. High-consequence small-bore connections must be identified and included in the programme.

The Caribbean and Guyana Context

Piping integrity management in Trinidad and Tobago, Guyana, and Suriname requires CUI susceptibility assessments that account for tropical humidity, salt-laden coastal atmospheres, and the wide diurnal temperature cycling in exposed piping systems. Carbon steel piping in coastal equatorial service can experience CUI initiation and progression at rates that make Canadian corrosion rate assumptions non-conservative by a significant margin. RISL engagements in these markets apply service-environment-adjusted corrosion rate assumptions — not Canadian or North American temperate defaults.

---