Case Study: Fixed steel jacket platform leg, 42-inch tubular X65, K-node tubular brace joint — Offshore oil & gas (fixed jacket platforms)

Asset & context

Offshore oil & gas (fixed jacket platforms) — US Gulf of Mexico (200 ft water depth). Fixed steel jacket platform leg, 42-inch tubular X65, K-node tubular brace joint. Crew of 7 for 18 days.

The challenge

A GoM Class 1 fixed jacket platform operating in 200 ft water depth, in service since 1991, was on a 5-year underwater inspection interval under API RP 2SIM for in-service inspection of fixed offshore structures.[1] The K-node joint at elevation -68 ft (where the 24-inch K-brace meets the 42-inch leg) had been a high-fatigue-loading concern from the original design FEA, with monitored stress concentration factor (SCF) of 5.8 — typical for a K-node geometry in a North Atlantic-pattern wave climate. A CVI by ROV during the 2026 inspection found a 95 mm linear indication at the K-node weld toe on the brace side. The operator had to: (a) characterize the indication (crack vs paint flaw vs marine growth artifact), (b) size any confirmed defect to drive fitness-for-service evaluation, (c) issue a class society notification per ABS Rules for Building and Classing Offshore Installations within the 24-hour window if a crack was confirmed.[2]

Approach

Method selection: ACFM and PAUT in tandem for underwater weld characterization

Underwater NDT for offshore jacket structures has a specific tool kit defined in API RP 2X and AWS D3.6M.[3][4] Close visual inspection (CVI) is the first screen. Magnetic particle inspection (MT, the underwater variant uses fluorescent wet-particle technique) is the surface-NDT workhorse for ferromagnetic structural welds but has practical limitations at depth (visibility, current, light from the diver helmet). Alternating Current Field Measurement (ACFM) is the modern alternative — diver-deployed handheld probe, no magnetic ink required, gives both crack detection and length sizing in a single pass, and works through paint up to 6 mm thick. The operator's underwater inspection procedure specified ACFM as the primary follow-up to CVI per AWS D3.6M §6.4.

PAUT was the depth-sizing tool. A diver-deployed phased array probe with a 5 MHz, 32-element linear array configured for shear-wave inspection of the K-node weld geometry can resolve through-wall extent of a surface-breaking crack to roughly ±1 mm accuracy. The PAUT scan plan followed ASME V Article 4 Appendix IV for weld inspection, adapted for the K-node geometry and the curved tubular surface.[5] Sizing a fatigue crack at a K-node requires both crack length (from ACFM) and crack depth (from PAUT) to feed into the fracture-mechanics fitness-for-service evaluation.

Saturation diving was the deployment platform. At -68 ft, surface-supplied air diving is possible but constrained on bottom time; saturation diving from a dive support vessel (DSV) gives 5-6 hour useful working windows and was selected for the multi-method inspection of the K-node. Two PAUT-qualified saturation divers were rotated through the inspection, with surface-side data analysis by a Level III PAUT inspector in real-time.

Procedure: ACFM screening, PAUT sizing, full K-node coverage

The first dive completed CVI on the K-node weld at 0.5 m visual range, photographing the indication and confirming it was a discrete crack-like feature rather than marine-growth artifact or paint disbond. Marine growth was removed from a 600 mm × 400 mm window around the indication using a low-pressure water jet (12 MPa) — high-pressure cleaning was specifically excluded by the procedure to avoid masking or worsening any crack indication. Surface prep brought the underlying weld and HAZ to bright metal.

ACFM scanning followed with a Tritex Eddyfi probe across the full K-node weld toe on both the brace side and the leg side. The brace-side scan returned a 92 mm crack signature with the characteristic ACFM Bx and Bz field-component plots consistent with a surface-breaking fatigue crack. The leg-side weld scanned clean. ACFM length sizing gave 92 mm ± 3 mm — confirmed in a second-pass scan with the probe rotated 90° to verify the orientation.

PAUT depth sizing followed on a third dive. The 32-element linear array was mounted in a flexible curved-surface holder that conformed to the 42-inch leg tubular geometry. The scan was performed in linear electronic scan mode at 45° refracted angle on the brace side, capturing the crack tip diffraction. The PAUT A-scan showed a clear tip signal at 8.2 mm depth, with the corner-trap signal at the crack-mouth surface. The crack was confirmed as a fatigue-type surface-breaking crack, 92 mm long, 8.2 mm deep into the 16 mm wall — roughly 51% through-wall extent.

Findings: fatigue crack at high-SCF K-node, no other K-nodes affected

Class society notification went out within 18 hours of PAUT confirmation. The fracture-mechanics fitness-for-service evaluation under BS 7910 §7 used the measured crack dimensions, the operating stress range, the wave-loading spectrum from the platform's metocean database, and the base-metal fracture toughness (Charpy V-notch data from the original mill certificates).[6] The critical crack size at end-of-life conditions returned 14 mm depth × 140 mm length — the existing 8.2 mm × 92 mm crack was inside the critical envelope but with reduced margin against a major storm-loading event.

The asset team triggered an expanded-scope inspection of every K-node on the jacket. Saturation divers ran ACFM on 24 additional K-nodes across the four jacket legs over the next 6 days. Twenty-three nodes returned clean ACFM. One additional node (different leg, similar elevation) returned a 45 mm × 4 mm crack — well inside the BS 7910 critical envelope and dispositioned as a monitored defect with re-inspection at 12 months. The fatigue-crack distribution matched the original FEA prediction: highest-SCF K-nodes at the mid-depth elevation showed crack initiation first; lower-SCF nodes were clean.

A separate ROV-deployed VT inspection of the jacket anode condition (a routine 5-year scope) confirmed 38% remaining anode material, consistent with the original cathodic-protection design service life. Anode replacement was added to the next 5-year inspection cycle rather than requiring immediate intervention.

Disposition: class-approved underwater grinding-and-weld repair

The 92 mm × 8.2 mm crack was repaired by underwater hyperbaric weld repair per AWS D3.6M Class A.[4] The procedure: (1) grind out the crack to verified clean metal, confirmed by MT after each grinding pass; (2) re-profile the weld toe to a smooth contour with a maximum 1:4 weld-toe-to-base-metal transition to reduce SCF; (3) deposit a new weld using diver-deployed shielded metal arc welding with low-hydrogen E7018 electrodes inside a habitat with controlled atmosphere; (4) post-weld ACFM and PAUT to verify defect-free repair. The repair habitat was sized at 1.8 m × 2.4 m × 1.5 m, with surface-supplied breathing gas, and the welder-diver was certified to AWS D3.6M Class A welding.

Repair execution took 9 days, including 2 days of habitat installation and removal. The smaller 45 mm crack on the second K-node was monitored rather than repaired, on the engineering data — repair cost on a second node would have extended the campaign by another 7 days and the BS 7910 evaluation confirmed adequate margin for 12-month monitoring. The class society accepted both the repair and the monitored-defect disposition.

The avoided cost case was substantial. The next-step alternative if underwater repair had been judged infeasible was a full leg-strengthening project — adding grouted-pile reinforcement or installed plate cover-plates over the K-node region. Operator engineering estimates for similar past projects on GoM jackets put the cost at $28-40M with a 4-6 month installation window. The actual underwater repair cost: $4.8M including DSV mobilization, saturation diving, NDT, and the habitat repair scope.

Methods deployed

Defects found

Outcome

Class society notification filed within 18 hours of PAUT confirmation. BS 7910 fracture-mechanics FFS confirmed sub-critical for 12-month interval. Underwater hyperbaric weld repair completed per AWS D3.6M Class A. Second monitored crack scheduled for 12-month re-inspection. Platform returned to service.

Cost avoidance

$28-40M against a leg-strengthening project (grouted-pile reinforcement or cover-plate installation). The underwater hyperbaric repair at $4.8M was 85-88% cheaper than the leg-strengthening alternative and completed in 9 days vs 4-6 months.

Certifications required

Crew qualifications: