Inspection of Link 2 of a Mooring Chain · 20663/7/11 TWI Ltd Inspection of Link 2 of a Mooring Chain J. R Rudlin TWI Ltd Background BP wished to confirm the integrity of a mooring

20663/7/11 TWI Ltd

Inspection of Link 2 of a Mooring Chain J. R Rudlin

TWI Ltd

Background

BP wished to confirm the integrity of a mooring chain in the chain stopper location which had been in service a number of years, and required an inspection technique that would give quantitative information. It was agreed that in order to be conservative the detectable flaw size with 90% probability of detection should be 20mm through wall and 100mm long. The restrictions for the inspection were that it had to be carried out in 15m depth of water inside the turret structure, with no demanding tasks on the diver such as probe scanning because of the difficult water conditions. The total distance from the inspection site to the location of the NDE equipment and inspector was around 100 metres. The cables pass through a hazardous area so all inspection cabling and equipment needed to be doubly insulated and protected against sparking of any type. It was expected that any planar flaws would have an orientation of 10° and 70° to the minor axis or variations from this, so this had to be taken into account in the design of the inspection equipment. The surface for the inspection was possibly very corroded, and the access conditions for the inspection device were constrained as sometimes Link 1 is tilted with respect to Link 2 which reduces the access. It was required to produce the inspection technique for trials in the space of 10 weeks. Concept of Inspection Tool

Because of the access and scanning restrictions, surface inspection methods such as eddy current or ACFM could not be used. It was proposed, therefore, to use an array of eight ultrasonic probes placed on the crown of the link, to carry out the inspection. The detection principle used both direct and corner echoes. This was modelled using CIVA software, (Figure 1) and tried practically, with manual deployment of the probes. This enabled probe positions to be chosen. All probes were 2MHz 45° with a 10mm crystal and were supplied with wedges, so that, the water was used as a couplant. A tool to hold the probes on to the surface as shown in Figure 2 was designed and manufactured by Phoenix Inspection Systems. A switch box was manufactured, to switch between probes and a testing procedure was developed. An EPOCH instrument was used. POD study A set of test samples, with slots cut at a range of angles was manufactured (Figures 3 and 4). The performance was verified, by means of inspections on samples with slots introduced. The range of slotted test samples was increased so that, a statistical probability of detection could be calculated, using the response vs size method [1], and the desired sensitivity was confirmed. The POD curves for the two flaw orientations are shown in Figure 5 and 6. Diving Trials In order to familiarise the diving team with the equipment and to test it in more realistic conditions, the equipment was deployed on a model stopper in a diving tank (Figure 7), with the NDT operators working with the divers and practising the procedure. This exercise resulted in some minor changes to the procedure and the equipment. Offshore Deployment The equipment was deployed by Plant Integrity and the inspection was successfully carried out. Large indications of anomalies in two links and small indications in two others were found. It was decided that more information about the size of the anomalies was needed. Since their likely orientation was known, and the redeployment was required within a month, the only option for this was a modification of the existing tool.

20663/7/11 TWI Ltd

Tool used for sizing The required concept was to use each probe, to detect the presence of a crack across the circumference, (in a similar way to the zonal inspections used for pipeline welds). Figure 8 shows the principle of this technique. Locating the probe positions required modelling, to establish new probe positions and adjustment of the system (Figure 9). Tests were carried out to confirm that the system could assess the flaw size, using samples that had been manufactured with large sawcuts, for pull tests. Second Offshore Inspection A second inspection was then carried out that gave a better indication of the size of the anomalies, and provided confidence to carry out the lifting operation to change the mooring chains. Comparison of UT with cracks in chains After the chains had been removed, the opportunity was then taken to inspect the appropriate links with magnetic particle inspection and other surface methods. Destructive testing to obtain full information on the flaw size was also undertaken. The results obtained corresponded to the UT indications, except where a ridge caused by wear of the chain had resulted in a false call of a small crack. All the other chains, where no UT was found, were checked with MPI and no further indications noted. Finally the indication amplitude from the slots and indications found in the chain were compared (Figure 10). The data is scattered (as might be expected from a wide range of orientations, but the amplitude slope was almost identical, suggesting that the slots were a reasonable approximation to the flaws found. Conclusions

1 An ultrasonic test tool was developed for inspecting each Link 2 of the mooring chains in situ, for anomalies such as fatigue cracks. This was developed and tested and refined through onshore trials in a diving tank. The tool performed well on site.

2 The tool was shown to have a probability of detection for slots representing both orientations of the expected cracks of 90% POD for a crack 20mm deep.

3 Large amplitude Indications were obtained on two chain links, 4 Modification to the equipment was carried out, using a zonal method, to enable better

interpretation of the size of the anomalies. 5 The tool, coupled with a failure analysis, enabled confidence in the operation for safe removal of

the chain. 6 The onshore inspection after the chains had been removed showed good correspondence

between the ultrasonic indications and flaws in the chains.

Reference [1] Schneider C R A and Rudlin J R, 2004: ‘Review of statistical methods used in quantifying NDT reliability’.Insight 46 (2) February Acknowledgements This was a major inspection and could not have been achieved without significant team effort led by BP (the contribution of Neil Stagg, Alwyn Mcleary and Craig Donaldson, amongst others to facilitate the inspection is particularly acknowledged). The following also contributed significantly to the technical work and support of this project. TWI: Channa Nageswaran, Capucine Carpentier, Dorothee Panggabean, Kim Hayward, Philippe Bastid, Joanna Nicholas, Nigel Lee Plant Integrity: Neil Dodsworth, Graham Edwards, Angela Norrie, Sean Fewell Technip: Peter Dickson, Cameron George and the diving team Phoenix ISL: Karl Quirk, Tony Price, Chris Gregory WI would like to thank BP for permission to publish this work.

20663/7/11 TWI Ltd

Figure 1 Modelling of Ultrasound paths in chain

Figure 2 Ultrasonic Tool

Side View Top View

20663/7/11 TWI Ltd

Figure 3 Range of slot positions

Figure 4 Test Samples

20663/7/11 TWI Ltd

Figure 5 POD Curve for slots oriented at 10 degrees top minor axis

Figure 6 POD Curve for slots oriented at 70 degrees to minor axis

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 5 10 15 20 25 30 35

Slot Depth (mm)

Prob

abili

ty o

f Det

ectio

n (m

m)

PODLower bound with 95% confidence

20663/7/11 TWI Ltd

Figure 7 Training of divers on a simulated chain situation

Figure 8 Principle of the Sizing Technique

Crack Profile 1

Crack Profile 2

Crack Profile 3

Crack Profile 4

A

B

C

D

E

Probe beam positions in plane of crack

20663/7/11 TWI Ltd

Figure 9 Modelling of Sizing method

Figure 10 Comparison of On-site Indications and slots