GOLDILOCK & REELING FAILURES

STATE OF ART
D.ALLUE – June 24

1
TABLE OF CONTENTS
1. Reeling Cracks
I. Background (5LPP/GSPP)
II. Ballymore Goldilock
III. Comments
2. Route Cause for cracks during reeling
3. Benchmarks
I. FJC Design
II. ULTRA
III. NEMO Hybdrid FJC
4. Characterization of Goldilock FJC
5. Odysea June 24 improvement action plan
6. Summary - Outcomes
PLP Tubular Essentials - Coating 2
1 REELING CRACKS

3
I. BACKGROUND 5LPP/GSPP

► Failure modes are divided into 4 Types for classification :

• Type A: Crack on Parent Coating Next to Field Joint

Type B: Crack on Parent Coating Outside Field Joint Area
Type C: Crack Within Field Joint Chamfer Area
Type D: Crack Within Field Joint (body of field joint)

► Main one is Type A

• “Up to 5%of joints in difficult projects” as per Richard Ross , TFMC Principal Technology Expert | Pipeline Materials & Manufacture

Typical examples of GSPP crack during spooling (courtesy R.Ross (left) and [1] (right)

Title of the presentation 2 October, 2024 4

II. GOLDILOCK – BALLYMORE PROJECT 2024

► Ballymore statistics (courtesy I.Santos – Odysea – June 24 (CONFIDENTIAL))

► Odysea comments
« There were no Type B cracks - There was a single Type C crack & a single Type D crack - Type C from our experience
originates from the Chamfer interface (first disbondment of FJC at chamfer, then crack forms) - Type D is anomalous, and
we had not seen this before, during any of our testing, the single instance points to a nonstandard cause.”
Title of the presentation 2 October, 2024 5
III. COMMENTS
► Quality wise
• Goldilock FJC failures during reeling present same characteristics than state of art 5LPP/GSPP
» Reasonable to proceed by analogy in terms of route cause analysis (see here after)
• Type A cracks (on Parent Coating, next to Field Joint) are the vast majority
» And this as well similar to 5LPP/GSPP (as per TFMC information)
► Quantity wise
• 1.6 % of installed joints suffered from cracks during Ballymore spooling
» To be compared with « 5% for difficult 5LPP/GSPP project » as per TFMC information
» (nota : the failed joints were repaired, interrupting spooling operations but delivered. Meaning a repair procedure
was adopted by operators (as for standard 5LPP/GSPP).
► In summary
• Goldilock system presented similar failures than a 5LPP/GSPP during reeling of Ballymore projects (mainly Type A
cracks in parent coating close to FJC, and in similar quantities or no more than what is considered a difficult project).
• The failed joints were repaired and the line pipe installed.

► Questions (to be addressed to S7)

• Criticality of Ballymore spooling ? (Vessel, Temperature, Reeling Speed, …)
• Repair Procedure ?
• Any adaptation of reeling conditions ? (speed, …)
Title of the presentation 2 October, 2024 6
2 ROUTE CAUSE FOR CRACKS DURING REELING

7
STRAINS VS CRACKS

► Rupture of a material or an assembly,

occurs when the applied force exceeds the strength of that material/assembly.

► We can here after identify several origins of strains generated in the coating
assembly, and in particular at FJC area, during reeling.

Title of the presentation 2 October, 2024 8

REELING MECHANICS

► Reeling process imposes repeated high plastic strains on the pipe [2]. Leading to
pipe ovalization, possible buckling, collapse,
as well as rupture into the Thermal Insulation coating.

Reeling at Orkanger [2] Longitudinal stresses in the pipe during reeling [2]

Title of the presentation 2 October, 2024 9

REELING PARAMETERS

► Literature suggests some key parameters driving intensity of generated strains in

coating material [2,3]
• Reel Radius – Nominal Reeling Strain at Outer Surface of pipe
• Reeling Speed
• Reeling Back Tension Range
• Reeling Temperature

Title of the presentation 2 October, 2024 10

FIELD JOINT COATING (5LPP CASE)

► Field Joint Coating Operation generates additional locals strains [1]

• Field joint coating shrinkage as a result of field joint coating volume contribution (injected at
approx. 250°C)

Scheme of FJC shrinkage effects in coating system [1]

Injection Moulded PP (Rae internet site)

Title of the presentation 2 October, 2024 11

PEAK PLASTIC STRAINS (5LPP CASE)

► Literature suggest Peak Plastics Strains to be generated with the

following key characteristics or origin [3]
• Peak plastic strain in the coating occurs at the interface between
field joint and parent pipe coatings at the end of the overlaid layer
(Type A) and at the top of the parent coating chamfer (Type C);
Type C Type A
• Plastic strains in the coating are governed by the shrinkage of the
IMPP during quenching and during reeling by the influence of
geometrical stress risers such as coating geometry, and
dimensional and material mismatches;
• Peak coating strain is likely to be noticeably higher than the
nominal reeling strain in the steel;
• Coating temperature is the parameter that most influences peak
strain (compared to reeling speed and reeling back-tension);
• Slightly higher than ambient coating temperature reduces coating
peak strain; obviously, however,coating must be suitably cured
prior to reeling operations;
• Reduced reeling speed reduces peak strain;
[3] ABAQUS Local Model (Tension Only Model) –
• The impact of back tension on strain levels is marginal; however, Typical Strain Field after Spooling
applied back-tension must beappropriate and approved for reeling
operations
Title of the presentation 2 October, 2024 12
3 BENCHMARKS

13
► This part of analysis focuses on existing solutions which are promoted for their
beneficial impact on reel ability of 5LPP/GSPP
• Intention here is to see if we can link with Route Causes

► As per Richard Ross (TFMC), main mitigations to GSPP failures :

• Having a qualified parent coating / FJC coating system
(fixed products, applications, and key tests such as FJC compatibility test for parent coating and Full Scale Bend Test)
• Avoid when possible to reel at cold temperature
• Design of FJC to reduce stresses
• Use of alternative FJC or Insulation system (depends on project characteristics)
» Hybrid FJC Nemo 1.1 (instead of full IMPP)
» ULTRA
– « More tolerant to low T° than GSPP »

Title of the presentation 2 October, 2024 14

RICHARD ROSS / ASSESSMENT OF FJC SYSTEM

► We collect Tensile, Compression and Flexural Modulus over a range of temperatures which the design engineers request for FEA
design. When I mentioned modulus I was referring to a general increase in stiffness in the material regardless of the property
measured but it is fair to say that in the context of coating integrity Tensile Modulus is usually considered.
► When we are evaluating a material as a coating candidate then elongation at break at the temperature of interest is a key indicator of
the materials ability to withstand strain.
► Charpy impact is a good test for comparing different candidate materials

► For linepipe and FJ systems there may be an FEA model used by some to determine the reelability which will use a particular
material model that requires specific properties.
► I would say that there is no specific model that accurately predicts when and where a failure in a coating system will occur. Most
models are based on experience or known failure cases and help to predict when you are in a general zone of concern. What seems
to be lacking in the industry is an understanding of the failure mechanism of the material and therefore the relevant property that
needs to be measured.

► In practice what we tend to do is perform bending and straightening trials with the pipe (linepipe and FJ) conditioned to the likely
environmental temperature which is meant to simulate the spooling process. Occasionally we prepare a lead string of pipe which is
spooled onto the vessel but this is an extremely expensive test and is usually limited to completely new systems we have never
used before , such as GDLX.

Title of the presentation 2 October, 2024 15

I. DESIGN OF FJC

► Operators developped their « receipe » to optimize geometry / strains

distributions along the chamfer (and mitigate over issues linked 3LPP)

Past insulation “direct” chamfer example Modern staged chamfer examples

Title of the presentation 2 October, 2024 16

II. ULTRA

► Literature [4] suggests ULTRA beneficial behavior could be linked with relevant
accommodation of Mechanical Properties of applied FJC vs Parent Coating
• avoiding “Material Mismatch” reported in [3], which in other words is to avoid having
Higher Young's modulus (elastic modulus) in the field joint coating material when
compared with the elastic modulus of parent coating material [1]
• And this even more when temperature is 0°C (Efjc/Eparent coating <<0 ) ! With a
specific foam for FJC keeping low E at 0°C

Reported ULTRA applied material tensile properties [4]

Title of the presentation 2 October, 2024 17

III. NEMO HYBRID FJC

► Literature [1] suggests two main drivers at origin of beneficial usage of Hybrid NEMO as FJC for
5LPP/GSPP
• « Hour Glass » IMPP (Thermal barrier to Nemo PU) , first layer of FJC (lowering shrinkage effect)
• Filler Nemo PU with lower E (110 Mpa) than PP material (FJC less « stiff » than parent coating)

Title of the presentation 2 October, 2024 18

4 CHARACTERIZATION OF GOLDILOCK FJC

19
TENSILE PROPERTIES (ISO 527-1)

► Tensile probes sampled from the Parent Coating (R-Red), FJC (B-Black), and at
Chamfer (BR) – test @ 23°C

Yield Strength Rupture Rupture

Sample Re (Mpa) E (Mpa) Rm (Mpa) R (Mpa) %
B (FJC) 21,4 1066 23,8 23,8 107
BR (Chamfer) 19,2 1065 19,2 19,2 5
R (Parent coating) 20,8 1091 23,2 23,2 85
PP Solid [2] 26 1700 95
PP Foam [2] 15 800 72

Title of the presentation 2 October, 2024 20

 TENSILE PROPERTIES (ISO 527-1)
► Comments
• Tensile properties of Parent Coating and FJC are similar
» Tensile constraint approx. 21 Mpa (similar than Odysea/VLR 2023 pipes samples)
, Rupture at 23 Mpa after approx 100% elongation (plastic deformation)
» Tensile moduli at 1000-1100 Mpa
» Need test curves for more comments
• Comparison with literature [2] suggest same range of tensile properties than 5LPP
» Need to confirm and investigate elasto plastic behavior (ex. : no striction in tensile test)

Goldi

• At Chamfer exibits Fragile rupture charateristics (no elongation,)

» Rm=R= 19 MPa (& < 10% deformation at rupture)
» Seems all ruptures in FJC Material (few adhesive at chamfer)
– As E similar for matching materials, hypothesis rupture initation at adhesion weakness
than propagates in material
Title of the presentation 2 October, 2024 21
TENSILE PROPERTIES (ISO 527-1)

► Benchmarks
• At positive temperature, literature [2] shows plastic deformation of Foam PP before
rupture (filaments formed from foam holes),
turning to fragile rupture at -30°C (no plastic deformation)
• Godilock exhibits fragile rupture when happens at chamfer interface
» To be investigated and compared with cracks in raw materials (FJC or PC)

Title of the presentation 2 October, 2024 22

PULL OFF (UNI EN 12814-2:2021)
► Bending 2° of samples did not lead to any crack
Rupture
► Pull off testings from 2° bent than straightened probes Samples Mpa (av.)
(some aged 120°C 1 month) Reference 16,8
• Lower strength than measured on normalized tensile probes (test effect ?) Bent 2° & Staightened 15,9
• Several testings artefacts to be discussed with Mated B&S + Ageing 1 month 120°C 16,4

• rupture generally occuring in Parent Coating, contradictory with

tensile tests possibly effect of test conditions

Pull off probes (right = References, center = B&S, right = B&S+Ageing

Title of the presentation 2 October, 2024 23
PULL OFF ISO 4624

► Pull off of sliced specimen along FJC

Elongation at
• 10 to 14 MPa measured in full FJC or PC areas Sample Area MPa Rupture
rupture (mm)
1 black 9,88 FBE/Goldilock 2,2
» In line with qualification pull off (dollies) 2 black 13,6 FBE/Goldilock 3,8
3 black 14 FBE/Goldilock 3
but Adhesive failures at lowest strength than material properties 4 overlap center 2,03 FBE/FJ Goldilock 0,2
5 overlap center 2,87 FBE/FJ Goldilock 0,2
• 2 to 5 MPa in overlap chamfer areas, 6 overlap center 3,2 FBE/FJ Goldilock 0,3
7 overlap 3,24 FBE/FJ Goldilock 0,4
» Unexpected Adhesive failures at FBE interface (and very low strength) 8 overlap 5,42 FBE/FJ Goldilock 0,8
9 overlap 13,3 FJ Goldi / Goldilock 3,3
» To be further explained 10 overlap 13,4 FJ Goldi / Goldilock 3,1

Title of the presentation 2 October, 2024 24

MAIN OUTCOMES

► Tensile properties of GDLX seems in same range than 5LPP reference

• Reassuring absolut values as for expected application (E 1000 Mpa, Re >20 MPa…)
• different visco plastic behavior to be investigated
► Same Tensile properties for Parent Coating and FJC
• No material mismatch at FJC area
► Questioning interfaces performances through Pull offs tests
• Several adhesive failures at lower strength than raw material intrinsic properties
• To be investigated (effect of the test ?)

► (several tests results to be obtained (tests curves ) and debated with Mated for
refined understanding)

Title of the presentation 2 October, 2024 25

5 ODYSEA JUNE 24 IMPROVEMENT ACTION PLAN

26
 ODYSEA INTERVIEW JUNE 24 - BALLYMORE FEEDBACK
► Type A is inherent to all field joint systems of similar design due to the higher modulus of the FJ not only as a Agree
material(in the case of SPP FJ and GSPP LP) but also as a structure/mechanical design as the FJ is often thicker than
the parent coating overall and this overlap area interface is an area of accumulation of stress.
Currently mitigated Mated
► In the case of GDLX we see that Type C is mitigated by a very strong bond of the chamfer interface and high results (need extra tests)
integrity of the GDLX when compared to GSPP.
► As for the Type A, the issue of dissimilar moduli is mitigated with both the LP and FJ being made of GDLX. Agree
► However the geometry in what is often considered the “industry standard field joint design”, as some
specifications require overlap in order to create a greater bond between what are typically two SPP structures (FJ
and Jacket) as the SPP/GSPP bond does not tolerate strain well. We consider this design an artifact of MLPP
systems, and suboptimal/unnecessary for GDLX, and a major contributing factor to the observed cracks. Tbd with Odysea

► Reeling customers have told us that changing the FJ geometry would be welcome. These efforts are already in Agree
process and consist of 3 main activities:
1. We have contracted Dr. Sylvain Popineau as a consultant to assist us in FEA modeling of GDLX field joints
aimed at optimizing the geometry and providing a parameterized tool for predicting stress buildup during
reeling
2. We are validating the FEA via rapid iterative testing on our Medium Scale Bend Rig using DIC to produce
quantitative results that give feedback to the model. We are also using MSBR to quickly screen the basic
integrity of design concepts
3. We are producing full-scale versions of the optimized geometries to perform simulated bend testing and
validate the resulting designs at scale.
Title of the presentation 2 October, 2024 27
Title of the presentation 2 October, 2024 28
COMMENTS ON ODYSEA IMPROVEMENT PLAN

► Action plan focused on GEOMETRY design of FJC

• One of key mitigation arising from literature and customer interview
► Still some potential weak points to clarify and that be detrimental ?
• Material,
» Clarify Tensile Modulus FJC/Parent coating
» Effect of Temperature (what happens at 0°C reeling ?)
► Other improvement ideas ? possibilities ?
• Use Nemo Hybrid on GDLX projects <150°C ?
» And a « Nemo Hybrid with GDLX » for > 150°C ?
• Modify GDLX properties /Material ? For a lower E ?

Title of the presentation 2 October, 2024 29

6 SUMMARY & OUTCOMES

30
► Reeling cracks observed on 5LPP systems are mostly located at the edge of FJC lipper
and propagating within the parent coating (for maximum 5% of installed joints)
► Goldilock presented exact similar behavior during Ballymore project 2023
► This can reasonably be understood by similar tensile mechanical properties of the two
systems and similar FJC designs
► To be noted that Goldilock presents fragile behavior at interfaces with FBE sublayer or with
FJC, which differs from 5LPP systems. This could be seen as a negative point by
customers even if it is not linked with reeling failure mechanism. For other laying methods,
adhesion strengths of Goldilock seems sufficiently high to withstand specified tests for
intended usage (such as push pull test).
► For reeling, it is thus reasnable to think that the same mitigation promoted for 5LPP
systems could apply for Goldilock, starting with design of the FJC, and the material used
for FJC (use of a less stiff material than parent coating). It was confirmed by Odysea that
these are the current solutions explored to improve reeling performance of the Goldilock
system.
Title of the presentation 2 October, 2024 31
REFERENCES

[1] « NEMO Hybrid System - A Novel Field Joint Solution », Normando Cunha, Shawcor Pipeline Performance, Offshore
Technology Conference Brasil, 24–26 October 2017.
[2] «Numerical Modeling of Low Temperature Reeling of Pipes », Frode Grytten (SINTEF), Virgile Delhaye (SIMLab NTNU),
Kjell Olafsen (Bredero Shaw), Trond Sjchelderup (Bredero Shaw), Proceedings of the Twenty-fourth (2014) International
Ocean and Polar Engineering Conference, Busan, Korea, June 15-20, 2014
[3] “Effect of Temperature, Reeling Speed and Pipe Tension on the Performanceof Field Joint Coating During Reeling of
Offshore Pipelines” , Rajaram Dhole, Subsea 7; Ismael Ripoll, Xodus Group; Sabesan Rajaratnam and Celine Jablonski,
Subsea 7, Offshore Technology Conference held in Houston, TX, USA, 16 - 19 August 2021.
[4] “Novel Thermal Insulation Systems for Deepwater and Arctic Pipelines Based on Styrenic Alloys” , Shiwei William Guan,
Tony Bacon, Tor Fredrik Bredeli, and Marianne Asbøll (Bredero Shaw),Adam Jackson (RAE Energy), Proceedings of the
Twenty-fourth (2014) International Ocean and Polar Engineering Conference Busan, Korea, June 15-20, 2014
[5] Modelisation numérique du revetement d’Isolation thermique de canalisations pétrolières sous-marines, Theophile
Hourdou (Thèse présentée le 14/12/23), ENSTA Bretagne – Université de Lille – SAIPEM (Sebastien Blassiau) - PSL Mines
de Paris
[6] Goldilock FJC characterizations, Mated Report , June 2024
[7] GDLX Qualification Program Summary, May 25 2021

Title of the presentation 2 October, 2024 32