2

TABLE OF CONTENTS

1 INTRODUCTION ................................................................................................................. 3
2 OBJECTIVE .......................................................................................................................... 3
3 MATERIALS TECHNOLOGY FOR THE OIL AND GAS INDUSTRY .......................4
3.1 Future development.......................................................................................................... 4
3.2 Concluding remarks .........................................................................................................5
3.3 Recent Materials Technology R&D initiatives ................................................................ 6
4 COMMON DENOMINATORS FOR INCREASED ROBUSTNESS ..............................7
4.1 Management of oil and gas development projects ........................................................... 7
4.2 Design ........................................................................................................................... 8
4.3 Knowledge gaps ............................................................................................................... 9
5 REFERENCES: ................................................................................................................... 10
 3

1 INTRODUCTION
Over the last three years the Norwegian Petroleum Directorate (NPD) has expressed some
concern about the situation with respect to material selection and safety, triggered by recent
experiences in the North Sea. During the period 1996 to 2002, 220 gas leakages (>0.1kg/s) have
been reported on the Norwegian shelf /1/. Of the 220 gas leakages, 5 are connected to different
types of failures on subsea pipelines, all involving through pipe wall fractures. The consequences
of such failures are significant. The total cost (including the loss of income) of the pipeline
failures is estimated to some 5-10·109 NOK. The impact on environment and human health and
safety has not been significant in these cases. Most of the 220 incidents are smaller topside gas
leakages. Luckily none have set on fire.

The final product is intended to be a methodology for robust material selection in the oil & gas
industry. A pre-project was initiated year 2002 and SINTEF Materials Technology was engaged
by NPD for examining relevant internal SINTEF reports, NPD reports and other easily available
information. Possible applications of the methodology are indicated in the report of the pre-project

• Drilling equipment
/2/:

• Well completion
• Structural materials
• Subsea, topside and onshore production and pipeline systems
• Flexible risers
• Chains and mooring lines for floating units

Subsequent to the finish of the pre-project report, a seminar with NPD, the Norwegian University
of Technical and Natural science (NTNU) and SINTEF was held in Trondheim in January 2003 to
achieve mutual understanding of the objectives of the main project. During this seminar, it was
decided that SINTEF should submit a new, more detailed plan for the main project.

For 2003 a focussed study on super martensitic stainless steels (S13Cr) pipeline material and

• the relative large number of unexpected incidents

flexible risers was requested. These cases were selected based on:

• significant concequences regarding production down time

• complexity of degradation mechanisms
• large differences on construction of pipeline and flexible riser

In this report a summary of the common denominators are presented. For more details, it is
recommended to take a closer look on the part reports produced on flexible risers /3/ and S13Cr
steels as pipeline material /4/.

2 OBJECTIVE
The objective of the present project is to summarize trends in material selection procedures in
larger oil & gas companies. The work is based on the use of existing knowledge in SINTEF,
Marintek and NTNU and their industry network. The work has focussed on a combination of
industry experience and research experience. The outcome of this project will be used as input to
the RMS methodology.
 4

3 MATERIALS TECHNOLOGY FOR THE OIL AND GAS INDUSTRY

Since the first findings of petroleum on the Norwegian Continental Shelf (NCS) back in the
1960s, the oil & gas nation Norway can look back on an impressing technological development
which is well recognized internationally. During the last 30 years the oil and gas industry acting in
the NCS, have positioned themselves as pioneers in several technological fields. The development
of OLGA 1D multiphase flow simulator, a joint SINTEF/IFE product, is one excellent example of
how Norwegian expertise made vital contributions to oil and gas transportation technology.
Development and safe use of new and more traditional materials is another example. Safe and
cost-effective use of materials gives the premises for realization of many other technology-leaps.
As an example, deep water development with associated sub-sea intervention can be mentioned.
The aggravating focus on increased safety and reliability stimulates research on associated
subjects. The safety aspect in the offshore industry acts as a basis for increased profitability,
safety and sustainability for future exploitation of our oil and gas resources. Also the judgment of
possible environmental impacts is a crucial aspect, especially in the debate of expanding the
offshore activities in the northern regions (Lofoten, Barents-sea).

Because of our oil and gas industry developed in hostile environments, Norway has paved the way
for use of steels (and other materials) under very demanding conditions. Development of materials
and fabrication technologies has been followed by development of standards and guidelines for
safe implementation. This situation is well recognized internationally, and will from our side in
the future be followed up by increased engagement in international standardization communities.

SINTEF Materials Technology/NTNU (and IFE in some specific areas) have been significant
R&D partners during the above described development period. In the 1970s much attention was
given to the development of low-alloyed steels with good welding properties, fracture and fatigue
(corrosion-fatigue) resistance. This was followed by the development of hyperbaric (under-water)
welding. Development and use of new types of stainless steels, titanium and composites were
addressed during the 1980s. The 1990s has been characterized by the development of high-
strength steels and corrosion resistant alloys (e.g. supermartensitic 13% Cr stainless steel quality),
especially designed for offshore pipelines.

The oil and gas installations and infrastructures will need continuously monitoring, upgrading and
maintenance, still for many decades ahead. The future challenges with respect to “maintenance-
free” sub-sea installations, general deep water solutions with high reliability, and down stream
natural gas technology development, are examples which all require new and upgraded
competence.

3.1 Future development

Since the beginning of the Ekofisk development in the early 1970-ties Norwegian exploitation of
oil and gas resources has moved into increasingly harsher environments. There is no indication
that this trend will discontinue. The next steps could be fields in water depths of 2000 metes or
more, or fields in ice. The past development has shown that new materials are taken into use at a
steady rate. S13Cr steel pipelines and flexible risers are examples. For future field developments
materials like high strength steels, titanium alloys, composites and advanced polymers may be
required for critical components. For “bulk” applications (structures, pipelines, risers/riser towers,
mooring systems, etc) the most likely development would be use of more corrosion resistant
materials, improved corrosion control, and a gradually increased strength (especially for steels).
Use of composites is likely to be somewhat increased where the combination of flexibility, weight
reduction and non-corrosive properties are wanted. For metallic materials development of
coatings and painting systems has to be improved. The general durability of coatings does have
 5

improvement potentials, and new products will soon be commercially available (e.g. smart
paintings for crack detection).

These developments will represent new challenges to the engineering community, with regard to
basic material science, degradation mechanisms and safety assessment.

In the light of this we think future recruitment, education and competence building will focus on;
- improved understanding of fundamental ageing and degradation phenomena
- how these mechanisms alter the structural integrity over the lifetime
- how these effects can be quantified and be connected to modern risk assessment and lifetime
prediction/assessment methodologies
- how this can be communicated via software/computational tools
- how this information can be linked to the e-Field1 concepts for real-time monitoring,
operation and decision making support

The classical disciplines Structural Engineering & Design, Materials Technology and Risk
Assessment, should in the future work much closer and interactively. By stimulating this cross-
disciplinary interaction, the challenges related to robust and cost-effective solutions in the oil and
gas industries, competitive in an international industry, will be overcome and represent a
significant added-value for the industry.

There are three trends within Materials Technology and related areas we think will be further
emphasized in future development for the oil and gas industry to ensure the most robust solutions;
- the development of quantitative structural integrity methodologies/tools (coupling between
structural engineering, materials technology, risk/safety assessment methods)
- the development of 3D flow assurance and multiphase flow tools, now going on in SINTEF,
will support further development of computational tools for degradation of materials
- the development of nano- and functional materials.

Furthermore, these trends and development scenarios could find many potential innovative
solutions when coupled with the development in the ICT area, heading for fully integrated and
real-time monitoring, control and operation tools within the frame-work of the e-Field concepts.

The e-Field conceptual way of thinking should be adopted in the new boost in offshore
engineering. Functional materials and smart/intelligent components will probably be introduced
along with future development in the oil and gas industry, where censoring and monitoring
functions are built-in for control and operation purposes. The e-Field concept also opens for
including real-time direct calculations of the consequences of failures (based on physical
degradation and crack propagation/resistance models).

3.2 Concluding remarks

The situation on the Norwegian Continental Shelf is generally characterized as mature, and the
solutions sought for will more and more be tailor-made for tail-production scenarios, and
production in more difficult accessible areas (deep waters, long transport). However, there is still
expected profitable oil and gas production for more than 30-50 and 70-100 years ahead,
respectively.

The National (Norwegian) Technology Strategy for Added Value and Competitive Advantage in
the Oil and Gas Industry (OG21; see www.og21.org) has recently been established with

1
E-field combines all information in intelligent decision tools to support key work processes for a selected mode of
operation
 6

commitment from all major oil companies operating in the North Sea basin. It is expected that
future research and development within the oil and gas area will be consistent with this strategy.

The strategy has been established recognizing the need to have a comprehensive and efficient
system for research and technology development on a permanent basis in the oil and gas industry
in Norway. The motivation is to develop new knowledge and technology to ensure economic and
environmentally sound development of the resources on the Norwegian Continental Shelf.

However, may even more important will be to ensure technology export on the international
marked by developing world-class technology solutions.

3.3 Recent Materials Technology R&D initiatives

Several initiatives have lately been undertaken (and are in progress), and some are planned in near
future:

a) Norwegian-Japanese cooperation on pipeline technology

b) OG21 “Positioning Document” for Materials Technology. Included in this work is the
proposal of the establishment of a large R&D program on Structural Integrity. As a
background for this document is the comprehensive technology gap studies within the defined

• Zero harmful discharge to sea

Technology Target Areas as defined by OG21:

• 30% reduction of emissions to the atmosphere

• stimulated recovery
• cost effective drilling
• real-time reservoir management
• deep-water floating technology
• long range transport of well stream
• seabed and downhole processing
• competitive gas production and offtake
c) Internal strategic projects in SINTEF on pipeline integrity management (including sensor,
monitoring and communication technology)
d) Cooperation with IFE and Statoil on residual stress field simulation (case: girth welding of
pipes)
e) Stålmat initiative on Constraint effects in fracture mechanics assessment (Application of
Constraint Corrections in Design and Failure Assessment)
f) International JIP initiative on hydrogen embrittlement
g) Fatigue behavior in high/ultra-high strength steels (updating the design basis), including weld
repair methods and quantification on their effects
h) DEEPLINE – Design, Installation and Operation of Deepwater Pipelines, a strategic
programme carried out by MARINTEK, funded by NFR and industry
i) BFLEX – computer program for calculation of failure and degradation mechanisms in flexible
riser pipe-wall
j) SIMLA – Numerical simulation tool to study stress distribution on a complete pipeline on the
seabed
k) LINKpipe – Numerical simulation tool to study how cracks or defects affects the performance
of pipes
 7

4 COMMON DENOMINATORS FOR INCREASED ROBUSTNESS

The common denominators presented here are based on the experience gained through the work
on S13Cr steels as pipeline material and the work on flexible riser systems /3/4/. The design and
construction of pipelines and flexible risers are very different. This also reflects the difference
regarding degradation and failure mechanisms. Still there are many common experiences which
could represent a more general basis to the term robust material selection.

4.1 Management of oil and gas development projects

In this case management is linked to the technical leadership of the operator/owner/purchaser with
focus on responsibility. The scope of the present project is mainly technical. However, the
working group of this RMS project has been exposed to aspects of management of oil and gas
development projects which undoubtedly must be taken into account regarding robustness.

Cost-benefit approaches are in many cases the driving force for technology developments. For
robust material selection this means increased utilization of materials, more narrow tolerances on
properties and dimensions and reduced time schedules. It is clear that the organization and
management of such field development project differs between the companies. But, the general
aspects to consider are the following recommendations. Some of the input is based on experience
from other projects, some is based on interviews with actual operators/suppliers and some is based
on a questionnaire on S13Cr steel as pipeline material. The recommendations to the management

• taking responsibility for securing relevant competence and to secure interactions between
are:

• improving the use of "Best international practice". Technology transfer from one project to
fields of competence through all stages from design to operation.

• taking responsibility regarding the connection between design criteria and operational
another should be outlined in written procedures.

conditions (i.e. elevate the accuracy of lifetime assessments, evaluation of risk of failure
etc.). Both flexible risers and S13Cr as line pipe material are considered as new
technologies and the operational experience is too short to verify reliability and robustness
in demanding applications. In this situation design must be based on test data. Accelerated
environmental assisted tests may in many cases need to run for long periods to be realistic.

• more detailed time schedules accounting for upcoming problems during construction
This has to be accounted for if test data is not available.

projects. The oil companies are mostly satisfied with the time schedules regarding pipeline
construction projects (cf. questionnaire on S13Cr steel). This is not the case looking closer
on the requirements on fabrication schedules and how upcoming problems in this phase of
construction have increased risk of poor design and lack of evaluations. Right now
different types of cladded pipes are focused. Reeling or J-laying, introducing larger plastic
strain to the pipe, is intended to be used for installation. Where is the time schedule and

• the riser system should be included as a primary element of a production system, and
budgets for fundamental research on the integrity of such pipes during operation?

reliability and robustness of the riser system should be considered at the stage of concept
evaluation of a new field. Long term realistic tests and verifications should be expected in
the evaluation of risk and lifetime assessments.
So far the riser system has been considered as proven technology. This is not the actual
case as the experience of long term durability is very limited and that the flexible riser
systems are introduced to more and more demanding environments and loadings. When a
flexible riser system is inherent in the field concept and the field conditions are more
demanding, the designer is left with selection of new materials in the riser system. These
new materials or material combinations are not verified for long term durability.
 8

• competition between suppliers of flexible riser systems has lead to unfortunate situations.
The design criteria differ and they keep the design basis as a secret. This makes
comparison of the suppliers difficult. The commonality on fundamental research and
development suffers also because of the secrets between the suppliers. It is recommended,
on a contractual basis, that the operators require full and open access to details on the
construction and the risk/lifetime assessments from the suppliers.

4.2 Design
It is identified a large potential for increased robustness by improving the completeness of
operational design of flexible risers and pipeline systems. The field experience on flexible risers
and flowlines indicates that the design basis has been insufficient. The statement of insufficient
design is based on experience from failure investigations, showing that the stress/loads in many
cases have been underpredicted relative to the maximum allowable design stress/loading. This is
often combined with wrong predictions of degradation- and failure mechanisms. New and
unexpected failure modes have been experienced, in many cases related to material properties and
degradation or failure mechanisms. Development of more accurate tools for design and life
prediction by means of numerical simulations combined with new test methods, full scale testing,
instrumentation and monitoring during operation and field experience in general should have had
more focus. More effort on iterative design and verifications is emphasized. This can be partly
realized through the following recommendations:

• enhanced requirements for third party evaluation of new pipe concepts or new applications
-for flexible risers:

of flexibles. Design of flexible risers involves advanced materials, interaction between

very different materials in a complicated structure, and time dependent degradation

• fatigue design of armour wire based on operational conditions: Aqueous annulus with H2S
mechanisms. This kind of independent third party expertise requires resources.

and/or CO2 permeating from bore. Sea water ingress, oxygen level and cathodic protection

• more work into the mechanisms of sealing and fixation in the end termination should be
are additional environmental factors.

undertaken, in particular regarding response of polymer materials to cyclic temperature

• investigation of fatigue initiation and crack growth in PVDF materials. Recent design
and stress, and to ageing mechanisms

modifications of end terminations may lead to fatigue problems due to crack growth
through the liner.

• take into account the strain history from installation on the properties of the line pipe
-for pipelines:

• take into account environmental impact on line pipe material properties regarding fatigue
material

• take into account the effect of general corrosion on pipe wall thickness and local corrosion
crack growth and fracture.

• take into account typical weld defects (Engineering Critical Assessment)

(notch effects).

• requirements on increased documentation of seabed topography along the planned route of

the pipeline and how it affects the stress distribution on the pipe during normal operational

• after installation and before operation, document the exact position of the pipe to verify
conditions and shut downs (high cycle and low cycle fatigue respectively).

whether the design criteria are met or not. If deviations are registered, re-design and
evaluation of remedial actions are highly recommended.
 9

• on pipe section(s), which are considered as most critical (highly loaded) during operation,
observe/register pipe movement during Ready For Operation procedure and initial

•
operation for comparison with design criteria.

•
improve the interaction between local and global design analyses.

•
avoid stress/strain concentrations in highly loaded sections.
in addition to the monitoring on the well head and pipeline end (platform/on shore),
improve monitoring of critical parameters on pipeline during operation as:
-local loadings (stress)
-cracking
-wall thickness
-pH
-flow

4.3 Knowledge gaps

This chapter looks closer on ways to reduce knowledge gaps and thereby reduce the risk of
incidents offshore.

The term "complete design" includes a complete understanding and quantification of all possible
sources of environmental impact on the construction/structure during operation. However large
holes are present today. This may be one of the major reasons why the predictions of lifetimes on
S13Cr pipelines and flexible risers are questioned world wide. The design basis, the operational
experience and operational monitoring/registrations are limited.

The experience that has been accumulated through the last few years has shown that even if a
flexible riser design or a S13Cr flowline has been fully qualified, failures have occurred in
unexpected modes. These failures may likely be defined as a result of knowledge gaps. The
following recommendations are likely reducing the risk of incidents and failure of such subsea
systems:

A. We have to close the gaps that are known and present today on pipeline- and flexible riser
systems. This requires establishment of new relevant test methods and extensive testing and
verifications programs. As soon as a test method is critically evaluated and regarded acceptable
for qualification purposes, it is of main importance to standardize the method. It is also vital to
implement the test results in relevant standards and specifications.

B. Bring the design analyses closer to completeness (see section 4.2).

C. It is clear that fundamentals regarding environmental assisted failure mechanisms are not fully
understood. It is recommended that more basic and long term research is initiated to provide a
better understanding of these issues (including Ph.D. programs).

D. Strategic strengthening of the communities in the offshore industry, universities and research
institutes on Environmental Assisted Mechanics (cf. EA Cracking, EA Fracture Mechanics, EA
Fatigue etc.). Strategic investment on relevant equipment on a national basis is also necessary to
bring forward, taking into account the large volumes of laboratory work connected to
characterization and optimization of different materials exposed to different operational
conditions.

E. It is believed that closer collaboration between oil companies/operators,

manufacturers/fabricators and research institutions/universities are likely an effective way of
 10

reducing technology gaps in the future. To be able to identify possible sources of failure, it is
required open communication. This is especially a major challenge for suppliers of flexible risers.

5 REFERENCES:

1. Oljedirektoratet: Sammendragsrapport, "Utvikling av risikonivå – norsk sokkel" Fase 3 –

2002, April 2003
2. J.M.Drugli, C.Thaulow, J.Ødegård, T.Rogne, R.Stokke, S.Berge, J.Berget: Pre-project,
"Robust material selection in the offshore industry", SINTEF Report STF24 F03202,
January 2003
3. H.Lange, T.Rogne, "Material selection of weldable super martensitic stainless steels for
pipeline material in the offshore industry SINTEF Report, STF24 F04222, February 2004
4. S.Berge, T.Glomsaker, "Robust Material Selection (RMS) in the Offshore Industry –
Flexible risers", Marintek Report, February 2004
 2

TABLE OF CONTENTS

1. INTRODUCTION.................................................................................................................. 4
2. FLEXIBLE RISER WALL STRUCTURE ......................................................................... 5
2.1 Introduction ...................................................................................................................... 5
2.2 Carcass ........................................................................................................................... 5
2.3 Liner ........................................................................................................................... 7
2.4 Annulus ........................................................................................................................... 8
2.5 Pressure armour................................................................................................................ 9
2.6 Tensile armour................................................................................................................ 10
2.7 Outer sheath.................................................................................................................... 10
2.8 Anti-wear tape ................................................................................................................ 11
3. FAILURE MODES..............................................................................................................12
3.1 Carcass failure modes.....................................................................................................12
3.2 Liner failure modes ........................................................................................................12
3.2.1 Pull out and rupture of pressure barrier in end termination ............................ 12
3.2.2 Collapse of carcass due to gas permeation...................................................... 13
3.2.3 Degradation of Rilsan .................................................................................. 13
3.3 Failure modes for pressure armour................................................................................. 14
3.4 Failure modes for tensile armour.................................................................................... 15
3.4.1 General ............................................................................................................ 15
3.4.2 Wear ................................................................................................................ 16
3.4.3 Fatigue and corrosion fatigue.......................................................................... 16
3.4.4 Hydrogen induced cracking ............................................................................ 17
3.4.5 Anti-wear tape ................................................................................................. 17
3.5 Failure modes for outer sheath .......................................................................................17
4. OPERATIONAL EXPERIENCE.......................................................................................19
4.1 Experience by Norsk Hydro ........................................................................................... 19
4.2 Experience by Statoil ..................................................................................................... 19
4.3 Information from suppliers of flexible risers ................................................................. 19
5. DISCUSSION ....................................................................................................................... 21
5.1 General ......................................................................................................................... 21
5.2 Metallic components ......................................................................................................21
5.2.1 Stainless steel .................................................................................................. 21
5.2.2 Carbon steel..................................................................................................... 21
5.3 Polymer components ......................................................................................................22
6. ORGANISATIONAL FACTORS ...................................................................................... 24
6.1 Choice of riser system ....................................................................................................24
6.2 Competition vs. openness in the market.........................................................................24
7. RECOMMENDATIONS..................................................................................................... 26
7.1 General recommendations.............................................................................................. 26
7.2 Specific recommendations ............................................................................................. 26
RELEVANT TESTS.............................................................................................................. 30
 4

INTRODUCTION

During recent years a significant number of flexible risers on the Norwegian Continental Shelf
have suffered from failure, for a variety of reasons. In several cases the failure mode was related
to material properties, and was not foreseen in design. Long term durability and reliability of riser
systems may be questioned.

In this report an overview of failure modes and critical factors for material selection is provided.
The report is based on interviews of industry representatives with competence on flexibles, and
general knowledge within MARINTEK/SINTEF. Recommendations are given for improved
robustness and reliability of materials for flexible risers.

T/prosjekt/70/700271/MT70 F04-048
 5

1. FLEXIBLE RISER WALL STRUCTURE

1.1 Introduction
Flexible risers are used for a range of functions: Production risers for gas and oil, water injection,
gas lift, gas injection, oil or gas export, test productions etc. Flexible risers are also used for
drilling and well maintenance. In this report the discussion is limited to the transport function.

Flexible risers used for production, injection or export are likely to be subjected to a number of
conditions that may be affect the integrity of the riser. Due to the rather complicated wall
structure where materials with very different properties are interacting, Figures 1 and 2, a large
number of failure modes are possible. Many of these failure modes are related to material
properties. In this section the different layers of a flexible riser are described with respect to
function, structure, material and possible failure modes.

Figure 1. Typical cross section of flexible pipe wall structure.

(1) Stainless steel carcass
(2) Thermoplastic liner
(3) Carbon steel pressure armour
(4) Carbon steel tensile armour, two contra-wound layers
(5) Thermoplastic outer sheath
Polymer tape, which is used to minimise friction and wear between layers of armour, is not
shown. Additional layers of material with low thermal conductivity may be applied in order to
obtain specific thermal insulation properties of the pipe.

1.2 Carcass

The carcass is the innermost layer of a pipe, and the only metallic component that is in direct
contact with the fluid in the bore. The carcass is made from stainless steel strip in a continuous
process onto a mandrel. The material is stainless steel, typically AISI 316L or similar, that needs
to be compatible with the chemical constituents of the transported liquids and/or gases.

T/prosjekt/70/700271/MT70 F04-048
 6

Figure 2. Pipe wall structure, in more detail.

The function of the carcass is to provide strength against external hydrostatic pressure, and
mechanical protection of the liner against pigging tools and abrasive particles. The carcass also
provides strength to resist crushing loads during e.g. installation operations. At large water depths
the hydrostatic as well as the crushing loads will increase.

The carcass is an open structure and does not provide any containment of internal pressure, i.e. oil
and gas can flow freely across the carcass. Flexibility is obtained by the ability of each profile to
slide with respect to the neighbouring profiles.

Figure 3. Typical carcass profile.

In the case of a damaged outer sheath, the external pressure will be acting directly onto the liner,
and must be carried by the carcass alone. A basic design criterion is thus the external pressure at
maximum water depth, assuming empty pipe.

Carcass collapse may also be caused by release of absorbed gas in the liner. During high pressure
operations the liner will become saturated with gas, which will be released in periods of low
pressure shut-down. For multilayer liners, if proper venting is not provided for the slits between
the layers of liner, pressure may build to cause collapse of the carcass. The carcass is normally
not designed for this condition, which must be avoided by design or by operational restrictions.

The collapse capacity is strongly dependent on whether the carcass is fully supported by the
surrounding structure (liner/pressure armour) or whether there is a gap. A gap may be caused by
several mechanisms:
T/prosjekt/70/700271/MT70 F04-048
 7

− Stretching and possibly deformation of the carcass due to force of gravity.

− Shrinking of liner due to deplastification, possibly counteracted by swelling due to absorption

− Pressure induced creep of liner into gaps in pressure armour and carcass profiles.
of hydrocarbons and/or gas.

The forming of a carcass profile from metal strip is a cold-forming process. For a given material
there are technological limitations with regard to the thickness of a strip that can be formed to a
carcass profile like the one shown in Figure 3. The collapse strength of the carcass, which is
essentially determined by the stiffness (moment of inertia) of the profile, has basic limitations for
this reason. Collapse strength to resist hydrostatic pressure as well as crushing loads from
installation equipment may be a limiting factor for deep water applications of large diameter pipe.

More compact and stiffer profiles are under development, Rytter and Rishøj (2002), that may
extend the deepwater range of flexible risers.

1.3 Liner
The liner is the sealing layer, made from a thermoplastic by extrusion over the carcass. In some
applications a multi-layer liner is used, with sacrificial layers on the inside and/or the outside of
the sealing layer. The purpose of the sacrificial layers is to provide protection against the metallic
components. The liner is exposed to the fluid in the bore, and limits the upper service temperature
of the riser and the chemical composition to the various fluids that may be transported through the
line.

Different materials may be used, depending on the design conditions. Three generic classes of
materials used as liner are:

− High density polyethylene (HDPE) and cross-linked polyethylene (XLPE)

− Polyamide (nylon) (PA11 or PA12)
− Poly vinylidene fluoride (PVDF)

Within each class of materials a large variability in properties are available. Some of the materials
used are brand names, protected by patents or licenses as shown in Table 1.

Table 1. Polymers used in the liner for flexible risers

Material Producer Used by supplier
Rilsan® PA11 + plasticizer Atofina All
Solef® 60512 PVDF/CTFE* Solvay NKT, Wellstream
Solef® 1015/078 PVDF + plasticizer Solvay Wellstream
Gammaflex® PVDF/HFP** Atofina + Technip Technip
Coflon® PVDF + plasticizer Atofina + Technip Technip
HDPE/XLPE
*
Copolymer with Chlorotrifluoro-ethylene
**
Copolymer with Hexafluoro-propylene

The trade names Coflon® and Gammaflex® are hold by Technip but are probably based on
Kynar® and Kynar Flex® respectively from Atofina, possibly modified by use of their own
additive package, particularly plasticizer.

T/prosjekt/70/700271/MT70 F04-048
 8

A major criterion for selection of liner material is the design temperature. In general, HDPE has
very good chemical resistance and maintain good mechanical properties up to ~60°C. However,
hydrocarbons like crude oil, methane, methanol etc., are absorbed and may work as plasticizer.
Therefore, if the bore fluid contains hydrocarbons, HDPE may be used at low and moderate
temperatures only, generally below 20 – 30°C. Crosslinking may in general improve high-
temperature properties and in addition reduce the absorption of hydrocarbons, and thus XLPE
may be used at somewhat higher temperatures than HDPE.

Polyamide materials may be used at higher temperatures but is very sensitive to humidity. In the
case of a high water cut, polyamide suffers from hydrolysis at elevated temperatures. The main
mechanism of hydrolysis of PA11 and PA12 is chain scissoring (reduced molecular weight),
causing brittleness. Prediction of service life in various environments has probably been a major
complication with this material, Ottøy (2001). The Rilsan User Group (RUG), which is a
collaboration between a range of operators and suppliers, have however developed a new
procedure to predict the life time and degradation rate for Rilsan® and published by API as a
Technical Report (API TR 17 RUG). According to API TR 17 RUG, the extrapolated service time
for Rilsan® at 60°C in a typical well flow (humid, pH4) is 20 years. Nevertheless, Rilsan is
considered as the most used liner material within the North Sea, MCS (2001).

PVDF may be used at higher temperatures, possibly 130°C with present brands. However,
typically 20% plasticizer is added to PVDF homoploymer, in order to improve processing
(extrusion) properties and reducing the possibility for defects like blisters. In contact with
hydrocarbons, the plasticizer tends to be extracted from the PVDF, leading to permanent
shrinkage of the material which again has contributed to several failure modes as described in
detail in the next chapter: Pull-out or rupture of the liner at the end termination and decompression
of the carcass due to pressurized gaps between the layers. The former problem has been mitigated
by use of accelerated deplastification of the end section of the liner before mounting of end
termination.

A recent method to improve the processability of PVDF without use of additives is to apply a
copolymer. Solef ® 60512 and Kynar Flex® are PVDF copolymers. For the Solvay material the
comonomer applied is, according to available information, CTFE (chlorotrifluoro-ethylene), and
for the Kynar Flex® the comonomer is HFP (hexafluoro-propylene). New development is also
ongoing for liner materials in order to resist even higher temperatures, MCS (2001). This
development is mainly run by the suppliers and the material industry.

1.4 Annulus
The section between the liner and the external sheath is the pipe annulus. This is an open
structure with no pressure barrières. In a pipe that is transporting a fluid under high pressure, gas
and liquid will permeate through the liner and cause pressure build-up in the annulus. To prevent
rupture of the external sheath, the annulus is vented at the end terminations, typically at 1 bar
overpressure.

In the as-fabricated state, void space in the pipe annulus is filled with atmospheric air. Fatigue
design of steel armour wire in the annulus has been based on an assumption that the environment
is benign. However, during operation the chemical composition may change, for several possible
reasons:

Leakage by damage of the outer polymer sheath, caused during installation or operation. This
would lead to sea water flooding of the annulus. Depending on the nature of the leak, sea water in
T/prosjekt/70/700271/MT70 F04-048
 9

the annulus could be depleted of oxygen, or possibly saturated with air. Efficiency of cathodic
protection is uncertain. Sea water may be combined with H2S and/or CO2 due to permeation
from the well flow.

Permeation of species from the product flow, notably water (H2O) which may condense and
accumulate in the annulus, in combination with gaseous components like hydrogen sulphide
(H2S), carbon dioxide (CO2) and methane (CH4).

Risers which have been subjected to sea water ingress may be repaired, flushed with inhibitor, and
re-installed. Inhibitor fluid, possibly with some residual sea water and with CO2 or H2S due to
permeation from the well flow, could have an effect on residual fatigue life.

These environments may have a significant effect on fatigue life of steel components, and need to
be considered in design and operation.

1.5 Pressure armour

The primary function of the pressure armour is to resist the hoop stress due to internal pressure.
The pressure armour is also a strength component with respect to external forces, e.g. crushing
forces due to handling or accidental loading.

The pressure armour is an interlocking profile made from rolled carbon steel with tensile strength
in the range 700 – 900 MPa. Three different profiles are currently in use, Zeta/Flexlok, C-clip and
Theta, Figures 4-6. Some of these profiles are protected by patents or licenses.

The interlocking of the pressure armour is a limiting factor for the minimum bending radius of the
riser. If this limit is exceeded, irreversible damage to the flexible line will occur, leading to
perforation of the liner when subjected to internal pressure.

Figure 4. Zeta/Flexlok interlocking profile used as pressure armour.

T/prosjekt/70/700271/MT70 F04-048
 10

Figure 5. C-clip interlocking profile used as pressure armour.

Figure 6. Theta-shaped interlocking profile (two variants) used as pressure armour.

For high pressure applications the interlocking layer may be strengthened by an additional layer of
flat steel profiles that are not interlocked, cf. Figure 3.

Zeta/Flexlok is used by Technip and Wellstream. However, the two suppliers have developed
proprietary variants of the profiles.

C-clip® is used by NKT, and Theta® is used by Technip.

1.6 Tensile armour

The tensile armour is two or four counterwound layers of armour wire and provides strength
against axial stress caused by internal pressure and by external loads. The tensile armour also
provides torsional strength to the pipe. However, for torsional loads in a direction leading to
unwinding of the outer layer of armour the strength and stiffness is poor.

1.7 Outer sheath

The function of the outer sheath is to provide a seal against the sea water in order to prevent
corrosion and to give mechanical protection to the steel armour. The loads typically applied on the
outer sheath is impact, erosion and tearing as well as, in certain cases, external or internal
pressure. The material is extruded thermoplastic – HDPE or Rilsan®. According to available

T/prosjekt/70/700271/MT70 F04-048
 11

information, MCS (2002), nearly 40% of the riser failures are due to external sheath damage and
most of these took place during installation.

1.8 Anti-wear tape

In a flexible riser subjected to cyclic bending the steel armour will slide cyclically. If two layers
of steel armour are in direct contact, wear may take place. For this reason anti-wear tape is
applied between layers of steel armour. The tape is not leak-proof, and fluids in the annulus may
flow through the tape.

The materials used are thermoplastic tape like Rilsan with a thickness in around one millimetre.
The tape is subjected to significant contact stress and large slip amplitudes. Cumulative slip for a
20 year design may be of the order of 50⋅103 m. The tape must thus retain a minimum strength at
the temperature of the armour, and to be wear resistant.

T/prosjekt/70/700271/MT70 F04-048
 12

2. FAILURE MODES
Due to the complicated and composite structure of a pipe wall, a large number of failure modes
are possible. In this section failure modes as known from service experience and full-scale tests
are described. Failure modes due to mechanical overload are not discussed in any detail. Focus is
on failure modes that are related to material properties and material selection.

2.1 Carcass failure modes

The carcass may be subjected to a large number of failure modes; overstretching, fatigue, radial
collapse, wear, erosion, corrosion and damage from pigging and similar operations.

Radial collapse has been reported in several cases, caused by pressure build-up in multi-layer
liners. Calculations have showed that in these cases the pressure was in excess of the design
collapse pressure for the carcass. The cause of the collapse was thus not related to material
properties or designed capacity, but due to an unexpected mode of loading.

Fatigue or wear damage of the carcass has not been reported in the open literature. Due to the
structure and loading on the carcass, fatigue is not a likely failure mode, except as secondary
damage due to initial damage in the production phase or due to pigging or the like.

The carcass may be subjected to erosion and erosion-corrosion in production risers for gas-
condensate fields with sand production. Full scale tests have demonstrated significant material
loss under conditions representing realistic operational conditions, Kvernvold (1992). Prediction
tools have limited accuracy, and the design envelope for safe operation is uncertain.

The full scale tests have shown that corrosive environments with CO2 will give enhanced erosion
rates. Plain corrosion has not been reported for the carcass.

2.2 Liner failure modes

2.2.1 Pull out and rupture of pressure barrier in end termination

The end termination of the liner must provide two functions:

- Sealing of pressure in the bore

- Fixation of liner to prevent pull-out

Various mechanical solutions are employed by the suppliers. In one early design the polymer
layers were restrained in the axial direction by a clamping force by use of a locking ring. The
clamping force was in the original design applied on the polymer sheet onto the carcass and was
therefore limited by the stiffness of the carcass. Upon thermal cycling the compressive forces
generated in the polymer layers at high temperature due to higher thermal expansion than the
steel, will relax and after some time or cycles, tension will appear in the polymer at low
temperature. This mechanism in combination with migration of plasticizer in the PVDF resulted
in 1994-95 at Snorre in pull-out from the end termination.

The mechanism of pull-out was verified in a mid-scale test by MARINTEK, Berge and Eide
(1999).

T/prosjekt/70/700271/MT70 F04-048
 13

After redesign of the end termination the problem was solved and has not appeared later in
service. However, the improved fixation has lead to a different problem, fatigue cracking initiated
by the imprint of the locking ring. Failures have been reported in mid scale testing and in service.
This is a problem which may require more research, Melve (2001).

2.2.2 Collapse of carcass due to gas permeation

Gas in the bore will under high pressure diffuse into the polymer liners and be solved in the
polymer. This is a very slow process and takes several weeks, up to a few months, in order to
reach steady state conditions. A large amount of gas may be solved in the polymer material. For
risers with multiple liners, pressure will build up between the liners and create a pressurized gap
when the bore pressure is reduced. For high operational bore pressure (typical of gas injection,
~30 – 40 MPa) there may be solved more than enough gas in the polymer to diffuse into the gap
and build up a gas pressure at the interface, large enough to collapse the carcass if the bore is
depressurized.

The mechanism has been predicted by use of simulation, Glomsaker (2002). Typical result from
simulation of the gas pressure inside the liner after pressurization of a riser from 1 bar to 390 bar,
and subsequently depressurization again to 1 bar, are shown in Figure 7.

a) Pressurization b) Depressurization

Figure 7. Gas pressure in a three layer liner of PVDF.

The outer gap is located at radius = 0.093 m

A second failure mode related to diffusion and solution of gas in the polymer is blistering of
internal voids. Very little information is available for this failure mode.

2.2.3 Degradation of Rilsan

Rilsan® (PA11) is until now the most used liner-material in flexible risers, MCS (2001). There
are several degradation mechanisms for polyamide like thermal degradation, oxidation,
photodegradation, absorption of water etc. However, for PA11 in the actual humid environment
free of oxygen, the dominating process will be hydrolysis. Hydrolysis results in scissoring of the
polymer chains and cause brittleness of the material. There has been large uncertainty about how
to predict hydrolysis rate in order to ensure 20 years life time of the product. The hydrolysis rate

T/prosjekt/70/700271/MT70 F04-048
 14

increases rapidly with temperature and somewhat less with increasing sourness. Most of the
failures related to degradation of Rilsan (as on Njord) however, seem to have been caused by
underestimated water content in the fluid. In addition, Atofina claims on their web page that there
has never been documented a flexible pipe failure caused by failure of polyamide-11 operated
within the recommended service window. The operators however claim that there were large
divergences between different aging models until the API TR 17 RUG was prepared by the Rilsan
User Group – founded in 1998 consisting of a large number of operators and suppliers. As
indicated in Figure 8, this work resulted in a significant reduction of predicted life time. The
degree of degradation is according to this procedure quantified by the inherent viscosity – which
is a magnitude calculated from the viscosity of the polymer in a suitable solvent, corrected for
content of plasticizer. This magnitude is named corrected inherent viscosity (CIV). CIV reflects
the degree of degradation because the viscosity is related to the average molecular weight of the
polymer, at a certain temperature and solvent, through the Mark-Houwink equation. The RUG has
established CIV = 1.2 as end-of-life criterion for Rilsan because at this level of degradation one
find an abrupt drop in mechanical properties. We also note that in the RUG-method there are no
option considering unsaturated or dry environment – life time predictions are suggested to always
include a saturated water phase.

30
pH=4 (API TR 17 RUG )
pH=5 (API TR 17 RUG )
25 pH=7 (API TR 17 RUG )
Saturated (API RP 17B)
Unsaturated (API RP 17B )
20
Life time [years]

15

10

0
50 55 60 65 70 75 80 85 90 95 100
Temperature [°C]

Figure 8. Predicted life time for Rilsan® for various service temperatures according to API RP
17B (2002) and API TR 17 RUG (2002)

2.3 Failure modes for pressure armour

The pressure armour is subjected to several degradation mechanisms:

In a pipe that is subjected to cyclic bending, the contact points of the profile will slide cyclically,
with considerable contact stress. Due to the varying bending, the contact pressure will vary

T/prosjekt/70/700271/MT70 F04-048
 15

cyclically. This is a problem particularly for the Zeta profile, Figure xx, where the contact stress
causes cross-wire bending. The sliding may thus result in significant cyclic stresses in the cross-
wire direction and possibly fatigue failure. The fatigue life may be affected by fretting at the
contact points. The associated failure mode is cracking parallel to the wire axis.

Ovalisation of the pipe due to curvature variations and possibly side loads from a bend stiffener or
a bellmouth will give cyclic stresses longitudinal to the armour profile, and possible fatigue
cracking normal to the axis of the armour.

Both these failure modes have been reported from full scale fatigue testing of risers with Zeta
profiles. Theta and C-clip profiles appear to have better fatigue properties.

The presence of aqueous environments with H2S and/or CO2 may have a significant effect on
fatigue strength. Such environments are probable. Gas will permeate through the liner from the
high pressure well flow. Fresh water may also permeate from the bore, or there may be sea water
ingress through damage in outer sheath.

At the contact points the armour may be subjected to wear.

Computational methods based on finite boundary elements are available for calculation of cross
wire stress of pressure armour, Sævik et al. (2001). Test methods have been developed for cross-
wire and longitudinal fatigue loading of pressure armour, Berge (2001). The effect of
environment may also be tested in small scale, Berge et al. (2003). The methodology for doing
fatigue design analysis is thus established and proven, Sævik (1998), Berge et al. (2001).

2.4 Failure modes for tensile armour

2.4.1 General
Potential failure modes for tensile armour may be listed as:

−
−
overload in tension, possibly in combination with internal pressure, causing tensile failure

−
overload in bending or compression causing wire disarray or birdcaging

−
overload in torsion causing unwinding of armour or birdcaging

−
fatigue

−
corrosion fatigue

−
fretting fatigue

−
wear

−
hydrogen induced cracking
corrosion.

The overload modes of failure are not related to material properties, but to design and operational
conditions.

The remaining modes may be linked, and a large number of synergistic mechanisms are possible –
at least in theory.

T/prosjekt/70/700271/MT70 F04-048
 16

2.4.2 Wear
In early design of flexible risers the layers of tensile armour was laid directly against each other,
with direct metallic contact. For this design wear of tensile armour was considered to be the
limiting factor for service life of flexible risers. Wear life was assessed by rotating bending
testing, which was a very conservative test. Since approximately 1990 fabricators have applied
layers of polymer tape – anti-wear tape – between layers of metallic armour. Provided the tape
can withstand the contact pressure, this has proved to be effective, and wear is no longer
considered a problem for tensile armour. However, the long term durability of anti-wear tape in
demanding conditions (large contact pressure and/or high temperature) has not been proved, as
disussed in Section 3.4.5.

2.4.3 Fatigue and corrosion fatigue

Design analysis has shown that fatigue of tensile armour may be critical with respect to design
life. In the as-fabricated state, void space in a pipe annulus is filled with atmospheric air. For this
reason fatigue strength criteria have been derived on the basis of fatigue tests in air, assuming the
environment in a pipe annulus to be benign. Service experience has shown that during operation
the chemical composition of the annulus is likely to become corrosive, for the following possible
reasons:

- Leakage by damage of the outer polymer sheath, caused during installation or operation, or
by malfunctioning venting valves. This would lead to sea water flooding of the annulus.
Depending on the nature of the leak and the distance from the leak, sea water in the annulus
could be depleted of oxygen, or possibly saturated with air. Efficiency of cathodic
protection is uncertain. Sea water may be combined with hydrogen sulphide (H2S) and/or
carbon dioxide (CO2) permeating from the bore. There is also a possibility for Microbial
Induced Corrosion (MIC) as sulphide reducing bacteria (SRB) may develop in stagnant
seawater.

- Permeation of species from the product flow, notably gaseous components like H2S and/or
CO2, possibly in combination with fresh water (H2O) which may condense and accumulate
in the annulus.

- Risers which have been subjected to sea water ingress may be repaired, flushed with
inhibitor, and re-installed, Taylor et al. (2002). Inhibitor fluid, possibly with some residual
sea water and with H2S and/or CO2 due to permeation from the well flow, could have an
effect on residual fatigue life.

Carbon steel in aqueous environment with H2S and/or CO2 is susceptible to corrosion, and the
fatigue strength is likely to be affected. SN curves for fatigue design are empirical, and need to be
assessed on the basis of relevant data. Until recently no standardised test methods were available
for this type of testing. A test protocol has now been worked out and is being applied in fatigue
testing of armour wire, Berge (2002). Guidance is given on the following:

−
−
specifications for the environment

−
specimen preparation

−
fatigue loading procedure
data processing

Recommendations are given on treatment of mean stress effects in fatigue design, and a unified
method is proposed for assessment of SN design curves from test data.

T/prosjekt/70/700271/MT70 F04-048
 17

No cases of service failure due to fatigue of tensile armour have been reported in the open
literature. On the other hand, very few dynamic risers on the Norwegian Shelf have seen more
than five years of service loading, which is approximately 25% of a typical design life. Until
service experience has been gained, corresponding to the design criteria, fatigue design must be
based on test data and design analysis.

2.4.4 Hydrogen induced cracking

Tensile armour is generally classified on the basis of “sweet” or “sour” service, following the
criteria given by NACE (National Association of Corrosion Engineers), NACE TM 02-84 (1996).

Sour service wire has a strength generally below 800 MPa. Sweet service wire has a tensile
strength in the range 1200-1400 MPa. One fabricator is using a “basic sweet” wire with strenght
in the range 800-900 MPa.

Norwegian operators have been conservative in their specification for tensile armour, using sour
service armour for all production risers even if the product flow is sweet according to the NACE
criteria. The reason for this conservatism is the possibility that wells that are initially sweet, may
become sour through production life. The main reason for development of hydrogen sulphide is
the use of sea water for re-injection, and development of sulphate reducing bacteria inside the
formation.

No cases of hydrogen induced cracking of armour wire have been reported in the open literature.

2.4.5 Anti-wear tape

The anti-wear polymeric tape that is applied between layers of metallic components is a secondary
element of a pipe wall, with no function in terms of sealing or strength. If the tape fails, however,
wear and fretting fatigue is likely to affect the service life of the steel armour, and the service life
may be significantly reduced, Berge and Sævik (1993).

Rilsan® and other types of thermoplastic material are used for the tape. In cases of high contact
pressure and in combination with high temperature the tape may be subjected to wear, creep
deformation and other types of degradation mechanism.

The API Recommended Practice and Specifications do not specify any requirements with regard
to the properties of anti-wear tape. Procedures used by the suppliers of pipe for qualification of
tape materials are not known.

No failure of anti-wear tape has been reported. It should be borne in mind that anti-wear tape has
been used for approximately ten years, and relatively few flexible riser systems with tape have
reached their design life. MARINTEK has developed a test facility for wear testing of polymer
tape, simulating the conditions of tape between layers of armour wire, Berge et al. (2001). It is
known from the testing that the properties of different tape materials can vary widely with respect
to friction, creep behaviour and wear. The API documents provide little guidance on test methods
and acceptance criteria for the tape. This is obviously a topic for further investigations.

2.5 Failure modes for outer sheath

For a flexible riser subjected to cyclic bending, the outer sheath will experience the largest cyclic
strains. The outer sheath is also at a low temperature, and is thus conceptually vulnerable to
T/prosjekt/70/700271/MT70 F04-048
 18

damage. However, no fatigue failures of the outer sheath due to normal operational conditions
have been reported.

Above the waterline, the outer sheath may be subjected to direct sunlight, which may cause
ageing. However, this has not been reported as a problem.

On the other hand, a large number of failures have been reported, due to two main reasons:

− Damage due to rough handling, impact, etc. during installation or operation.

− Failure of the venting system, causing pressure build-up in the annulus and failure of outer
sheath by bursting.

Water ingress has also been caused by malfunctioning or even lacking venting valves.

None of these failure modes are related to material properties, and will not be discussed herein.

T/prosjekt/70/700271/MT70 F04-048
 19

3. OPERATIONAL EXPERIENCE

3.1 Experience by Norsk Hydro

A survey of the flexible risers operated by Norsk Hydro is given in Table 3.

Table 3. Summary of Norsk Hydro’s experience on flexible risers

Platform Number of Liner material Replaced Failure Mechanism
flexible risers
Visund1) 28 PVDF 5 Collapse of carcass
due to gas permeation

Njord 20 PVDF 132) Collapse of carcass

due to gas permeation
PA-11 (Rilsan®)
Degradation of Rilsan

Snorre A1) 5 PVDF 4 Pull-out

Troll B 33 PA-11 (Rilsan®) 1
Troll C 35 PA-11 (Rilsan®) 0
Others 30 3
Total 151 26
1)
Currently operated by Statoil.
2)
Some risers have been replaced more than once.

The history of failures has been reported in the open literature, Olsen et al. (2002).

3.2 Experience by Statoil

Statoil has not (so far) been able to release written information to the project on failed pipes.
However, some information has been released through oral discussions.

Statoil has in general experienced somewhat less problems with flexible risers compared to Norsk
Hydro, cf. Table 3. This may have been fortuituous. From discussions with key persons in the
two companies it appears that the approach to the use of flexible risers is based on the same
knowledge and technology, and the use of the same suppliers.

3.3 Information from suppliers of flexible risers

There are currently three suppliers of flexible risers world-wide:

- NKT Flexibles (Denmark)

- Technip Coflexip (France)
- Wellstream Co (UK and USA)

T/prosjekt/70/700271/MT70 F04-048
 20

The market for flexible risers is very competitive, and the three suppliers are very reluctant to
disclose details about the materials and material selection procedures. In general the procedures
comply with API 17B and API 17J. In addition the suppliers carry out much work in-house and
as commissioned work through research institutes. Very little of this is public domain
information.

T/prosjekt/70/700271/MT70 F04-048
 21

4. DISCUSSION

4.1 General
The operators Norsk Hydro and Statoil both base their purchase procedures on the general
documents API Specification 17J (API-17-J) and API Recommended Practice 17B (API-17-B) in
addition to API Technical Report 17 RUG (API-TR-17-RUG) for life time assessments for
Rilsan® liners.

API-17-J put the responsibility for design and material properties on the flexible pipe
manufacturer. Both the design and the material properties shall be verified by a third party.
Corrosion protection and operational conditions (internal conditions, external conditions and
insulation), necessary for the manufacturer to define load cases, are the responsibility of the
purchaser. There is no requirement in the standards for the purchaser to verify their analysis by a
third party.

4.2 Metallic components

4.2.1 Stainless steel

Stainless steel is used for the carcass structure, which is in contact with the fluid in the bore.
Standard grades are generally used, 304, 316 and similar. No problems related to material
properties (corrosion, fatigue, wear, etc.) have been reported in the open literature.

The main design problem of the carcass is to provide structural strength against external loads.
This problem is related to material grade through material strength properties, and formability for
fabrication. However, this aspect is considered to be outside the scope of work for the present
project.

4.2.2 Carbon steel

Carbon steel is used for pressure armour and for tensile armour. The profiles are produced by
rolling and/or drawing. The main criterion for selection of material grade is whether the product
flow is “sour” (containing H2S) or “sweet”.

The NACE criterion is used as a definition of sour service. As discussed in Section 3.4.4 Statoil
and Norsk Hydro have adopted a conservative approach towards selection of material grade for
armour wire, using sour service wire for all risers that are designed as production risers, regardless
of chemical composition of the initial well flow.

No problems related to HIC have been reported in the open literature.

Due to the finding that the annulus of a flexible riser is likely to be filled with water, much work
is currently ongoing to verify the integrity of risers that are in operation, taking into account the
real environment. The design condition for flexible risers may in the future be based on
waterfilled annulus. In that case corrosion fatigue may become a major design criterion.

Currently each supplier of flexible risers has its own proprietary (and confidential) set of SN
curves for fatigue design of armour wire. There are reasons to believe that the SN curves from the

T/prosjekt/70/700271/MT70 F04-048
 22

individual suppliers may differ significantly, even for in-air conditions. The suppliers have very
little data on fatigue strength in aqueous conditions containing H2S and/or CO2.

In a Joint Industry Project that is carried out by MARINTEK/SINTEF an extensive test

programme has been undertaken, to provide a basis for design of flexible risers with water-filled
annulus. In the programme, the following scenarios are studied:

- Leakage by damage of the outer polymer sheath, caused during installation or operation.
This would lead to sea water flooding of the annulus. Depending on the nature of the leak,
sea water in the annulus could be depleted of oxygen, or possibly saturated with air.
Efficiency of cathodic protection is uncertain. Sea water may be combined with hydrogen
sulphide (H2S) and/or carbon dioxide (CO2). There is also a possibility for Microbial
Induced Corrosion (MIC) as sulphide reducing bacteria (SRB) may develop in stagnant
seawater.

- Permeation of species from the product flow, notably water (H2O) which may condense and
accumulate in the annulus, in combination with gaseous components like H2S and/or CO2.

- Risers which have been subjected to sea water ingress may be repaired, flushed with
inhibitor, and re-installed, Taylor et al. (2002). Inhibitor fluid, possibly with some residual
sea water and with H2S and/or CO2 due to permeation from the well flow, could have an
effect on residual fatigue life.

SN curves for fatigue design are empirical, and need to be assessed on the basis of relevant data.
No standardised test methods are available for this type of testing. For this reason a test protocol
was worked out for the JIP, Berge (2002). In the protocol guidance is given on the following:

- specifications for the environment

- specimen preparation
- fatigue loading procedure
- data processing

Recommendations are given on treatment of mean stress effects in fatigue design, and a unified
method is proposed for assessment of SN design curves from test data.

All three suppliers of pipe are participants of the MARINTEK/SINTEF JIP. An anticipated
outcome of the JIP is a unified approach to fatigue design of armour wire. The JIP is planned to
run until 2005.

4.3 Polymer components

Typical characteristics for failures related to polymers can be classified as follows:
- Unexpected chemical exposure (humidity in bore fluid – Rilsan degradation
- Unexpected failure mechanism (gas pressure between layers of multilayer PVDF liner
- Unexpected mechanical loading (rough handling during installation, dropped objects,
fatigue of liner in end connection due to thermal cycling
- Failure of ancillary equipment (Blocking of vent ports)
- Unexpected material behaviour (pull-out in end termination)

Thus, the failures may be linked both to the design input, design and material knowledge as well
as handling of the pipe during installation and operation. Concerning robust material selection,
T/prosjekt/70/700271/MT70 F04-048
 23

actions and guidelines may be introduced in the design phase as well as in the design input phase
– i.e. estimation of operation conditions. Particularly important is that many of the failure
envelopes have started with the polymer materials.

The industry has already taken a step forward concerning degradation rate of Rilsan® by the
Rilsan User Group (RUG). It seems like the API-TR-17-RUG document has succeeded in
establishing a more reproducible procedure for life-time estimation of Rilsan®. Thus this work
has probably increased the safety of the liners although there have not yet been any failures of
flexible risers solely due to incorrect life time estimation. The procedure claims to be conservative
in the aspect that the criterion is applied on the inner wall, and thus ignoring the reduced ageing
rate of the outer wall due to temperature gradient and diffusion of water through the wall.
However, the document is not consequent in this issue because, elsewhere in the same document,
it is claimed that wall thickness has no effect on degradation rate. This should mean that the
diffusion rate of water is high compared with the aging (hydrolysis) rate, which is also found by
Jacques et al. (2002). There are also still topics that need further investigation; this includes
effects of organic acids, ethanol and, indeed, how to model cumulative exposure. Because the
procedure is based on a minimum CIV, it also applies a time based safety factor on the critical
CIV value that decreases with more severe environment – from 2 at pH = 6 to 1.6 at pH = 4.

Nevertheless, the impression is that the oil companies have little influence on the material
selection concerning polymer liners in the flexible risers. Given the operational conditions and the
chemistry of the produced fluid, it is left to the manufacturers (and a third party evaluator) to
document that a material is proper.

T/prosjekt/70/700271/MT70 F04-048
 24

5. ORGANISATIONAL FACTORS
The scope of the present project is mainly technical. However, material selection for flexible riser
design is also contingent on organisational factors. Some of these will be discussed briefly.

5.1 Choice of riser system

For the water depths and general design conditions on the Norwegian Continental Shelf, two riser
systems are feasible:

− Top tensioned steel risers

− Flexible risers

Top tensioned steel risers can be used with fixed platforms or tension-leg platforms. If a floating
platform – ship or semisubmersible – is chosen, there is no option other than flexible risers, due to
the dynamic response of the system.

When selecting a field concept, fixed versus floating production systems are evaluated very
thoroughly, being a major decision. In this evaluation, the riser system has generally been seen as
proven technology. Thus, the robustness and reliability of the riser system tend to be non-issues
in this phase. This relates to the general design as well as the selection of materials.

The outcome of such a process may be that the designer of the riser system is left with a decision
to use flexible risers, even if this is the less optimum riser system for a certain application. As a
result, flexible risers have been used in gradually more and more demanding applications, at a rate
that has not allowed accumulation of any significant service experience. The large number of
riser failures that has been experienced on the Norwegian Continental Shelf in recent years is an
indication that the application of flexible risers to more and more demanding applications may
have been too rapid.

The failure statistics given by Norsk Hydro shows that all failures related to materials were linked
to un-expected behaviour of the thermoplastic liner, in most cases linked to long term durability.
With more effort into long term and realistic testing, many of these failure modes could have been
predicted.

The lesson learnt is that research and product development should have a wider time horizon, with
more time and effort for the long term tests which are required for verification of service life and
robustness.

5.2 Competition vs. openness in the market

The API documents (API 17B and API 17J) have lead to a major improvement with regard to
commonality in design practices. However, the API documents do not always specify the design
criteria, like fatigue design SN curves, or temperature limitations for use of specific liner
materials. To quote from the Scope of API 17B,

“In general, flexible pipe is a custom-built product that can be designed and manufactured in a
variety of methods. It is not the intent of this document to discourage novel or new
developments in flexible pipe. On the contrary, it is recognized that a variety of designs and
methods of analysis are possible. For this reason, some topics are presented in general terms

T/prosjekt/70/700271/MT70 F04-048
 25

to provide guidance to the user while still leaving open the possibility of using alternative
approaches.”

For this reason, on many aspects of pipe design it is left to the supplier of pipe to establish the
specific criteria, and to provide the justification for these criteria. In this process the API
documents provide general advice only, and much of the evaluation is left to the supplier and a
possible third party.

The three suppliers of flexible risers are acting in a very competitive market. Technological
advancements in design and materials are kept as well protected secrets, and use of third party is
kept to a minimum. Even failure statistics and failure analysis reports which could contribute
towards improved reliability over-all, are shielded from insight. Due to the market forces,
operators seem to accept this situation.

As a result, there may be little commonality between the approaches taken by the three suppliers,
even for design problems that are very similar and not product specific. One example is fatigue
design criteria for tensile armour, which is a rather standard product made from Carbon steel. The
three suppliers have their own proprietary and confidential set of design criteria, which may vary
significantly for virtually the same type of component and material, in some cases even supplied
from the same group of vendors. It is difficult to find a rationale for this situation.

A similar description applies to the thermoplastic materials used for the liner. The three suppliers
run separate courses with regard to development and qualification of new materials, in some cases
involving patent rights. Some of the failure modes and corresponding design criteria are related to
complicated time-dependent material performance, like creep, thermal cycling and ageing, and to
the interaction between the thermoplastic liner and metallic components, that cannot be modelled
properly. Time is limited for accumulation of long term experience through service or full scale
testing. A more open research environment could obviously give a broader perspective, and
improved insight.

T/prosjekt/70/700271/MT70 F04-048
 26

6. RECOMMENDATIONS

6.1 General recommendations

− The riser system should be included as a primary element of a production system.

Reliability and robustness of the riser system should be considered at the stage of concept
evaluation.

− Flexible risers is a new technology, and the operational experience is too short to verify
reliability and robustness of risers in demanding applications. In this situation design must
be based on test data. However, accelerated tests may in many cases need to run for long
periods to be realistic. Research with a sufficient time horizon should be encouraged, to
verify design criteria related to time degradation processes.

− The experience that has been accumulated through the last few years has shown that even if
a flexible pipe design has been fully qualified, risers have failed due to new and unexpected
failure modes. This indicates that the behaviour of the materials and in particular the
interaction between the different materials in a pipe wall is not fully understood. It is
recommended that more basic and long term research is initiated to provide a better
understanding of these issues.

− The competition between the suppliers of flexible risers has lead to a situation where
confidentiality about product properties and design criteria has become an obstacle for
advancement of reliability and robustness. The suppliers should be encouraged to work
towards a larger degree of commonality in design practices and design criteria.

− Design of flexible risers involves advanced materials, interaction between very different
materials in a complicated structure, and time dependent degradation mechanisms. Third
party evaluation requires very advanced knowledge of materials and structure of a flexible
riser. There is a need for independent expertise, with the resources that are required.

6.2 Specific recommendations

− Fatigue design of armour wire has until very recently been based on SN curves obtained
from testing in air. Service experience has shown that a likely environment in the annulus
of a pipe will be aqueous, with H2S and/or CO2 permeating from the bore. Sea water
ingress, oxygen level and cathodic protection are additional environmental factors. Fatigue
design criteria for armour wire need to be reconsidered.

− Sealing and fixation of the liner in the end termination has proved to be a problem.
Improved end coupling design has appeared to alleviate the problem of liner pull-out.
However, with improved fixation, fatigue crack growth through the liner has become a
potential failure mode. Due to the short history of the improved design, no pipe design has
yet been proved through service history to have a 20 year design life. More work into the
mechanisms of sealing and fixation in the end termination should be undertaken.

T/prosjekt/70/700271/MT70 F04-048
 27

− Liner materials are subjected to long term deterioration mechanisms, like hydrolysis,
deplastification creep and ageing. In order to meet ever more challenging operational
conditions, new material grades of thermoplastics are being introduced at a rate which does
not allow accumulation of service experience. Thus, qualification is to a large extent based
on accelerated tests under simulated conditions. The methods used for qualification testing
and the design criteria should be evaluated critically.

T/prosjekt/70/700271/MT70 F04-048
 28

REFERENCES

API TR 17 RUG (2002). The ageing of PA11 in Flexible pipes (final draft for publication 21-10-
2002 SJG).

Berge S and Sævik, S (1993). Correlation between theoretical predictions and testing of two 4-
inch flexible pipes. Proc. Energy-Sources Technology Conference (ETCE), American Society
of Mechanical Engineers (ASME), Houston, USA.

Berge S and Eide OI (1999), Facility for thermal cycling of end terminations of flexible pipe.
Third European Conf. on Flexible Pipes, Umbilicals and Marien Cables – Materials Utilisation
for Cyclic and Thermal Loading. J. A. Witz (ed.), London, UK.

Berge S, Sævik S, Langhelle N, Holmås T and Eide OI (2001). Recent developments in

qualification and design of flexible risers. Int. Conf. Offshore Mechanics and Arctic
Engineering (OMAE), Rio de Janeiro, Brazil.

Berge, S (2002). Test protocol: Corrosion fatigue testing of armour wire for flexible risers.
MARINTEK Report MT70 F02-127 (confidential).

Berge S, Bendiksen E, Gudme J and Clements R (2003). Corrosion fatigue testing of flexible
riser armour – procedures for testing and assessment of design criteria. Int. Conf. Offshore
Mechanics and Arctic Engineering (OMAE), Cancun, Mexico.

Glomsaker T (2002). Calculation of permeation of gas into PVDF liner for Njord GI riser.
SINTEF MATEK Memo, 2002-05-03 (confidential).

Jacques B, Werth M, Merdas I, Thominette F and Verdu J (2002), Hydrolytic ageing of polyamide
11. 1. Hydrolysis kinetics in water. Polymer, 43, pp 6439 – 6447.

Kvernvold O (1992). Erosion-corrosion in inner steel carcass of flexible pipes. Int. Seminar
Flexible Pipe Technology, O. Olufsen (ed.), Trondheim, Norway, 1992.

MCS International (2001). State of the Art – Flexible Riser Integrity Issues, Study Report
prepared for UKOOA.

Melve B (2001). Principles for life time estimation of PVDF pressure barriers in high
temperature flexible pipes based on fracture mechanics. Int. Conf. Offshore Mechanics and
Arctic Engineering (OMAE), Rio de Janeiro, Brazil.

Olsen PG and Rongved K (2002). Operators experience with flexible risers. Int. Conf. Offshore
Mechanics and Arctic Engineering (OMAE), Oslo, Norway.

Ottøy MH, Finstad H, Mathiesen MW, Moursund B and Nakken T (2001). Field experiences on
Rilsan-ageing compared to data from ageing model. Int. Conf. Offshore Mechanics and Arctic
Egnineering (OMAE), Rio de Janeiro, Brazil

Rytter J and Rishøj N-J (2002). A novel compression armour concept for unbonded flexible pipes.
Offshore Technology Conference (OTC), OTC 14059, Houston, USA.

T/prosjekt/70/700271/MT70 F04-048
 29

Sævik S (1998). A new method for calculating stresses in flexible pipe tensile armours. Int. Conf.
Offshore Mechanics and Arctic Engineering (OMAE), Lisbon, Portugal.

Sævik S, Gray LJ and Phan A-V (2001). A method for calculating residual and transverse stress
effects in flexible pipe pressure spirals. Int. Conf. Offshore Mechanics and Arctic Engineering
(OMAE), Rio de Janeiro, Brazil.

Taylor TS, Joosten MW and Smith F (2002). Technical solutions applied for the treatment of
damaged dynamic risers. Int. Conf. Offshore Mechanics and Arctic Engineering (OMAE),
Oslo, Norway.

T/prosjekt/70/700271/MT70 F04-048
 30

RELEVANT TESTS

API 17B and API 17J list a large number of tests that are relevant for qualification of flexible
risers. Some of these tests are non-standard, and described in the API documents. Relevant
standards are given below.

RELEVANT DOCUMENTS:

API RP 17B, Recommended Practice for flexible pipe, American Petroleum Insitute (1998)

API Spec 17J, Specification for unbonded flexible pipe, American Petroleum Institute (1998).

API Std 1104, Welding of pipelines and related facilities.

ASTM A29, Specification for steel bars, carbon and alloy, hot-wrought and cold-finished –
general requirements.

ASTM A370, Test methods and definitions for mechanical testing of steel products.

ASTM D256, Test methods for impact resistance of plastics and electrical insulating materials.

ASTM D413, Test method for rubber property – Adhesion to flexible substrate.

ASTM D570, Test method for water absorption of plastics

ASTM D638, Test method for tensile properties of plastics.

ASTM D671, Test method for flexural fatigue of plastics by constant-amplitude-of-force.

ASTM D695, Test method for compressive properties of rigid plastics.

ASTM D746, Test method for brittleness temperature of plastics and elastomers by impact

ASTM D789, Test method for determination of relative viscosity, melting point, and moisture
content of polyamide (PA).

ASTM D1044, Test method for resistance of transparent plastics to surface abrasion.

ASTM D1238, Test method for flow rates of thermoplastics by extrusion plastometer.

ASTM D1242, Test method for resistance of plastic materials to abrasion.

ASTM D1525, Test method for Vicat softening temperature of plastics.

ASTM D1693, Test method for environmental stress-cracking of ethylene plastics.

ASTM D2143 Test method for cyclic pressure strength of reinforced, thermosetting plastic pipe.

T/prosjekt/70/700271/MT70 F04-048
 31

ASTM D2583, Test method for indentation hardness of rigid plastics by means of a Barcol
impressor.

ASTM D2924 Test method for external pressure resistance of reinforced, thermosetting plastic
pipe.

ASTM D2990, Test method for tensile, compressive, and flexural creep and creep rupture of
plastics.

ASTM D1141, Standard practice for the preparation of substitute ocean water, American Society
for Testing and Materials (1998).

ASTM E466, Standard practice for conducting force controlled constant amplitude axial fatigue
tests of metallic materials, American Society for Testing and Materials (1996).

ASTM E739, Standard practice for statistical analysis of linear or linearized stress-life (S-N) and
strain-life (ε-N) fatigue data, American Society for Testing and Materials (1991).

ASTM E831, Test method for linear thermal expansion of solid materials by thermomechanical
analysis.

Handbook on Design and Operation of Flexible Pipes, Berge. S. and Olufsen, A. (eds.), SINTEF
Report STF70 A92006, 1992.

ISO 8457-2, Steel wire rod, Part 2 – Quality requirements for unalloyed steel wire rods for
conversion to wire.

ISO/DIS 13628-2, Flexible pipe systems for subsea and marine applications.

NACE MR 01-75, Sulfide stress cracking resistance metallic materials for oilfield equipment,
National Association of Corrosion Engineers (1996).

NACE TM 01-77, Standard test method: Laboratory testing of metals for resistance to sulfide
stress cracking and stress corrosion cracking in H2S environments, National Association of
Corrosion Engineers (1996).

NACE TM 02-84, Standard test method: Evaluation of pipeline and pressure vessel steels for
resistance to Hydrogen-induced cracking, National Association of Corrosion Engineers (1996).

T/prosjekt/70/700271/MT70 F04-048
 32

APPENDIX A

Sources for this investigation:

The written sources, mainly found in the public domain literature, are listed in the report.

It has been very difficult to obtain any written information on pipe failures and failure modes from
operators and suppliers of flexible risers. Permission was given to use the table showing the
service experience of Norsk Hydro, Table 3.

A similar table from Statoil has not (yet) been released, due to unresolved discussions within
some of the licenses. Suppliers of pipe are involved in those discussions.

Representatives of Statoil and Norsk Hydro were interviewed. Both companies are commended
for giving very open and detailed information about practices and experiences with flexible risers.
The views that were presented in the interviews are reflected in the report. However, since the
information was given orally, the authors of the present report are not in a position to quote the
sources or to give specific references.

Representatives of all three suppliers of flexible risers were interviewed. The suppliers indicated
that they are prepared to answer general questions only, preferably on a questionnaire format. On
explicit questions it was made clear that very little information could be released on on-going
technical developments in materials technology. It was furthermore indicated that due to the
sensitivity of the issues, approval to release information would take much time. For this reason,
with the time allocation that was given for the project, it was not found feasible to pursue this
idea.

MARINTEK/SINTEF has been engaged in a variety of projects on flexible risers, also involving
materials technology. All of the work has been commissioned work with rather strict
confidentiality requirements. The experience from this work is reflected in the general
recommendations given in the report.

T/prosjekt/70/700271/MT70 F04-048
 2

TABLE OF CONTENTS

1 INTRODUCTION ................................................................................................................. 3
2 OBJECTIVES ........................................................................................................................ 3
3 MARTENSITTIC STAINLESS STEEL FOR PIPELINES..............................................3
3.1 Super martensitic alloy grades ......................................................................................... 3
3.2 Properties of super martensitic stainless steels.................................................................4
3.2.1 Tensile properties .............................................................................................. 4
3.2.2 Hardness ............................................................................................................ 5
3.2.3 Welding metal chemical composition ............................................................... 6
3.2.4 Corrosion properties.......................................................................................... 6
3.2.5 Fabrication, installation and operation of S13Cr pipelines ............................... 8
3.3 The historical background for selecting S13Cr line pipe materials ............................... 10
3.4 ”First users”.................................................................................................................... 11
3.5 Joint Industry Project initiative on SMSS, 1998 ............................................................ 12
3.6 SINTEF strategic funding .............................................................................................. 13
3.7 EU project JOTSUP .......................................................................................................13
3.8 Recent initiatives ............................................................................................................ 13
4 FAILURES OF SUB SEA PIPELINES .............................................................................14
4.1 Vital properties for design of pipelines .......................................................................... 14
4.1.1 Hydrogen embrittlement ................................................................................. 14
4.1.2 Local geometry and stress concentration ........................................................ 15
4.1.3 Welding and post weld heat treatment ............................................................ 15
4.1.4 Tensile properties and temperature ................................................................. 15
4.2 Subsea failures of pipelines on the Norwegian continental shelf................................... 16
4.2.1 Åsgard and Tune reeling incident and Tune hyperbaric repair weld failures . 16
4.2.2 Gullfaks failure................................................................................................ 17
4.2.3 Fracture at anode pads, Åsgard ....................................................................... 17
4.2.4 Failure of Åsgard Hub’s.................................................................................. 17
4.3 Failure caused by high temperature corrosion ............................................................... 18
4.4 Summary of the subsea failures...................................................................................... 19
5 UTILISATION OF RESEARCH IN THE OFFSHORE INDUSTRY ...........................19
5.1 Fracture toughness and high strength steels ................................................................... 19
5.2 Corrosion history and related R&D ............................................................................... 21
5.3 Environmental Assisted Cracking .................................................................................. 21
6 QUESTIONNAIRE TO THE OFFSHORE INDUSTRY ................................................23
6.1 Summary of the questionnaire........................................................................................ 23
6.1.1 Knowledge gaps .............................................................................................. 24
6.1.2 Increased robustness by improving the completeness of operational design.. 25
6.1.3 Technical management of pipeline construction projects and pipeline
operation......................................................................................................................... 26
7 DISCUSSION ....................................................................................................................... 26
8 REFERENCES.....................................................................................................................27
 3

1 INTRODUCTION
Over the last three years the Norwegian Petroleum Directorate (NPD) has expressed some
concern about the situation with respect to material selection and safety, triggered by recent
experiences in the North Sea. During the period 1996 to 2002, 220 gas leakages (>0.1kg/s) have
been reported on the Norwegian shelf /1/. Of the 220 gas leakages, 6 are connected to different
types of failures on subsea pipelines, all involving through pipe wall fractures. The consequences
of such failures are significant. The total cost (including the loss of income) of the five failures is
estimated to some 5-10·109 NOK. The impact on environment and human health and safety has
not been significant in these cases. Most of the 220 incidents are smaller topside gas leakages.
Luckily none have set on fire.

As background for mapping the development of risks with respect to failures/accidents within the
offshore industry, NPD has desired an investigation with respect to robust material selection
(RMS) on line pipe materials with emphasis on super martensitic stainless steels (S13Cr steels). A
parallel study is performed on flexible risers and the common denominators of these to offshore
applications are reported in the main report of the project. The basis for the selection was a
meeting with the Norwegian Petroleum Directorate after delivery of the pre-project report /2/.

2 OBJECTIVES
The objective of the present project is to summarize trends in material selection procedures in
larger oil & gas companies. The work is based on the use of existing knowledge in SINTEF and
NTNU and their industry network. The work will focus on a combination of industry experience
and research experience.The outcome of this project is to prepare for a RMS methodology.

3 MARTENSITTIC STAINLESS STEEL FOR PIPELINES

3.1 Super martensitic alloy grades

Common for all stainless steels is a minimum content of 11% Chromium. S13Cr steels have
typically 11-12% Chromium. The microstructure is austenitec at temperatures beyond ca 900°C.
Rapid cooling suppresses the formation of ferrite and martensite is mainly formed. With higher
contents of Chromium than 12%, Nickel (and/or Manganese) must be added to be able to form
martensite. Some remaining austenite and ferrite is typically present after cooling. The amount of
remaining austenite is part of controlling the strength of S13Cr line pipe materials. Increased
amount of remaining austenite reduces the strength. For the low strength S13Cr steels (yield
strengths between 500 and 600MPa) it is assumed approximately 15-25% remaining austenite.
Increased alloying with Molybdenum increases the corrosion resistance. The low content of
carbon is the main contributor for the weldability.

Super martensitic stainless steels are normally divided into three types /3, 4/:

1. Lean grade, 11Cr2Ni

2. Medium grade, 12Cr4.5Ni1.5Mo
3. High alloyed grade, 12Cr6Ni2.5Mo

Typical chemical composition for the three grades is outlined in Table 1. In the development of
this steel type, several alloy sequences have been designed to meet requirements to an increased
stress corrosion cracking resistance in the presence of H2S. In Table 2, major alloying sequences
to meet targeted corrosion resistance are summarized. In addition to these environmental factors,
 4

sufficient low temperature toughness is required. In order to meet toughness specifications the
ferrite content should be minimized. Hence, the ferrite forming elements such as Cr, Mo and Si
must be counterbalanced by austenite stabilisators. In practice, this means Ni since both C and N
levels should be lowest possible to maintain the optimum weldability through reduction of
hardness. As rough approximations, Schaeffler type of diagrams can be used, although these were
developed to predict the microstructure of weld metals. An example of this philosophy is the
following set of Cr and Ni equivalents (elements in wt%):

Creq = Cr + 1.37 Mo + 1.5 Si + 2 Nb + 3 Ti (1)

Nieq = Ni + 22 C + 14.2 N + 0.31 Mn + Cu (2)

When 13% Cr steels are selected, it should be noticed that service behavior also will strongly
depend on the consumables employed.

Table 1: Typical chemical composition of supermartensitic 13% Cr stainless steels /3/.

Alloy grade
Element 11Cr2Ni 12Cr4.5Ni1.5Mo 12Cr6.5Ni2.5Mo
(lean) (medium) (high alloyed)
C (max.%) 0.015 0.015 0.015
Mn (max.%) 2.0 2.0 2.0
P (max.%) 0.030 0.030 0.030
S (max.%) 0.002 0.002 0.002
Si (max.%) 0.4 0.4 0.4
Cu (max.%) 0.2-0.6 0.2-0.6 0.2-0.6
Ni (%) 1.5-2.5 1.5-2.5 1.5-2.5
Cr (%) 10.5-11.5 11.0-13.0 11.0-13.0
Mo (%) 0.1 1.0-2.0 2.0-3.0
N (max.%) 0.012 0.012 0.012

Table 2:Alloy design to meet target corrosion resistance /4/.

Alloy grade
Environmental
11Cr2Ni 12Cr4.5Ni1.5Mo 12Cr6.5Ni2.5Mo
parameters
(lean) (medium) (high alloyed)
T 20-100°C 20-100°C 20-100°C
P (CO2) 10 bar 20 bar 20 bar
P (H2S) - 0.005 bar 0.050 bar
pH 3.5-4.5 3.5-4.5 3.5-4.5
Cl− 600-100,000 ppm 600-100,000 ppm 600-100,000 ppm

3.2 Properties of super martensitic stainless steels

3.2.1 Tensile properties

550MPa is the typical minimum specified yield strength (SMYS) of S13Cr line pipe steels.
Experiences at SINTEF show that the yield strengths normally are between 600 and 730MPa.
However, SMYS for the TUNE field was 640 MPa with a specified maximum of 760 MPa. The
relatively large scatter is probably a result of the manufacturing process (i.e. thermo mechanical
history). The results indicate that increased wall thickness normally is connected to reduced
 5

strength. This may be connected to a lower cooling rate of heavy walled pipes, giving less
martensite.

For S13Cr line pipe steels it is observed that the yield strength increases with increasing
temperature /10/. From room temperature to approximately 150°C the yield strength increases 50
to 90MPa. The larger increase is found on low strength S13Cr, which indicates that increased
amount of remaining austenite is connected with increased ability for martensite transformation.
The transformation is believed to be induced by deformation and that increased test temperature
increases the degree of transformation. This assumption is based on the observation of reversal to
original tensile properties after heating and cooling. Testing between 150 and 200°C shows that
the yield strength stabilizes. Above 200°C the tensile properties are reduced. It should also be
mentioned that the tensile strength is reduced with almost the same magnitude as the yield
strength increases. This will clearly affect the ability of strain hardening at elevated temperatures.

3.2.2 Hardness
Prior to the large qualification work of the 13Cr steel class in the late 90-ties, a preliminary
programme was carried out to study the relationship between the weld thermal programme
(cooling rate, peak temperature, subsequent tempering) and heat affected zone (HAZ) mechanical
properties (Charpy V notch toughness, tensile strength and ductility, hardness), using the weld
simulation technique. Here, a wide range of martensitic steels was included, ranging from the
“older” class with relatively high carbon content (~ 0.1% C) and no alloying with exception of Cr,
to the new so-called “supermartensitic” grades with improved weldability, based upon low carbon
(<0.02%) content and balanced additions of alloying elements such as Ni and Mo to improve the
corrosion properties and to maintain the microstructure balance.

This work was kind of a pioneer study, and was reported already in 1995 /11/. The results
obtained showed that the “old, traditional” 13Cr steels contain too high carbon content, which
give a very high hardness level of about 500kg/mm2. This is far beyond an acceptable level, even
when applying tempering. With a hard martensite present, the steel will be very susceptible to
hydrogen cracking, and expensive operations like preheating and postheating may be necessary,
together with certain precautions regarding welding consumables and shielding gas moisture must
be taken. In addition to the very high hardness, high carbon steels will inevitably give low
Charpy-V impact values in the as welded condition. It was also shown that toughness varied
substantially with the chemical composition of the steel. The toughness seemed to be impossible
to predict from chemical composition only, whether it is high, medium or lean alloy. On the other
hand, this early work showed that the toughness is nearly independent of the weld cooling rate
(∆t8/5). Finally, it is interesting to note that the alloys with the best toughness had a very similar
composition as those steels now being commercially available.

One important observation is that it is possible to calculate the Ms temperature. In practice this is
very important, since it may control the phase transformation start, and hence, the tome available
for hydrogen diffusion during welding, and build up of weld residual stresses. This equation is as
follows (discussed in ref./12/):

Ms (°C) = 539 – 423 (%C) – 30.4 (%Mn) – 12.1 (%Cr) – 17.7 (%Ni) – 7.5 (%Mo) (3)

According to NACE MR-0175 the hardness requirement for weldable martensitic stainless
pipeline steels is set to HRC=22 (HV10=249) for sour service applications (i.e. H2S partial
pressure >3.5 mbar). This is practical impossible without reducing the strength considerably.
Normal thermo mechanical procedures applied for S13Cr steel pipes give a hardness in the range
290-320 HV10 in the base material. Even for a so-called low strength material (YS=520MPa)
 6

tested for the Tangguh development in Indonesia, the hardness was in the range 270-280 HV10 in
the base material /10/. In the heat affected zone the hardness is increased to 340-360 HV10.
Applying post weld heat treatment for 5 minutes at 620-650 °C has shown to be beneficial and
decreases the hardness of the heat affected zone by 10-20 Vickers.

Today the martensitic steels are welded using duplex/superduplex consumables. To avoid
precipitation of sigma phases, reducing the mechanical and corrosion properties of the weld metal,
a longer post weld heating time is not recommended. However, going for matching consumables
this may be an option.

3.2.3 Welding metal chemical composition

Girth welding of supermartensitic 13% Cr steels is frequently performed with mechanised
GMAW process or manual GTAW, frequently with superduplex wire /15/. The weld metal
strength has been shown to be very sensitive to the microstructural phase balance, the yield
strength being fully controlled by minor amounts of soft austenitic and/or ferritic constituents /16/.
The impact properties of weld metals are strongly dependent on the oxygen content. A sharp drop
in toughness, from about 160 to 60 J, has seen when the oxygen content increased from 150 to
250 ppm. In practice, the properties of pipeline girth welds will thus be influenced by the welding
process, the shielding gas, the welding consumables and the welding performance, see Fig.1,
which contain data from GMAW, GTAW and SAW.

3.2.4 Corrosion properties

Since the new weldable martensitic stainless steels were introduced in the mid 90'ties, the
understanding of potential degradation mechanisms has increased their application limits. Also
the way of testing and qualification has slightly changed with the new knowledge. Findings have
also given input to revising standards as European Federation of Corrosion standard 17 specifying
testing of corrosion resistant alloys and also input to change of company specifications. It has also
had valuable input to revising the NACE TM 0177 standard and given input for the new ISO
standard.

Corrosion degradation mechanisms that have had special focus and may limit the use of these

• General corrosion at low pH's

materials are:

• Sulphide stress corrosion (SSC)

• Stress corrosion cracking (SCC) and high temperature intergranular corrosion (HTIC)
• Hydrogen embrittlement (HE) or hydrogen induced stress cracking (HISC) and its impact
on fracture mechanics behavior.

These environmental assisted mechanisms are briefly discussed in the following

3.2.4.1 General corrosion at low pH

In the first qualification program for the Gullfaks satellites and the Åsgard field, SINTEF reported
high general corrosion for all samples tested for sulphide stress corrosion (room temperature, both
on simulated condensed water and simulated formation water). This resulted in large discussions
and an enlarged test program was executed in several labs. Early in the discussions SINTEF
explained this type of corrosion by the use of Pourbaix diagram showing that these materials could
turn active under a certain pH. Later research has shown that this pH limit is in the range 3.7-4.0.
However, it is also discussed by Drugli et. al. that this corrosion will depend on the buffer capacity
in the solution, temperature, H2S and flow conditions /6/. At low buffer capacity solutions as
condensed water and moderate flow rates acceptable pH will be lower due to increased pH at the
 7

surface during service, caused by the corrosion reactions. It is also shown that the passivity is
improved at elevated temperature.

3.2.4.2 Sulphide stress corrosion (SSC)

Welded S13Cr steel, type 12-13Cr/4-4.5Ni/1-1.6Mo, was tested with constant load testing (load
~90% of YS) in a buffered 5% NaCl solution with start at pH 3.5. The results were drawn from
720 hours testing. All welds failed in the buffered solution at pH 3.5 with 0.01 bar H2S and the
one test with 0.004 bar H2S. At pH 5, a forge weld did not fail at 0.01 bar H2S while a TIG weld
failed at 0.1 bar H2S. The results indicated that the tested alloy-range has threshold values of H2S
between 0.1 and 0.01 bar H2S at pH 5 in a 5% NaCl solution and below 0.004 bar H2S at pH 3.5.
The threshold values will also depend on the chloride content, partial pressure of H2S and pH
besides of the alloying of the material. Published data for higher grade S13Cr steels type 13Cr/4-
6Ni/0.7-1.5Mo and 13Cr/4-6Ni/2-2.5Mo have been gathered and the data indicate threshold
values of 0.001-0.01 bar H2S at pH 3-3.5 and 0.01-0.03 bar H2S at pH 4.5-5 for both materials.
Most data are for base materials, only a few for weldments. Different test methods are involved as
slow strain rate, 4-point bend and constant load.

A martensitic consumable with matching strength to the base material is of most interest.
Consumables have been developed the last years and data have been published by SINTEF during
the involvement in an EU project /7/. The results show similar values as indicated above.

3.2.4.3 Stress corrosion cracking (SCC) and high temperature intergranular corrosion
(HTIG)
Multipass welding has shown to sensitise the martensitic structure by precipitations along the
prior austenite grain boundaries in the coarse grained heat affected zone /8/9/10/13/. These
precipitations have been shown by Ladanova et.al. to consist of CrC (Cr23C6) /14/. Adjacent to
these precipitates a narrow zone with depleted Cr occurs. The depletion zone is sensitive to
corrosion. It is also found that these precipitates are mostly of TiC type for the Ti alloyed S13Cr
steels. A short post weld heat treatment procedure at 650 °C for 5 minutes has shown to be
beneficial and a healing of the depletion zone is shown to take place. Intergranular corrosion has
not occurred on post weld heat treated samples. A time aspect in testing is highlighted especially
after the NAM failures of lean grade materials, see section 4.3. The NAM failures show that
corrosion of depletion zones may take time to initiate and the corrosion rate will be dependent on
the test conditions. No systematic study has been undertaken to define the borderline conditions
for this type of corrosion. The corrosion will depend on pH, buffer capacity, chloride level and
temperature besides of the material quality (i.e. alloying).

3.2.4.4 Hydrogen embrittlement (HE) or hydrogen induced stress cracking (HISC) and its
impact on fracture mechanics behavior
In air, S13Cr and duplex steels have a CTOD (Crack Tip Opening Displacement) fracture
toughness of about 0.7-1mm. Through fracture toughness characterizations on a high grade S13Cr
steel, performed in the Fram Vest project, it was demonstrated a drop of CTOD to <0.02mm
under normal cathodic protection (i.e. -1050mV SCE) at 4°C /17/. The cracking mechanism is
stepwise brittle fracture. This has been further verified in an ongoing Joint Industry Project. The
same mechanism is observed on super duplex and duplex steels, which drops to a CTOD value
<0.03mm. These effects were unknown in 1995.

In standard tensile specimen slow strain rate testing, several factors for the ratio (area reduction or
time to failure) between environmental testing and testing in air, ranging from 0.5 to 0.9, have
been suggested. However, for typical S13Cr steels the ratio ends up with values <0.1. The
industry has argued that this method with slow deformation until fracture, is giving a high degree
 8

of plasticity and unrealistic stresses. Therefore, they argue, that the slow strain rate method is not
considered proper to evaluate the materials in specific environments. Other methods as constant
load testing and 4 point bend were considered more appropriate with a given load/stress/strain and
the results were more understandable and less conservative. These effects were known in late in
1995 and published by SINTEF in 1997 /18/.

In fracture mechanics tests, a very sharp defect is simulated by a pre-fatigue crack. The stress
intensity at the crack tip during loading represents high local deformation and builds up of
hydrostatic pressure (i.e. stress tri axiality and plain strain). Large deformations are also reflected
in slow strain rate tests of smooth tensile specimens. The fracture toughness is reduced by a factor
of >35 times. This is even more than what found in a simpler test as slow strain rate (10-20 times
reduction in the ductility). However, both test methods show that the material is extremely
sensitive to hydrogen induced cracking during normal cathodic protection and high local stress
and strain. Hydrogen embrittlement is increasing at low sea water temperatures.

3.2.5 Fabrication, installation and operation of S13Cr pipelines

For the girth welding of pipes, welding procedure qualifications regarding mechanical and
corrosion properties are well specified. More often the offshore standard OS F-101 (DnV 2000)
makes the basis for the extent of the qualifications. Testing is performed on as welded test
specimens and on specimens that are deformed to simulate the installation process. The latter is
also artificial aged to simulate aging of line pipe material during operation. Testing of deformed
and aged material is related to operational issues.

For installation purposes the none deformed and aged condition (i.e. as received or "as welded"
condition) is evaluated to match the actual properties. However, the deformation during the
different stages of an installation process implies variations of the mechanical properties. This will
affect the evaluation of where in the installation process criticality regarding maximum defect size
is obtained. This is also linked to the circumferential position on the pipe, as the mechanical
properties depend on the deformation history. The fracture mechanics test matrix would have been
too extensive to be accepted by the industry if all property states should have been fully accounted
for in an engineering critical assessment analysis. A simple and less time consuming approach is
to simulate the different stages of installation by tensile tests. The tensile stress-strain curve of a
material makes the basis for the failure assessment diagram, which limits the accepted range of
maximum allowable defects. Experience gained through engineering critical assessments of
pipeline construction projects has been a major approach to tailor-make engineering critical
assessment analyses for safe utilization of pipelines.

Deformed and aged material has shown to be more sensitive to brittle fracture compared to the as
welded condition. The sensitivity is dependent on the strain amplitude of the deformation cycle,
the number of deformation cycles (i.e. accumulated strain) and the aging. The strain amplitude is
normally based on calculations considering the pipe dimensions and bending curvatures during
installation. The number of deformation cycles is dependent on the installation method and
considerations of reversed installation for repairs etc. The sensitivity to brittle facture is often
related to the properties of the weld deposit and the fusion boundary. Carbon steels are more
sensitive than S13Cr steels and duplex steels. This is believed to be connected to the increased
marking of the upper yield strength after deformation and aging of carbon steels. All steels show
elevated yield strength, reduced strain hardening (less Rm/Re ratio) and some reduction of the
ductility after cyclic deformation and aging.

The installation process (c.f. J-laying and reeling) apply roughly the opposite deformation of the
12 o'clock and 6 o'clock positions. The 12 o'clock positions complete the deformation cycles
typically in a compressive state and the 6 o'clock position typically in tension. For input to a
 9

fracture mechanics analyses, the properties of the 6 o'clock position has shown to be most critical.
Evaluations of criticality related to deformation history and artificial aging are lacking in todays
specifications and the knowledge on deformation and aging is not fully understood and quantified.
Traditionally the properties of the 12 o'clock position are used.

Eventual weld defects are detected and measured by means of automatic ultrasonic testing during
fabrication. The lower bound limitation of the ultrasonic equipment used is normally a defect
height of 2mm (accounting for the uncertainty of the test method). The equipment can detect
much smaller defects, but the accuracy becomes quite limited. This type of equipment has shown
to increase the level of detections and increase the applicability (as it measures/quantifies defect
size) for girth welding of pipes compared to other methods. Normally girth weld defects are
extended in the circumferential direction. Typical defects are interpass lack of fusion and
extended pores. The engineering critical assessment is performed to predict the maximum
allowable defect during installation. Due to the resolution of the ultrasonic equipment used today,
the maximum allowable defect size must normally be larger than 2mm height. This requirement is
normally associated with a 50mm length of the defect. For girth welded line pipe steels, with
ductile behavior during the fracture mechanics testing, the experience is that the maximum
allowable defect height ranges between 3 and 5mm (x 50mm length).

The JIP guideline is established for installation methods introducing plastic deformation.
Operational issues are not fully accounted for. As mentioned in section 3.2.4.4, the fracture
toughness is significantly reduced under normal cathodic protection due to hydrogen
embrittlement. Recent research in the ongoing JIP indicates the same effect on crack growth rates
during fatigue loadings. This has clearly impact on the operational limitations and the prediction
of lifetime. Moreover it indicates that the criticality to maximum allowable weld defects size
could be moved from the installation process to the operational static and fatigue loadings.

3.2.5.1 Operational engineering critical assessment

In the following an example of operational engineering critical assessment is performed to show
the interaction of design stress and operational limits.

A 12 inch steel pipeline with wall thickness 15mm has been qualified for the J-laying installation
method. In other words; the maximum allowable defect size is larger than 2x50mm. How is the
integrity of the pipeline during operational conditions, including hydrogen embrittlement from
normal cathodic protection?

A normal design criterion for the offshore industry is to utilize the line pipe materials up to 70-
80% of the yield strength during operation. This could be related to the specified minimum yield
strength (SMYS) or the actual yield strength (YS). Both approaches have been used by the
industry. In this case the minimum specified yield strength is used (i.e. most conservative) and a
level of 75% utilization.

It is presumed that the maximum allowable defect size for all cases is the lower limit (2x50mm),
which is the most conservative starter crack for operation. During installation ductile crack growth
extends the height of the defect with 1mm to a total defect size of 3x50mm. Possible interactions
with other minor defects are estimated to add to an accumulated defect size of 3x60mm.
Operational engineering critical assessments have been performed for X65 carbon steel (SMYS
450MPa), S13Cr steel (SMYS 550MPa) and duplex steel (SMYS 450MPa) with the dimensions
as above. Residual stress from welding is accounted for by means of the relaxation model of
BS7910. Stress concentrations from the weld toe and eventual pipe misalignments are not
accounted for in the analysis. The fracture toughness of duplex steel and S13Cr steel are extracted
from tests performed under normal cathodic protection at 1 atm pressure/4°C at SINTEF.
 10

Relevant results on X65 carbon steel are found in literature /19/. The results are presented in table
Table 3.
Table 3: Results of the engineering critical assessments
Material SMYS Max. Actual CTOD Defect Critical
design YS fracture size membrane
stress toughness a x 2c stress
[MPa] [MPa] [MPa] [mm] [mm] [MPa]
X65 carbon steel, BM 450 338 556 0.7 3 x 60 564
S13Cr steel, welded 550 413 648 0.02 3 x 60 371
Duplex steel, BM 450 338 630 0.03 3 x 60 419

The environmental impact on the carbon steel is very limited and the operational critical
membrane stress is above the actual yield strength and well beyond the maximum design stress.
The duplex steel is also considered safe in the assessment, but the safety of duplex steel is clearly
limited compared to the X65 carbon steel. If the actual yield strength of the super duplex material
was used as a design basis, the maximum design stress would have extended well beyond the
calculated critical membrane stress. Such case is considered unsafe. For the S13Cr line pipe
material the maximum design stress is clearly outside the range for safe operation. The critical
membrane stress (i.e. 371MPa in Table 3) correlates well with results obtained on single edge
notch tensile fracture mechanics tests under normal cathodic protection. A critical net section
stress of 370MPa was found for initiation of hydrogen induced stepwise brittle fracture /21/.

The performed analyses establish the critical stress level for fracture of the pipe during static
loading. The maximum fatigue stress is lower than the stresses presented in Table 3. The fatigue
crack growth rate is vital input for operational critical assessments. Such data are not published
yet, but results are produced in an ongoing Joint Industry Project (see section 3.8).

3.3 The historical background for selecting S13Cr line pipe materials
The last two decades have faced an increasing application of corrosion resistant alloys for pipings,
pumps and valves in the production of oil and gas. Expensive high alloyed corrosion resistance
alloys as the duplexes have been applied to a large extent, as an alternative to C-steel and
inhibitors. In the search for less expensive solutions for field developments in the oil production
industry, the oil companies are looking for cheaper materials with satisfactory mechanical strength
and corrosion resistance. Delivery capacity was also a factor that was considered. In recent years,
steel suppliers have developed new martensitic stainless steels with 12-13 % Cr and small
amounts of Ni (5-6%) and Mo (1.5-2.5%). These steels are today more or less referred to as Super
13Cr or weldable (low carbon) martensitic stainless steels. These have significant increased
corrosion resistance in sweet service compared to traditional martensitic grades, and satisfactory
welding procedures have been developed. Extensive research has been conducted around the
world to establish a better basis for evaluating their application for transporting unprocessed oil
and gas.

Cost benefit
Stainless steels are used extensively in the oil- and gas industry because of the hostile
environmental conditions. The steels are mainly used in piping systems and especially flowlines.
Candidates are in many cases a choice between duplex stainless steels, carbon steels and S13Cr
steels. Selection of carbon steels requires the use of inhibitor.

Duplex steels have excellent corrosion properties, but are rather expensive. The potential for
savings can be illustrated by a simple example. The price relation between duplex and S13Cr is
 11

approximately 3:2. The yield strength of duplex is within the range 400-450MPa and 600-700MPa
for S13Cr. The utilisation of the material is often related to the yield strength, and the total
potential for cost reduction is 50-60% compared to duplex line pipe materials.

The increased strength and good fracture toughness of S13Cr are relevant for installation methods
involving larger plastic strains as reeling. Reeling is a widely used method today and represents
cost and time benefits compared to traditional fabrication and installation techniques.

Development of new fields

The savings by use of S13Cr have contributed to more cost effective construction and installation
compared to duplex line pipe materials. The operational savings are significant compared to
carbon steel line pipes as well, if the costs related to the S13Cr pipeline failures are excluded. This
is mainly caused by the removal of the inhibitor. These savings can be vital information when
extensions of the number of fields that can be developed are considered. This is especially
important today as the remaining fields on the Norwegian shelf are many in numbers, but are
significantly limited in volume compared to the larger productions fields. It is also important that
the Norwegian oil industry is in the front line in cost effective development and maintenance in
order to be a preferred partner in the development of oil/gas projects on an international basis.

Environmental aspects
The higher strength of S13Cr compared to duplex steels will reduce the material consumption
with 25-35%. This development will be positive from an environmental point of view, first of all
because the consumption of steel is reduced, but also because the reduced weight makes it easier
to transport and handle (e.g. pre-installation, installation and removal after final production/
lifecycle).

Inhibitors are frequently used in pipelines made of carbon steels to control the corrosion
behaviour. The use of inhibitors is, however, considered negative from an environmental point of
view. By substituting carbon steels with stainless steels the pollution factor is removed.

3.4 ”First users”

Hydro took the first initiatives on designing S13Cr as pipeline material for the mild sour and
sweet environments, typical for the Norwegian shelf. In 1992 Hydro started a co-operation with
Kawasaki Steel on optimising chemical composition and to develop welding procedures. Testing
of corrosion resistance and mechanical properties was undertaken. Statoil took part in the
development of S13Cr steels in 1994. Prior to the final decision to select weldable S13Cr for
subsea pipelines, Statoil run a large programme to study the mechanical and corrosion properties
of different weldable S13Cr steels. The programme was performed in the period 1994-1996.
However, pipes and field welding were not available and heat affected zones were produced by
weld simulation. The process converged to weldable low-carbon S13Cr steels of the same type
used today, alloyed with Molybdenum and Nickel. See section 3.2 for typical alloying content and
properties.

First user was Statoil on Gullfaks Sør and the Åsgard field, using combination of duplex and
super duplex filler material, based on prior experience of the fabricator Coflexip Stena Offshore
Norway (today Technip Offshore Norway).

The properties of the girth weld and adjacent pipe heat affected zone was focused as defects are
normally concentrated in the weld and at the fusion line. The fusion line is in conjunction with the
heat affected zone and the microstructure of the heat affected zone is normally most susceptible
for corrosion and embrittlement mechanisms. The first laboratory tests on sulphide stress
corrosion revealed that in simulated condensed water, all samples suffered from adhanced general
 12

corrosion of the base material when tested according to existing specifications like EFC 17 and
NACE TM-0177. The specifications at that time implied a strong buffer of NaAc (4g/l) and a pH
of 3.5. More testing at Statoil, SINTEF and manufacturer showed the same results. More basic
research studies at SINTEF, funded by Statoil, showed that this phenomenon was related to the
strong buffer, which maintained a low pH at the surface. The electrochemical reaction will
normally result in higher pH in low or non buffered solutions. The specification has been changed
and a diluted buffer is applied today using 0.4 g/l NaAc /6/. See also chapter 3.1.4.1.

The pre-qualification and qualification programs on Åsgard, using test methods from which
experience had been gained on carbon steels, was promising. After the buffer adjustments
mentioned above, the results showed acceptable corrosion resistance to relevant environments.
The fracture toughness was improved compared to carbon linepipe steels, especially the risk of
brittle fractures at low operational temperatures. In addition, the strength was significantly
improved.

After final adjustments on welding process and -wire (see section 4.2), the Åsgard pipeline project
seemed to be a straight forward qualification programme. However, internal research at SINTEF
indicated significant reduced ductility of S13Cr under conditions of normal cathodic protection.
See section 5.2 and 3.2.4.4 for more details.

3.5 Joint Industry Project initiative on SMSS, 1998

S13Cr as pipeline material was installed on the Norwegian shelf in 1997/1998 on several projects.
The results were considered positive on both a technical and economical basis. The projects had
international focus and they gave the Norwegian offshore industry a positive promotion. A
preproject was initiatied and financed through Stålmat Forum (forum for Norwegian steel users
and fabricators /21/) with the purpose to establish a programme for further research and
optimisation of S13Cr as a candidate for subsea pipelines. Stålmat Forum is sponsored by the
member companies and the Norwegian Research Council through the VARP R&D programme.

• Develop fabrication procedures where the welding process is optimized for the actual alloys
At that time, the most important tasks for the S13Cr research progamme were /22/:

• Develop acceptance criteria for mechanical properties and defect tolerances for pipes under

• Determine the corrosion properties in welds, HAZ and base material

pipelaying and service conditions

• Extend the application range of S13Cr materials

The proposed programme was a combination of applied and fundamental research with a time
schedule of 4 years and a budget of 42 MNOK. 60% was intended financed by the industry (e.g.
O&G companies, fabricators, consumable manufacturers, pipe manufacturers) and the remaining
by the Norwegian Research Council. The Norwegian Petroleum Directorate emphasised the
importance of the proposed S13Cr research programme and recommended the project for the
Nowegian Research Counsil. However, the programme was not supported by the Norwegian
Research Council due to low prioriety in materials technology and low budgets. The industry
became reluctant and the JIP initiative terminated.

Extensive pre-qualification and qualification programmes, based on traditional test and analysis
techniques, and limited project R&D have partly compensated the planned investigations in the
JIP, but the fundamental and systematic scientific approach is limited. The results are
implemented in company- and construction project specifications, leaving especially three topics

• few implementations of results in national or international specifications and standards

out:
 13

• limited systematic testing, comparison and evaluation of matching S13Cr welding

•
consumables and duplex/super duplex consumables
limited research on fundamental understanding of degradation mechanisms (3 Ph. D.
candidates were proposed in the JIP)

Investigations on fractured pipelines on the Norwegian shelf (see chapter 4.2) have exposed
scientific areas with fundamental lack of knowledge. Through the failure investigations, it has for
example been documented that hydrogen embrittlement can be very critical. Field experience and
“first users” are vital for fully exposing critical parameters from pipe manufacturing, fabrication,
installation and operational conditions.

Even though the proposed S13Cr research programme contained the subtask “Hydrogen assisted
cracking”, it is not evident that all the relevant mechanisms and sources of hydrogen would have
been exposed as critical through the proposed programme.

3.6 SINTEF strategic funding

Based on alarming results from slow strain rate testing on smooth S13Cr tensile specimens in
1995-1996, SINTEF Materials Technology decided in year 2000 to invest in new equipment for
fracture mechanics testing under environmental impact. This was financed through internal
strategic funding and included a R&D programme with a total of 2.5-3.0 MNOK. The results from
this R&D programme is highlighted towards oil industry, engineering companis and steel
manufacturer and and published /18/. The results has been discussed in several fora as the EFC
working party on O&G.

3.7 EU project JOTSUP

The JOTSUP project started in January 2000 and was terminated in January 2003. In this project
focus was put on high production welding and little on corrosion. From Norway Statoil and
SINTEF were partners with a budget of 550-600 KEURO each and 50% was funded by EU and
the 5th framework. The other 50% was sponsored by Statoil and SINTEF themselves. Focus here
was put on materials developed in Europe (Industeel) and new matching consumables developed
by ESAB and Bøhler Thyssen. Corrosion verification was limited to sulphide stress corrosion
limits undertaken at SINTEF /7/.

However, one important finding in the JOTSUP project was that hydrogen may strongly influence
the fracture stress and ductility /19/. Here, matching consumables were used in terms of a
Thermanit 13/06 Mo MIG wire from Bøhler Thyssen and a metal cored OK Tubrod 15.55 wire
from ESAB. The results showed the fracture stress was remarkably reduced as the weld metal
hydrogen content increased, i.e., from 800-900 MPa for 4 ppm to 400-500 MPa for 15 ppm
hydrogen. Similar reduction in fracture ductility was observed. For low weld metal hydrogen
content, the fracture appearances were ductile with typical dimples. At high hydrogen
concentration, th fracture surfaces consisted of dimples and cleavage facets (fish eyes). The fish-
eyes observed had diameters up to 80 µm, and represent a strong indication of hydrogen
embrittlement. By performing heat treatment (225ºC for 24 hours) of the weld metal deposited
with full moist shielding gas, with subsequent tensile testing, an increase in the fracture stress
from 300 MPa to 1050 MPa was obtained. This means that the weld properties may be restored by
the use of hydrogen diffusion treatment.

3.8 Recent initiatives

In 2003 a Joint Industry Project was initiated on hydrogen embrittlement and pipeline integrity.
The project is supported by Hydro and STATOIL and is executed as a collaboration between
DNV (Det Norske Veritas) and SINTEF. The experiences from previous R&D projects and failure
 14

investigations are summarised /25/. The basis for a wider international Joint Industry Project is
established through this work. The content of the future project has focus on closing knowledge

• develop test methodology within the area of environmental assisted fracture mechanics
gaps by:

• use of the new test methods to establish critical limits for actual pipeline materials as S13Cr

• optimise material specification and utilisation

steels, duplex steels and carbons steels under operational conditions

• optimise welding procedures and evaluate matching consumables for S13Cr

• addressing fundamentals related to mechanisms of hydrogen embrittlement for pipeline
materials as a major scientific task (combined research institute work and Ph.D. student
programs)

The content of the project has most probably international interest. This assumption is made on a
basis that not only S13Cr steels have shown significant reduced properties under cathodic
protection, but duplex steels as well. Duplex steels as line pipe materials are from most operators
looked at as a robust and safe selection. This situation is now changed. It is also known that
hydrogen has impact on high strength carbon steels. The proposed programme have the intensions
to study carbon steel grades representing line pipe grades from X80 to X110.

In Norsk Hydro's development of the Ormen Lange field, it is decided to use duplex/superduplex
for the manifold and pipes from manifold and carbon steel mainline. The recent experiences have
lead to an initiative from Norsk Hydro to verify the robustness of these materials with respect to
hydrogen induced cracking. A project has been started in 2004 executed by SINTEF and DNV in
cooperation.

4 FAILURES OF SUB SEA PIPELINES

4.1 Vital properties for design of pipelines

4.1.1 Hydrogen embrittlement

Hydrogen embrittlement is normally associated with cathodic protection in conjunction with
welds, more specific heat affected zones. The weld toe, or similar surface offsets, will act like a
stress concentration area and the microstructure of the heat affected zone is likely more sensitive.
In addition, the residual stress from welding adds on the local stress level. Cracking usually
initiates at the fusion line or close to the fusion line in the heat affected zone. Cracking or failures

• hydrogen from the welding wire or moisture in the shielding gas or condensation on the
have been documented to be caused either by

• by hydrogen uptake and diffusion from cathodic protection.

pipe/groove surface

Based on diffusion measurements, the critical surface concentration of hydrogen from cathodic
protection seems to be between 1.3ppm and 9ppm, corresponding to hydrogen obtained at -
900mV and -1050mV SCE respectively. Long duration tests have shown that the hydrogen level
in martensite can build up to 10-15ppm when saturated, which is far above the critical level
needed to embrittle typical S13Cr line pipe materials. The critical local hydrogen concentration
where cracking takes place may be much higher. As the areas that are susceptible to cracking are
local, an exact value of the hydrogen level is very difficult to measure. Through diffusion,
hydrogen may also be increased at the fusion line and adjacent heat affected zone due to the high
content of hydrogen that can be dissolved in the duplex/superduplex weld metal.
 15

4.1.2 Local geometry and stress concentration

Critical stress concentration is the second important factor. Global stresses well below the
material yield strength could, because of local geometry, increase beyond yield and apply local
plastic deformation. Typical geometries that affect the local stress are:
-material strength mismatch
-pipe misalignment
-counterbore
-surface irregularities/defects
-transition at weld toe
-transitions to other pipe dimensions
-attachments on the pipe like anode pads

Operational shut downs, shut ins, pressure testing (RFO), tie in, vibrations, free spans and tides
will add on with fatigue on the pipeline. A critical event is most probably obtained when the local
strain breaks the chromium oxide by repeated loadings or by increased accumulated deformation.
The breaking of the chromium oxide will lead to a significant increase of local hydrogen uptake.
Cathodic protection at a potential of -1050 to -1100 mV SCE (normal sacrificial anode potential),
polarize the corrosion resistant alloys to a potential were the materials are in the immune area.
That means that surface oxides that are removed will not be reformed.

4.1.3 Welding and post weld heat treatment

Few comparable tests have been performed comparing S13Cr weld metal and duplex weld metal
on S13Cr pipeline. A few slow strain rate tests on round sub sized cross weld tensile specimens
have been performed and the results showed that the S13Cr weld metal gave a large scatter on
ductility compared to the duplex weld metal when cathodic protection was applied. The reason for
this can not be explained based on the existing data. Less investigation has been conducted on
S13Cr filler and the effect of post weld heat treatment. If post weld heat treatment has a beneficial
effect, longer holding times can be applied compared with todays practice on duplex welds, since
the microstructure of S13Cr is not as sensitive as duplex regarding precipitation of sigma phase.
Sigma phase reduce the corrosion properties and fracture toughness of duplex. In addition the
S13Cr weld metal has a favourable higher strength, causing less deformation of the weld.

The effect of post weld heat treatment has not been fully documented regarding hydrogen
embrittlement. So far, it is indicated a slightly decreased sensitivity with post weld heat treatment.
More investigations are needed to confirm this statement. On the other hand post weld heat
treatment is recommended for increased resistance to internal corrosion in the heat affected zone,
as the post weld heat treatment most probably reduces the depletion of Chromium carbides at the
grain boundaries /14/.

4.1.4 Tensile properties and temperature

The yield strength of duplex and super duplex steels decreases with increasing temperature which
is normal for metallic materials. S13Cr is the opposite. The yield strength increases with
approximately 50-90MPa if the test temperature is increased from room temperature to 120ºC /5/.
This means that evenmatching weld metal (which is the normal case of super duplex welds on
S13Cr pipes) at ambient temperature will change to a significant undermatch situation at typical
service temperatures. If high loads are present on a pipe section during operation, significant
deformation of the duplex weld metal can occur. During shut in with increased temperature and
increased internal pressure, both addressing an expansion of the pipeline, a significant undermatch
situation can lead to increased risk of buckling with subsequent cracking. The unfavourable
undermatching of duplex welds will probably be avoided by applying S13Cr fillers. Still work has
to be done to increase the ductility and fracture toughness of the S13Cr welds. Investigations
 16

regarding the alloying of the welding wires, type of welding method, welding process and post
weld heat treatment are needed.

4.2 Subsea failures of pipelines on the Norwegian continental shelf

Approximately 320km of S13Cr pipelines are installed on the Norwegian continental shelf. Most
of these pipelines are girth welded approximately every 12m, giving a total of 27 000 to 30 000
girth welds. Within the next few years it is indicated the total length will be increased to
approximately 500km of S13Cr pipelines on the Norwegian continental shelf.

• fracture of Åsgard pipeline during reeling. Detected during reeling on, 1998
The following failures on subsea pipelines on the Norwegian shelf have been investigated:

• anchor damage Åsgard. Internal pitting corrosion detected during post-incident investigations,

• fracture of the TUNE pipeline during reeling on/off. Leakages detected during subsea pressure
1998

• fracture of the TUNE pipeline after hyperbaric repair welding. Detected during subsea
testing, 2001

• fracture of Åsgard HUBs (not S13Cr). Initially detected during pressure testing and later
pressure testing, 2001/2002

• two fracture in conjunction with anode pads on the Åsgard field. Detected during operation,
during operation, 2001

• fracture of Gullfaks pipeline at towhead. Detected during operation 2003

2002

• fracture of superduplex at Garn-Vest. Detected in June 2003.

The common denominator of all the failures listed above is hydrogen embrittlement or hydrogen
induced stress cracking, except the anchor damage at Åsgard. The internal corrosion found during
the investigation of the anchor damage is considered irrelevant and is not commented further. But
the high temperature internal corrosion experienced by NAM in Holland is presented (see section
4.3).

In the following the failure investigations are shortly discribed and discussed /25/.

4.2.1 Åsgard and Tune reeling incident and Tune hyperbaric repair weld failures
Leakage and failure have been obtained for both duplex and matching S13Cr welding
consumables (Åsgard reeling incident and Tune, respectively). But the sequence of cracking is
probably different, due to the inherent large difference in hydrogen diffusion ability between the
two cases (i.e. much lower diffusion rate of hydrogen in duplex). The Åsgard reeling incident was
explained by diffusion (release of hydrogen) from the duplex weld metal into the S13Cr heat
affected zone during subsequent coating and thermal insulation process (at approximately 100-
150°C) resulting in fusion line cracking. By contrast, weld metal cracking was found in the Tune
case. Investigations have shown that diffusion of hydrogen in ferrite-austenite is slow, but most
studies focus on base metal microstructures, which are very different from those formed during
low heat input multipass welding. Here, more investigation is needed to verify the suggested
mechanisms.

The Åsgard reeling incident concluded that duplex consumables for S13Cr pipeline welding shall
be heat treated (so called baked) and delivered with maximum 3ppm Hydrogen. This requirement
is today implemented in internal company specifications. Implementation into national and
international specifications is recommended.
 17

The Tune hyperbaric repair welding clearly demonstrated that moisture in the shielding gas during
welding of S13Cr is unfavourable. The repair welds fractured in a brittle manner, and it was
concluded that the moisture arose from the subsea transport hoses of the shielding gas. The Tune
flowline was replaced with a carbon steel line pipe material.

4.2.2 Gullfaks failure

The recent failure at Gullfaks indicates that hydrogen from welding caused cold cracking to take
place /26/. The cracking was initiated in the transition girth-weld-toe to pipe-surface. The weld toe
of the girth weld acted as a stress concentration on the pipe. The cracking extended from the weld
toe/fusion boundary and into the S13Cr heat affected zone. The cracking could have been
introduced by lack of degassing of the duplex consumables to reduce eventual unfavourable
content of hydrogen, see section 4.2.1. There is also a possibility that moisture in the shielding gas
or condensation on the pipe wall could have been the hydrogen source. To verify the source of
hydrogen in this failure investigation is impossible without proven documentation of the welding
wire and welding process. It is believed that the extension of the cold cracking did not penetrate
the pipe wall.

The through thickness fracture, giving leakage, was brittle with evidence of secondary cracking.
Secondary cracking is typical for hydrogen embrittlement. Long term diffusion of hydrogen from
the weld deposit could be the embrittlement mechanism of the final fracture, even though the
measured levels of hydrogen were below critical values in the duplex weld and neighbouring
S13Cr pipe. Hydrogen from cathodic protection is assumed to play a minor role as the coated and
thermal insulated field joint, surrounding the area of the failed pipe section, is believed to have
given sufficient bonding to the pipe, insulating sea water at least 300mm away from the fracture.
Even though calculations on hydrogen diffusion have shown that hydrogen can diffuse far and
increase the local hydrogen concentration to levels above critical levels for cracking, further
studies are needed to document the diffusion in such cases.

4.2.3 Fracture at anode pads, Åsgard

Operational pipelines are exposed to global loading that locally can exceed yield due to local
stress/strain intensities caused by geometry and material mismatch. Calculation for the Åsgard
anode pad gave a stress concentration factors in the range of 2-2.5, which implies that local plastic
deformation has taken place /27/. This is a result from a sum of residual stresses from welding,
local geometry and external loading.

The fractures occurred in pipe sections with high global stress given by free spans of the pipeline.
The seawater has penetrated through the thermal insulation following breaks/openings between
the insulation and the pad-to-anode connections. Hydrogen from cathodic protection combined
with local plastic deformation is believed to be the mechanism of the observed fractures.

4.2.4 Failure of Åsgard Hub’s

Traditionally the transition pipe-to-flange (e.g. Hub) is made of heat treatment steels of typical
carbon steel grade AISI 8630. Heat treatment steels are not the topic of this report. Still it is
considered interesting as the failure analyses points at bi-material joints, residual stress, hydrogen
embrittlement, microstructure/heat treatment and quality assurance. The heat treatment steel,
which have a carbon content of approximately 0.3%, is considered not weldable for structural
applications. To avoid the brittle heat affected zone of such high carbon steels, the fabricators
build up a weld deposit on the end surface called buttering weld. Normally inconel 625 alloys
(high strength austenitic stainless steel) are used. Heat treatment at 650°C/4 hours is performed to
obtain acceptable toughness on the heat affected zone of the flange material. The buttering weld is
 18

often an elongation of the internal clad weld of the Hub. Cladding is performed to sustain the
internal corrosive environments.

The failures of the Åsgard Hubs showed that the fractures were located at the fusion boundary
between the buttering weld and the flange material /28/. The conclusions of the failure

• unfavourable groove geometry (e.g. perpendicular to the loading axis)

investigation were as follows:

• the heat treatment has activated diffusion processes giving significant embrittlement of the
inconel side of the fusion boundary due to carbide precipitations and formation of martensite.
The heat affected zone of the carbon steel (e.g. flange material) was embrittled by significant

• a significant local strength mismatch was experienced with micro Vickers hardness in the
decarburisation, giving formation of coarse ferrite

range 450-600 on the inconel side of the fusion line and ca 220-240 in the heat affected zone

• cooling from the heat treatment temperature to room temperature has introduced significant
in the flange material

hoop and radial residual stresses (in the magnitude of 300-600MPa through elastic finite

• high levels of measured hydrogen at the fusion boundary indicates poor control of filler
element analyses)

materials and/or welding process. It is indicated that cathodic protection could have increased
the hydrogen concentration during operation

72 Hubs was replaced at the Åsgard field. Several thousands of other Hubs are located in subsea
pipeline systems around the world…

Later investigations showed that the mechanical properties are very sensitive to the heat treatment
temperature /29/. It is indicated that heat treatment of buttered and cladded Hubs most probably is
unfavourable regarding the microstructures on both sides of the fusion line and the residual stress.
It is also clear that the qualification procedures for the buttering weld does not represent the same
verification level as for load carrying girth welds of a pipeline system, even though the loading
characteristics at the transition to the flange at least matches the loading of the pipeline.

A new design developed by Statoil, using cladded weldable carbon steel type F65 to avoid the
buttering weld, was mainly used in the replacements.

The new carbon steel design also reflects the limited short term availability of large diameter
forged qualities of duplex materials internationally. The duplex qualities of such dimensions are
also often connected to unfavourable brittle sigma-phase precipitations, which will reduce the
corrosion resistance and fracture toughness significantly.

4.3 Failure caused by high temperature corrosion

Experiences by NAM in Holland have put attention to these materials susceptibility to high
temperature corrosion. One case of lean grade (low alloyed) seam welded (laser welded) onshore
pipeline showed intergranular corrosion after a few years in service, and recently two seamless
pipeline (also lean grade) leaked after short time in service. All fields were sweet service and
temperature below 80°C. The latter corrosion mechanism was pointed out by SINTEF as a
possible problem first time at the Eurocorr 2001 /8/ and a summary was given at the
supermartensitic 2002 conference /13/. Here also mechanisms were suggested and discussed in a
paper by Ladanova et.al./14/.

NAM decided to replace the pipelines with duplex stainless steels and has not considered S13Cr
steels for sweet or sour service applications since.
 19

4.4 Summary of the subsea failures

From the failure investigations it can be concluded that S13Cr was in the category “new material
in known environment”. The corrosion properties and limits for S13Cr steel was focussed and
other environments was neglected as a problem due to existing knowledge at that time on
materials as duplex and super duplex stainless steels. The unknown environment was initially
unfavourable hydrogen content in the welding consumables causing cracking on S13Cr and not on
the duplex steels. Secondly it was humidity in the shielding gas or condensation during the
welding process. Thirdly is was clear that normal cathodic protection systems, to protect the
outside of the pipeline from corrosion, was a major source of hydrogen during operation. All these
sources of hydrogen have showed to be detrimental for S13Cr. It is clear that the indications on
hydrogen embrittlement on S13Cr from the research institutes and universities in the mid 90-ties
was not targeted and understood by the offshore industry prior to the use of the new pipeline
material.

The Hub failures also points at unsatisfactory quality assurance and design. It should be clear for
manufacturers of Hubs how their product is used in a pipeline system, and more specifically, that
the buttering weld is loaded to stress levels similar or even higher than in the pipeline because of
geometry constraints from the stiffer flange and the very high levels of residual stress. The
qualification and requirements of such should therefore at least match the type and amount of
testing required for qualification of pipeline girth welds. This was not the case. The lack of
documentation should have been picked up by the operator, which in first place is responsible.
And, as mentioned earlier, the design was poor by means of unfortunate material combinations,
heat treatment and groove geometry.

The anode-pads are a good example of poor design. Bringing a relative stiff body onto a rather
flexible pipe will generate stress concentration. In addition, the connector between the anode and
the pad is penetrating the pipe coating and thermal insulation. This has shown to be a weak link
regarding the possibility of seawater access to the pipe wall. New designs are already
implemented. Pipe failures linked to anode pads have occurred in conjunction with free spans (i.e.
in highly loaded pipe section). Even with a poor pad design, the failures could have been avoided
by better planning and control of the position of the pipe on the seabed to reduce the stress in the
pipe, or planning of positioning of the anode pads outside highly loaded pipe sections.

5 UTILISATION OF RESEARCH IN THE OFFSHORE INDUSTRY

5.1 Fracture toughness and high strength steels

The driving force for research regarding increased strength on offshore construction steels is not
mainly focused on the strength itself. The fracture toughness of the heat affected zone from a weld
has shown to be a limiting factor. Also the requirements of overmatching strength and acceptable
fracture toughness of the weld deposit are vital parameters in the process of development of high
strength offshore steels.

The driving force for the development of fracture mechanics have been the need to design and
operate safe nuclear power plants, where failure has to be avoided by all means. The standards for
fracture mechanics testing and defect assessment is influenced by this way of thinking. Lower
bound solutions are used so that the fracture mechanics testing and analyses becomes simple and
safe for many situations. The negative side of this is that the analyses become very conservative
and that they do not catch sensitive parameters for increased utilization.
 20

For many offshore applications the standardized defect assessment procedures become too
conservative, and workmanship criteria that in some cases are less conservative than standard
Engineering Critical Assessment analyses are used instead. The danger is that the workmanship
criteria become too static. There is a continuous development of new welding procedures and
pipeline steels. Workmanship criteria that do not reflect these changes in material properties are
potentially dangerous to use when there is a rapid development in materials.

Critical defect size in welds and at fusion boundaries for offshore applications have, for decades,
been predicted through Engineering Critical Assessments, following the procedures of PD6493
(now BS 7910) and standard test methods for determination of fracture toughness (like BS 7448).
It has been demonstrated, through large scale wide plate testing, that the assessments have been
very conservative. The conservatism is mainly linked to the significant different constraint of
planar defects in the construction compared to deeply notched standard test specimens. How
residual stress interacts with the fracture toughness is not clearly understood. But it is clear that
the approach used today is conservative (cf. BS 7910). Still these assessments are considered as
vital input for design. However, it has continuously been called for procedures to limit the
conservatism. Hence, we have to be aware that the historically high degree of conservatism in the
testing and analyses may have been a main contributor to the robustness of the material selection
and probably compensated for unexpected or unknown effects when such have occurred.

The last years research have been performed to develop more tailor-made fracture mechanics
testing procedures and defect assessment analyses for offshore pipelines. The methods is more
accurate and is better suited to catch critical changes as a result of the development of new
linepipe materials, welding procedures and pipeline geometries. More accurate methods are
important tools to avoid unnecessary conservatism and to quantify safety.

Today the conservatism of fracture mechanics assessments on pipelines is quite limited (e.g. for
reeling and J-laying in particular, TWI/DNV/SINTEF JIP Report 2003; "Project Guideline for
Pipeline Installation Methods Introducing Cyclic Plastic Strain" /30/). This JIP Guideline will be
converted to an open DnV RP within year 2004. The guideline specifies several changes from the
general standardized fracture mechanics test and assessment procedures, for instance the
introduction of the new small scale fracture mechanics test specimen (Single Edge Notch Tensile
specimen) that is tailor made for pipelines. The new test specimen geometry is specified. The new
guideline is much more accurate than general standards for defect assessment procedures. The
result is confidence to defect assessment analyses by the offshore industry. The use of defect
assessment analyses of pipelines that can catch eventually critical combinations of defects,
loading and material properties is important for a continuous development and safe use of new
technology for offshore pipelines.

Another important contribution for increased robustness of pipelines will be the new “Design
Guideline for Offshore Pipeline" that is under preparation. This work is performed within the
Joint Industry Project “Fracture Control of Offshore Pipelines". In this work the effect of internal
and external pressure will be implemented, partial safety factors will be developed, and the format
of design equations, developed for failure, will be on the same format as for other failure modes.
This opens for increased use of fracture mechanics in the design phase of a pipeline construction
project.

As the fracture mechanics assessments of today predict the maximum allowable defect size close
to the actual limits of the pipeline during installation, it is of main importance to take action and
compensate for eventual reduced fracture toughness during operational conditions. If not, the
operational assessments are clearly non-conservative. Today the fracture toughness obtained in air
is often used for both installation and operation. This practice should change as lower bound data,
 21

representing the fracture toughness under worst case operational conditions, is available for some
steels to day.

5.2 Corrosion history and related R&D

Three major findings or problem areas with respect to corrosion and corrosion protection have
been uncovered in the qualification and R&D of the S13Cr stainless steels. These are as discussed
previously; 1) general corrosion at low pH caused by the use of strong buffer, 2) high temperature
intergranular corrosion caused by sensitisising (Cr-carbide precipitation) of the heat affected zone
by multipass welding and 3) hydrogen induced stress cracking caused by high local levels of
hydrogen.

The first mechanism was solved rather quickly and explained. It was more or less a result of
standards for testing being adopted from carbon steel testing to stainless steel testing without a
proper evaluation.

The two last degradation mechanisms have caused more concern and resulted in significant
research and development. Besides of studying the robustness in post weld heat treatment to heal
Cr-depletion, a PhD study were undertaken to study the mechanism and under which welding
conditions carbides were formed. All answers are not given and further research should progress,
especially towards environmental limitation and welding procedures to avoid precipitations.

The last uncovered degradation mechanism was the impact of hydrogen either from welding or
from cathodic protection. To avoid hydrogen from welding has been overcome through better
welding procedures as to avoid humidity/dirt/oil in any form and by using consumables that are
degassed to reduce hydrogen uptake. Hydrogen from cathodic protection is more unforeseen
because it depends on several other factors as the coating and coating damages. Some
constructural matters can be and has been done with respect to design of anode pads and how to
place anodes and also to reduce the polarisation by controlling the potential to a level where
hydrogen is not produced (above -730mV SCE) and at the same time not exceeding the potential
for corrosion (below -500mV SCE). The effect of hydrogen has shown not to be elucidated with
traditional test methods, which has led to focus on fracture mechanics testing under environmental
impact. The enabling of fracture mechanics analyses are more thoroughly discussed in the next
chapter.

A co-operation in this area has been established with NKK (JFE group) and a student from the
Norwegian University of Science and Technology (NTNU) worked together with researchers at
JFE during the Spring 2003. The cooperation is continued in 2004.

5.3 Environmental Assisted Cracking

It has been demonstrated that the fracture toughness of duplex steels and S13Cr steels drops
significantly under normal cathodic protection. Major impact is also indicated on the crack growth

• establishment of maximum applied operational stress for the pipeline

rate during fatigue. These parameters are of main importance for design criteria as;

• estimation of operational lifetime

• accept the lower bound fracture toughness and crack growth rates during normal cathodic
To avoid hydrogen induced cracking on S13Cr pipelines, three possible designs can be applied:

• controlling the galvanic potential to >-900mV SCE. The effect of the cathodic protection
protection accounting for them during operational design.

is probably neglectable. This is based on tests performed at 1 atm pressure. Sub sea effects
with increased external pressure on the pipe surface must be investigated.
 22

• if waterproof coating is ensured, cathodic protection is not needed.

Several test methods have been applied in characterizing the susceptibility to hydrogen

• single edge notch tensile fracture mechanics test specimen (SENT)

embrittlement:

• single edge notch bend fracture mechanics test specimen (SENB)

• slow strain rate testing on round tensile test specimens (SSRT)
• four point bend testing on as welded surface and machined surface (4pb)
• full scale testing with and without artificial surface flaws
• constant load
• others

Traditionally sensitivity to hydrogen embrittlement has been investigated by applying a given

displacement onto a four point bending specimens with original pipe surface representing 90-
100% of the actual yield strength or by applying a constant load on round smooth tensile
specimen representing 90% of the actual yield strength. Loading this type of samples, with a
sensitive microstructure/material present, to yield or even locally above yield (especially on four
point bend tests) has shown not to be a representative test method for operational conditions. The
loadings have been performed in air followed by submerging and cathodic protection. The factors
should act simultaneously. This will prevent oxides to form subsequent to the loadings. However,
most important is to have hydrogen charging when the material is plastic deformed because in
these areas a higher and critical hydrogen concentration is achieved.

Slow strain rate testing of typical round tensile specimens combines the above mentioned factors.
And that is probably why this method exposed the sensitivity to hydrogen embrittlement in the
first place. But the only measure to evaluate from a slow strain rate test is the fracture contraction
(i.e. ductility measure). This could be a simple way to qualitatively compare and range materials.
But converting a ductility measure to a relevant design acceptance criterion is difficult and must
be supported by several empirical and analytical simplifications. From a fracture mechanics point
of view, this would be the same as finding a parameter describing the fracture toughness of a
material by means of the fracture contraction or fracture elongation from a standard tensile test.
As a measure for design input, this is considered as an inadequate description of the fracture
mechanics behavior of a material.

Slow or constant load fracture mechanics testing by using the SENT or SENB test specimens have
shown to be very sensitive to environmental impact in the same way as standard slow strain rate
testing. And the advantage with fracture mechanics test specimens compared to slow strain rate
specimens is that the data can be applied for fracture mechanics assessment.

The crack growth rates show the same tendencies as the fracture toughness. These findings are
very recent and are produced in an on-going Joint Industry Project on fatigue.

Four full scale tests have been undertaken. Two of the tests were conducted with cathodic
protection. One of the full scale tests and several standard small scale 4-point bend tests, loaded to
specified stress levels around yield, were submerged into the ocean to 600m depth to simulate
actual loadings and environment. However, no cracking was observed on any of the tests. Critical
conditions with respect to cathodic protection, local loading/strain concentration and pipe
movement have probably not acted simultaneously.

From the above results it is obvious that testing has to reflect operational conditions in a reliable
way. Taking the step forward to fracture mechanics testing under environmental impact a more
 23

reliable and representative test method can be developed. It is therefore important to understand
and reproduce the expected local stress behavior during operation applying especially designed
test geometry.

As the properties are significantly reduced under cathodic protection, it is of main importance to
increase the robustness by implementation of the actual properties in the design phase of coming
pipeline construction projects. It is also vital to ensure that the actual loading on the pipe is within
the design limits.

6 QUESTIONNAIRE TO THE OFFSHORE INDUSTRY

Focus on industry experience is the main topic of the questionnaire. The questionnaire is mainly
related to S13Cr pipeline integrity, but some of the questions have a more general approach. The
questionnaire is outlined to get an acceptable basis for a systematic evaluation of risk.

The basis for the questionnaire was meetings regarding robustness in material selection with
Norsk Hydro and Statoil and questions presented in the pre-project report /2/ (Chapter 5.2).

The intension of the questionnaire was to engage leaders/representatives of material

groups/departments of larger oil & gas companies by answering the questions, providing relevant
information on an open and free basis. To get an acceptable level of internal company
information, it was in the introduction of the questionnaire emphasised that the information should
be treated as confidential and reported anonymised (in general terms). The questionnaire was
distrubuted to leaders/representatives of the material groups in the Oil & Gas divisions of the
companies:

•
•
Norsk Hydro

•
Statoil

•
BP

•
ExxonMobil

•
Shell

•
Total
NAM

In addition, the company Intetech Ltd by Dr. Liane Smith, was asked to comment the
questionnaire.

Norsk Hydro, Statoil and NAM replied on the questionnaire on all sections. Intetech gave
comments where applicable. SINTEF is very grateful for the contributions.

ExxonMobil answered 7 out of the 22 questions. The 7 answers are not included in the evaluation
because of the very limited amount of relevant information. BP, Shell and Total did not respond
on the questionnaire. The intension of evaluating the best international practice is therefore
limited.

Details of the questionnaire are presented in Appendix A. A summary is presented in the next
chapter.

6.1 Summary of the questionnaire

The answers of question 21 (see Appendix A) indicate that the industry would prioritise
 24

1. to improve the knowledge on failure mechanisms and new test methods

2. to improve the design basis
3. to increase the implementation of test results on S13Cr line pipe materials into standards and
specifications
4. to evaluate time schedules for pipeline construction projects more closely

Priority 1 and 3 correlates well with actual activities on research of line pipe materials today (see
section 3.8 and 6.1.1). Priority 2 is, from an industry point of view, clearly improved during the
recent years by increased collaboration between design engineers and material specialists.
However, it is emphasis that large potential on increased robustness can be gained by
improvement of the design basis (see section 6.1.2). Priority 4 is related to management of
pipeline construction projects and procedures for such (see section 6.1.3).

Efforts of combining the answers of the questionnaire to the offshore pipeline failures and
research experience are presented in the next sections.

6.1.1 Knowledge gaps

The term "complete design" includes a complete understanding and quantification of all possible
sources of environmental impact on the construction/structure during operation. However holes
are still present today. This may be one of the major reasons why the predictions of lifetimes on
S13Cr pipelines are questioned world wide. The design basis, the operational experience and
operational monitoring/registrations are limited.

The experience that has been accumulated through the last few years has shown that even if a
S13Cr flow line has been fully qualified, failures have occurred in unexpected modes. These
failures may likely be defined as a result of knowledge gaps. The recommendations are as
follows:

A. Close the knowledge gaps that are known and present today on pipeline systems. This requires
establishment of new relevant test methods and extensive testing and verifications programs
within the frame of environmental assisted cracking. As soon as a test method is critically
evaluated and regarded acceptable for qualification purposes, it is of main importance to
standardize the method. It is also vital to implement the test results in relevant standards and
specifications.

B. Bring the design analyses closer to completeness. See next chapter for more details.

C. It is clear that fundamentals regarding environmental assisted failure mechanisms are not fully
understood. It is recommended that more basic and long term research is initiated to provide a
better understanding of these issues (research institute programs combined with Ph.D. programs).

D. Strategic strengthening of the communities in the offshore industry, universities and research
institutes on Environmental Assisted Cracking (Cracking, Fracture Mechanics, Fatigue etc.).
Strategic investments on relevant equipment are also necessary to bring forward. Large volumes
of laboratory work are connected to characterization and optimization of different materials
exposed to different operational conditions.

E. It is believed that closer collaboration between oil companies/operators,

manufacturers/fabricators and research institutions/universities are likely an effective way of
reducing knowledge gaps in the future. Increased communication, also inter disciplinary, is one
way of identifying coming sources of failure.
 25

6.1.2 Increased robustness by improving the completeness of operational design

Both SINTEF and the offshore industry consider the design basis for pipelines as insufficient.
This is exposed through the answers of the questionnaire and more directly from evaluations of
the investigations on pipeline failures. The pipeline failures show deviations between the actual
stresses/loads on the pipeline and the maximum allowable design stress/loading. The actual stress
on sections of the pipelines is often well beyond the design criteria. This is could be combined
with wrong predictions of degradation- and failure mechanisms.

It is recommended that tools for improved accuracy, by means of numerical simulations combined
with monitoring during operation and field experience, should have more focus in the future.
More effort on iterative design and verifications is emphasized. It is proposed that this can be
realized through improved completeness of operational design through the recommendations

• take into account the strain history from installation on the properties of the line pipe
listed below:

material during operation. For installation, reeling and J-laying are commonly used. Both
these methods apply plastic deformation to the pipe material. The mechanical properties
differ, especially between the top-of-pipe and bottom-of-pipe positions. Through the
recent year of pipeline qualifications and fracture mechanical assessments, it is indicated
that installation deformations ending in tension are unfavorable regarding maximum

• take into account environmental impact on line pipe material properties regarding fatigue
allowable defect size during both installation process and operation.

crack growth and fracture if the pipeline is not ensured waterproof or the galvanic
potential is not controlled. Fracture toughness and fatigue crack growth obtained under

• take into account the effect of internal general corrosion on pipe wall thickness and
relevant environments are vital input for design and lifetime analyses.

eventual local corrosion giving notch effects (stress concentrations). This is typical for

• requirements on increased documentation of seabed topography along the planned route of

carbon steels and not stainless steels.

the pipeline and how it affects the stress distribution on the pipe during normal operational
conditions and shut downs. It is the interaction of seabed topography, fixation points of
pipeline and burying/trenching etc. together with change of internal pressure and thermal
expansion/contraction during shut downs that likely could produce critical local loadings

• after installation and before operation, it is suggested to increase the documentation of the
on the pipeline.

position of the pipe to verify whether the design criteria are met or not. If deviations are
registered, re-design and evaluation of remedial actions are recommended. Remedial
actions are in this context related to changing of support, fixation, burying etc. to reduce

• monitoring of pipe movements on selected pipe sections. On pipe section(s) that are
the local loading to an acceptable state according to the design criteria.

considered as most critical (highly loaded, but within the design criteria) during operation
it is suggested to monitor the pipe movement during Ready For Operation procedure and
initial operation for comparison with design analyses. These movements can act as a good
reference and verification of the global and local design stress analyses. improving the
interaction between local and global stress analyses taking into account local geometry and

• more exact planning of the route of the pipeline, which improves the possibility of moving
material mismatch.

unfavorable geometrical stress concentrations outside highly loaded sections of the

pipeline. It should be possible to predict the positions in the length direction more exactly
(as well as other directions) when the exact route is planned. Together with the global
design stress analyses, this opens for improved planning of the positions of geometry
 26

dependent stress/strain concentrations. This would improve the utilization of materials and

•
the robustness of the design.
improved monitoring of critical parameters on pipelines during operation. The confidence
on operational design is dependent on verification of operational conditions. Increased
focus on monitoring is suggested as one tool to increase the confidence. Today a limited
number of well heads are instrumented for optimization of production. It is suggested to
extend the experiences from well heads on operational monitoring of pipelines, including
registrations of pipe wall loadings.

6.1.3 Technical management of pipeline construction projects and pipeline operation.

Cost-benefit is in many cases a major driving force for pipeline design and material selection.
This implies optimum utilization of the materials. To reduce time schedules on construction and
installation is clearly a tool to reduce the costs of pipelines. For the management of pipeline
projects, which differ between companies (cf. questionnaire), it is indicated that increased

• taking responsibility for securing relevant competence and to secure interactions between
robustness can be obtained by

• improving the use of "Best international practice". Technology transfer from one project to
fields of competence through all stages from design to operation.

• taking responsibility regarding the connection between design criteria, test data and
another should be outlined in written procedures.

operational conditions (i.e. elevate the accuracy of lifetime assessments, evaluation of risk
of failure etc.). S13Cr as pipeline material is considered as new technology and the

• more detailed time schedules accounting for upcoming problems during construction
operational experience is too short to fully verify reliability and robustness.

projects. The oil companies are mostly satisfied with the time schedules regarding pipeline
construction projects (cf. questionnaire). This is not the case looking closer on the
fabrication schedules and how upcoming problems in this phase of construction have
increased the risk of poor design and lack of evaluations.

7 DISCUSSION
The implementation of S13Cr as pipeline material on the Norwegian continental shelf was
initially believed to be within the category "new material in known environment". Failures during
fabrication, installation and operation have shown that the S13Cr material faced unexpected
"corrosion" behaviors. If S13Cr had been treated as "new material in unknown environment", the
risk of meeting the number of failures experienced on the Norwegian continental shelf (and
elsewhere) would probably have been reduced.

A historic review of the publications and reports, related to how our knowledge on S13Cr
materials has grown, shows clearly that nobody is perfect. This involves the operators/oil
companies, the research institutes/universities and official authorities/Norwegian Research
Council. What did we know and at what date? Why did we not do the right prioritizing?

The development of new test methods is primarily based on simulating a specific degradation
mechanism. This approach provides at least some understanding of the mechanisms. The history
indicates that there are problems related to selection of test method approaches, which means that
some failure mechanisms were not identified in advance of field experience. And field
experiences could, in some cases, not have been anticipated. But it does suggest that the
laboratory work may have been too narrowly focused. It is emphasized that traditional test
methods for corrosion and mechanical characterization may not represent the actual case.
 27

Today the Norwegian O&G companies are in the forefront regarding research and development
on typical pipeline materials. On S13Cr steels, they are also the users. The investigations on
pipeline failures on the Norwegian continental shelf have clearly been the main contributor to
developments within discipline Environmental Assisted Cracking and especially Environmental
Assisted Fracture Mechanics. This is now a focused area of competence and larger research
programs are supported by the industry. International engagement is expected as hydrogen
induced cracking has shown significant impact on super duplex and duplex steels as well. These
materials have been considered as a robust selection until recently. And super duplex and duplex
steels are widely use abroad and are also planned for the Ormen Lange development.

The pipeline failures have not only exposed knowledge gaps on degradation mechanisms. The
failures indicate a need for improved interaction between global and local design. This is from the
industry mainly linked to reduction of geometry dependent stress concentrations. SINTEF is in
addition emphasizing improved utilization and robustness by including interactions with seabed
topography in the design analyses. The planned route of a pipeline should be part of the design
basis and deviations between the planned route and the actual route should be documented and
evaluated.

Finally, remote monitoring of relevant parameters on a pipeline could be a tool to avoid

unexpected longer shut downs and repair/replacement, as it opens for evaluations of remedial
actions to reduce the risk of failure and for continuous evaluations of remaining lifetime.

8 REFERENCES

1. Oljedirektoratet: Sammendragsrapport, "Utvikling av risikonivå – norsk sokkel" Fase 3 –

2002, April 2003
2. J.M.Drugli, C.Thaulow, J.Ødegård, T.Rogne, R.Stokke, S.Berge, J.Berget: Pre-project,
"Robust material selection in the offshore industry", SINTEF Report STF24 F03202,
January 2003
3. J.J. Dufrane: Supermartensitic Stainless Steels 1999, Brussels, Belgium, 2002, Belgian
Welding Institute, 19-24.
4. P. Toussaint, J.J. Dufrane: Supermartensitic Stainless Steels 2002, Brussels, Belgium, 2002,
Belgian Welding Institute, 23-27.
5. T.Rogne, M.Svenning, H.I.Lange, S.Åldstedt, H.Fostervoll, "Testing of large diameter
weldable 13%Cr steel pipes for the Tangguh Production conditions", SINTEF Report
STF24 F00246, October 2000
6. J.M. Drugli, T. Rogne, M. Svenning, S. Axelsen and J. Enerhaug: ”The Effect of Buffered
Solutions in Corrosion Testing of Alloyed 13 % Cr Martensitic Stainless Steels for Mildly
Sour Applications”. NACE Corrosion ’99, Paper No 586, San Antonio, Texas, April 1999.
7. T.Rogne and M.Svenning: “Effect of chloride and partial pressure of H2S on the SSC
susceptibility of martensitic SS”, OMAE 2003, Paper No 37191, Cancun, June 2003
8. T,Rogne at. al.: “ Elevated temperature corrosion/cracking of large diameter weldable 13%
Cr linepipe”, Eurocorr 2001, Riva Del Garda, Italy, October 2001
9. T. Rogne et. al.: “ Elevated temperature corrosion/cracking of large diameter weldable 13%
Cr linepipe”, 6th International pipeline conference & exhibition, Merida, Mexico,
November 2001
10. T. Rogne et.al.:”Intergranular corrosion/cracking of weldable 13%Cr steel at elevated
temperature”, NACE CORROSION 2002, Paper No. 02428, Denver, April 2002
11. O.M. Akselsen, M. Bjordal, T. Rogne, G. Rørvik: "A study of the properties of 13% Cr
martensitic stainless steels", STF24 F95605, Nov. 1995.
 28

12. O.M. Akselsen, G. Rørvik, P.E. Kvaale, C. van der Eijk: “Study of properties of new 13% Cr
martensitic stainless steel”, Weld. J. (printing in progress, 2004)
13. T. Rogne and M. Svenning.:” Intergranular corrosion of supermartensitic stainless steel – a
high temperature mechanism?”, Supermartensitic Stainless Steel 2002, Brussel, 3-4 October
2002
14. Elena Ladanova, Jan Ketil Solberg and Trond Rogne.”Transmission electron microscopy
investigation of precipitation reactions in coarse-grained heat affected zone in two 13%CR
supermartensitc stainless steels”, Supermartensitic Stainless Steel 2002, Brussel, 3-4
October 2002
15. T. Håbrekke, P.E. Kvaale: OMAE 2001, Proc. Int. Conf., Rio de Janeiro, Brazil, June 3-8,
2001.
16. O.M. Akselsen, G. Rørvik, C. Van der Eijk, P.E. Kvaale: “Mechanical properties of
experimental 13% Cr stainless steel weld deposits”, Paper presented at the Nordic welding
conference, September 20-22, 2000, Iceland.
17. T. Rogne et.al.: “Evaluation of Hydrogen Embrittlement of S13Cr Stainless Steels Based on
SSR and CTOD Testing”, NACE CORROSION 2003, paper No 03534, San Diego, March
2003
18. T. Rogne and M. Bjordal: "Testing of welded 13 % Cr grades martensitic stainless steels
for sours service applications." NACE Corrosion '97, Paper No. 62, New Orleans, 8-14.
March 1997
19. R. Aune, H. Fostervoll and O. M. Akselsen: OMAE 2003, Proc. Int. Conf.,, Cancun,
Mexico, June 8-13, 2003.
20. N.Hagiwara, N.Oguchi, "Fracture toughness of line pipe materials under cathodic
protection", NACE Corrosion Conference 1997, Paper No. 200
21. H.I.Lange:”SENT testing under constant load with cathodic protection – FRAM VEST
project, 13%Cr pipeline”, SINTEF test report, project 240672.70B, 2002-10-18
22. Stålmat forum, http://www.sintef.no/units/matek/Stalmat/hovedside_2002_2.html
23. Project Proposal of Stålmat forum, "New martensittic stainelss steels for cost effective and
environment friendly transport of oil and gas", November 1998
24. R.Aune, H.Fostervoll and M.Svenning:”JOTSUP workpackage 1, Task 1.4
Environmental/corrosion requirements – IRC test of supermartensitic stainless steels with
matching filler wires”, SINTEF report STF24 F03227, March 2003
25. T.Rogne et. al., "A State-of-art on Hydrogen Induced Stress Cracking of welded
supermartensitic stainless steels – SINTEF experiences" SINTEF Report STF24 F03282,
January 2004
26. Vigdis Olden, Ragnhild Aune, Andre Mikkelsen, Trond Rogne, Odd Magne Akselsen and
Synnøve Åldstedt, "Gullfaks Satellite C2 Towhead leakage incident. Failure analysis of
S13Cr steel weldment – revison 2", SINTEF Report STF24 F03274, November 2003
27. B.Nyhus, Z.Zhang, E.Østby and J.Ødegård:”Åsgard Anodeproject – Fracture Mechanics
Assessment”, SINTEF report STF24 F03229, May 2003
28. H.I.Lange, S.Åldstedt, E.Østby, "Investigation of fractured weld connection on Hub No.
AB-103, Ågard B. Revision 1", SINTEF Report STF24 F01283, November 2001
29. V.Olden, S.Åldstedt, W.Dall, M.Raaness, A.Hellesvik, A-K.Kvernbråten, "The Effect of
PWHT on the Material Properties and Micro Structure in Inconel 625 and Inconel 725
Buttered Joints", SINTEF Report STF24F02321_rev1, January 2003
30. S.Wästberg (DnV), H.Pisarsky (TWI), B.Nyhus (SINTEF), "Project guideline for
engineering critical assessments for pipeline installations introducing plastic strain. Revision
1", DnV Technical Report 2003-3135, July 2003
 APPENDIX A A-1
SINTEF Materials and Chemistry

QUESTIONNAIRE TO THE OFFSHORE INDUSTRY

In the following, the questions and answers are summarized and commented:

1. In the first half of the 90-ties S13Cr was introduced as a new pipeline material. Does Your company
have (or had) responsibility for a sub sea S13Cr pipeline? Could you please outline what pipeline
(sector/area), the year of installation and the approximate amount of km.

-Total of about 350km S13Cr sub sea.

-About 50km S13Cr/lean grade on shore.

SINTEF comments:
Today the Åsgard field is the largest user. In the coming years, large volumes S13Cr line pipe
materials are planned used on the Norwegian shelf

2. Please describe your procedures (international standards, NORSOK specifications, DnV

specifications, internal specifications, other specifications….) in material selection, step by step, from:

-Procurement specifications
-Input from the steel/pipe producers
• the very early stage
-NORSOK M-001
-Project specific Material Selection Report
• design/engineering -Project specification(s), DnV OS F-101, ASME/API
• fabrication -Project specification(s), DnV OS F-101
• installation -Project specification(s), DnV OS F-101
• service/operation -Operation plans for inspection and maintenance

SINTEF comments:
Through post-failure investigations on fabrication and installation problems, it is believed that
today's procedures for such are satisfactory regarding robustness. Still benefits can be obtained
through more detailed registrations of sub-, but near, critical weld defects and more narrow
dimensional tolerances to reduce stress/strain concentrations (see Chapter 6).

It is believed that large potentials on increased robustness are found in the interaction between
design and operation. How is it documented that the design criteria are fulfilled on an operating
pipeline? How accurate is the design? How accurate are the positional registrations of the
pipeline on the sea bed? This is more extensively commented in Chapter 6.

3. Who has the responsibility on each stage (oil company, engineering company, fabrication company,
material manufacturers, consultants…)

• for decisions -Oil company together with partners

• for accepting -Oil company
deviations -Company specific procedures
• for quality control -Oil company / project / fabricator / manufacturer
• for accepting -Oil company
alternative/new -All above
solutions -Company specific procedure
 APPENDIX A A-2
SINTEF Materials and Chemistry

SINTEF comments:
It is proposed to gain robustness by bringing the best company experiences into official
guidelines/specifications

4. Has there been a shift of responsibility from the oil companies to the delivery industry? Who is
controlling/following up the contractors? Do the contracts focus too much on price/time and too little
on material quality? Please outline the trends during the last 10 years.

• The implementation of the Norsok principles has increased responsibility for the overall

•
quality, also for the suppliers
Wider use of Engineering Procurement Construction Installation (EPCI) type contracts, where
the contractor is responsible for procurement and follow up of the sub-contractors

SINTEF comments:
Although it is seams clear for the oil companies that the final responsibility is laying on their
shoulders, it should be emphasised that contracting out has resulted in less control, through:
-leaner project groups
-mismatch between official and actual competence of contractor/sub-contractor

In one of the answers it was mentioned that a reversion of this trend was visible in the company!

5. Has the number of employees in the “material group” significantly changed the recent 10 years?

•
•
Stable conditions are outlined in the answers.
If the amount of employees is reduced, it is partly connected to reduced number of projects.

SINTEF comments:
None of the "material groups" has increased in size even though the type of materials are
continuously increasing in numbers, the utilization is increasing and the materials/material
combinations get more specialized and sophisticated.

6. Does your “material group” have written plans for Yes Partly No
• the competence the “material group” should hold? V
• the role the “material group” play in the company? V
• the role of the material specialist in pipeline construction projects? V V
• how the “material group” serve R&D and external R&D on research
institutes/universities? V V

SINTEF comments:
As the answers diverge in the two last questions, it may be of importance to emphasise the
relevance of well documented strategic plans and action plans for "material groups"

7. How do the plans (of question no. 6) match the actual situation?

• As shown in question 6

SINTEF comments:
In other words; they all operate according to plan. Probably this is not the exact situation.
 APPENDIX A A-3
SINTEF Materials and Chemistry

8. How do you transfer knowledge from previous pipeline projects and operation/service to a new
project? Please describe your systems and your philosophy.

•
•
Revision of technical specifications.

•
The use and update of Best Practices

•
Teambuilding’s for experience transfers

•
Seminars and regular internal meetings
Project close out reports (regarding construction)

SINTEF comments:
Are there written plans and responsible persons for revisions and updates? It is vital to secure the
distribution of important information.

9. Does your company monitor the international technology development on sub sea pipeline materials?
If yes: How is this organised and who is responsible for the activity?

It seems like the technical organisation of the different companies has groups responsible for
following up international R&D via publications/conferences/seminars.
Networks are established for different technical disciplines.
Responsibility is in some cases more diffuse.

SINTEF comments:
Like in other institutions and organisations, the material groups are divided into technical
disciplines. How is inter-discipline transfer of technology secured?

10. Recent years, technology findings on S13Cr (and on other high strength C-Mn steels) have exposed
significant sensitivity to Hydrogen (from welding, cathodic protection and internal H2S corrosion).
What are your conclusions on these findings?

For the Norwegian O&G companies, progress on characterisation and understanding are focussed

• establish better limitations for safe operation of S13Cr pipelines

through:

• external (sea water) environment: Focus on CP, robust anode connections, coating,

• internal environment: Establish better H2S limits (may not be suitable for even very mild sour
deformation/stresses during operation, installation method (= effect of plastic deformation)

• investigations to understand how to mitigate internal intergranular cracking from sensitation

service condition).

• material, fabrication, welding and installation: More stringent requirement to base material,
during welding

• take one step backward summing up the S13Cr experiences

fabrication+weld+PWHT+NDT and restrictions on installation methods might be imposed.

SINTEF comments:
The Norwegian O&G companies are the motors, driving the R&D of S13Cr line pipe materials.
They are also the users, so far. Hopefully, a coming JIP will engage more of the international
community. The intensions of the JIP are to close vital holes of lack of knowledge and to increase
the understanding of the failure mechanisms, concluding with design limitations to S13Cr and
other line pipe materials. Modifications of S13Cr are probably brought forward.
 APPENDIX A A-4
SINTEF Materials and Chemistry

11. Do you have procedures to secure communication of such conclusions (see question no. 10) to the
“design/engineering group”?

The companies emphasise a change, by means of closer cooperation between material specialists
and pipeline engineers today.

SINTEF comments:
This is a trend that probably will migrate to a higher degree of confidence between design stresses
and actual stresses.

12. Do you use any tool/procedures on risk based material selection? If yes: Please comment with
supplementary information.

•
•
risk analyse based on sound safety calculation models, where material data is an input factor

•
sound engineering practice and present know how
go/no-go, based on predicted corrosion rates and lifetime

SINTEF comments:
The answers are likely calculation models on maximum design loading/stress combined with
corrosion properties and predicted lifetime. No specific statistical tools or risk-models are
mentioned.

13. When a pipeline construction project is established in your company, are you satisfied with the role
you play and the time and resources you have available in accomplishing a pipeline construction
project?

Based on proven technology for pipeline construction work, sufficient recourses are present to
perform the work needed. Time could always be questioned.

SINTEF comments:
It is believed that this was not the situation some years ago. The O&G companies have gained
experience on pipeline construction projects and the time schedules seem relevant. However, Sintef
has observed lack of time in late stages of fabrication.

14. Contracting pipe construction projects involves transfer of responsibility. Do you have a system to
secure the competence of the fabricator/supplier and that the actual personnel of the contract are
present and are focussed on your project?

•
•
follow verification plans, milestones

•
material lead engineer is monitoring/following the project

•
material lead engineer is responsible for all material activities in the project.

•
the material lead engineer is covered by an experienced engineer
hire personnel with special competence for special tasks

SINTEF comments:
None.
 APPENDIX A A-5
SINTEF Materials and Chemistry

15. Do you have comments/experience on using Engineering Procurement Construction contracts in

pipeline projects involving technology development? How do you solve such projects on a contractual
basis?

•
•
technology development work is set out as a separate contract

•
O&G company taken the responsibility for technology developments
EPC's shall be based on proven technology

SINTEF comments:
None.

16. If the contractor gets problems following the time schedule, or you observe deviations in product
quality, or it is exposed lack of competence, how do you normally react? What do you do to
compensate such cases?

• question nearly impossible to answer. Different actions may be necessary based on the

•
problem using alternative solutions
more manpower to solve the problem or assist the contractor to replacement of the

•
contractor/supplier
responsibility for field developments lies with the O&G company as the operator, and proper
actions have to be taken in order to reduce the risk for any contractor to causing delays

SINTEF comments:
None.

17. Increased utilisation puts stronger requirements on both material and design. How would you
describe the interaction with the “construction and design group” in pipeline construction projects?

•
•
establishment of close cooperation between the pipeline design and material engineering

•
cooperation vital for limits for utilisation of both duplex and S13Cr steels in seawater
similar utilisation limits may also be valid for high strength carbon steels (>X80), however
further research is recommended

SINTEF comments:
The relation between design stress limits and actual operational stresses must be brought forward
as one of the major knowledge areas to increased robustness on pipelines.
Documentation of operational stresses by different type of registrations is needed.

18. Have you experienced that “late-production”, close to finalisation of a pipeline construction project (as
f ex type and mounting method for sacrificial anodes, implementation of instrumented pipe segments,
type of flange connections, transitions etc and the position of such on the pipeline (highly or low
stressed sections)), could be more exposed to suboptimal design/material selection?

•
•
all parts installed in a pipeline construction shall follow the same quality principles

•
solutions, including design and material selection shall be optimal
of course there have been such situations, in particular with respect to quality of line pipe
fittings

SINTEF comments:
It is believed that the two first answers purely focus on the ideal situation. Short cuts are almost
inevitable when time schedules are hard to keep. Compromises occur. By bringing safety up front,
it is clear that unfavourable compromises will be reduced. It must be better to sort out the problem
 APPENDIX A A-6
SINTEF Materials and Chemistry

during construction than creating a much bigger problem in some later operational stage.

19. Do you have any comments on the statement; “implementation of expert systems/electronic tools
reduce technical overview”?

•
•
no specific
there is no substitute for seeing yourself...

SINTEF comments:
Within the term "complete design" it is important to evaluate all materials and all different
joints/connections throughout the pipeline system. It must be ensured that eventual poor designs
and poor material combinations/joints are detected.
Electronic design tools tend to remove the focus from the actual pipeline.

20. Are you sure that there will be NO future failures/leakages of the sub sea S13Cr pipelines you are
responsible for (i.e. within the calculated lifetime of the pipeline)?

•
•
none can give such guarantee, neither for S13Cr pipelines or any other pipelines!

•
it is impossible to give such guarantees
yes we have removed all S13Cr pipeline. The reason for removal being that we could not be

•
sure that the installed S13Cr lines would be safe to operate
S13Cr is not selected in any of mine construction projects

SINTEF comments:
None.
 APPENDIX A A-7
SINTEF Materials and Chemistry

21. If you are unsure about failures/leakages of your pipelines, could the uncertainty, seen from a
material specialist point of view, be connected to: (please rate your opinion from 0 (not significant) to
3 (very significant))
AVERAGE
SCORE:
A: Management of pipeline construction projects (3 answers) 1.0

B: Planning of pipeline construction projects (3 answers) 1.0

C: Economic resources (3 answers) 0.0

D: Competence of the owner or contractor (3 answers) 0.7

E: Authority of material specialists in the project (3 answers) 1.0

F: Time from material selection to fabrication of the pipeline (3 answers) 1.3

G: “Late-production” (see question no. 18) (3 answers) 1.0

H: Inexact, diffuse specifications/standards (3 answers) 1.7

I: Incomplete design (3 answers) 2.3

J: Lack of overview (3 answers) 1.7

Technology gap (lack of thorough understanding of failure mechanisms, i.e. lack of
K: 2.3
R&D, relevant test methods and documentation) (4 answers)
L: Monitoring of technology developments (4 answers) 0.5
Uncertainty connected to laboratory tests. Do they represent the actual case? (i.e.
M: accelerated tests, small scale testing, sharp crack versus machined notch etc) 1.8
(4 answers)

SINTEF comments:
It is believed that question K and M are linked as uncertainty to laboratory tests are probably
based on lack of thorough understanding of failure mechanisms.

The scores show that the industry would prioritise:

1. improved knowledge on failure mechanisms and new test methods (question K and M)
2. improve the design basis
3. implement the experience on S13Cr pipelines into standards and specifications
4. more closely evaluate time schedules for pipeline construction projects

22. Do You have comments or additional questions/information?

No.