High Performance Polymers For Oil - Rapra

High Performance POLYMERS for Oil & Gas 2014
Organised by:
Sponsors & Exhibitors:
INNOVATIONS WITH IMPACT
Supporting Associations & Media Partners:
World ils
Crowne Plaza Roxburghe
Edinburgh, Scotland
15-16 April 2014
Organised by
ISBN: 978-1-90930-99-2
© Smithers Information Ltd, 2014
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted
in any form by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written
permission of the publisher, Smithers Information Ltd, Shawbury, Shrewsbury, Shropshire, SY4 4NR, UK.
The views expressed in this publication are those of the individual authors and do not necessarily correspond
to those of Smithers Information Ltd. This publication is published on the basis that no responsibility or
liability of any nature shall attach to Smithers Information Ltd arising out of or in connection with any utilization
in any form any material contained in this publication.
Contents
OPENING SESSION: THE ROLE OF NON-METALLICS IN FUTURE OF OIL AND GAS EXPLORATION
Paper 1 Opening Keynote

Operator Perspective: challenges and polymeric requirements - current and future
John Lawson ETC, Senior Technology Advisor, Chevron
Paper 2 Offshore Drilling: Use of elastomeric materials and future challenges

Nicolas Arteaga, Process Engineer, Cameron paper unavailable at time of print
Paper 3 Case study: Installation of the world's first subsea rehabilitation system based on composite
material
Robert Walters, Founder & Chairman, APS Dubai
CORROSION AND FAILURE: DEVELOPING QUALIFICATIONS AND SPECIFICATIONS FOR FUTURE

INNOVATIONS
Paper 4 Nonmetallic Program for the Oil and Gas industry

- Practical challenges against the utilization of nonmetallics at Saudi Aramco and solutions to
successful applications.
Mr. Abdullah Al-Dossary, Nonmetallic Engineer, Saudi Aramco
paper unavailable at time of print
Paper 5 Rapid Gas Decompression Resistance of Elastomeric O-Rings to Supercritical CO2

Peter Warren, Head of Materials Engineering, James Walker
Paper 6 Industry Standards - A Blessing or a Curse?

Alexandra Torgersen, Engineering Manager, FMC Technologies
Paper 7 CNT Technology and Dispersion

Michaël Claes, Global Technical Director, Nanocyl
UNLOCKING THE POTENTIAL IN NANOTECHNOLOGY FOR THE OIL AND GAS INDUSTRY
Paper 8 Anti-Corrosion Epoxy Coatings Containing Clay in Smectic Liquid Crystalline Order
Dr H.J. Sue, Professor, A&M University
Paper 9 Polymer Graphene Nanocomposites for Oil and Gas Processes

Gobet Advincula, Case Western Reserve University
HARSH TEMPERATURE ENVIRONMENTS - CHALLENGES AND SOLUTIONS
Paper 10 Challenges of Temperature Extremes for Elastomer Materials

Glyn Morgan, Sector Manager, Oil & Gas, Element Material Technology
Paper 11 Cold Temperature Effects on Polymers - Cryogenic Spill Protection

Sebastien Viale, Ph.D., Polymer Specialist - Advanced Subsea Architecture, Technip
Innovation & Technology Center
Paper 12 Arlon 3000XT

Kerry Drake, Senior Scientist, and Burak Bekisli, Scientist, Greene, Tweed & Co
MECHANISMS AND MANIFESTATIONS OF POLYMER AGEING AND FATIGUE, TECHNOLOGICAL

SOLUTIONS, AND FUTURE OUTLOOK
Paper 13 Generation of Polymer Fatigue Data for the Oil & Gas Industry
Andrew Hulme, Principal Consultant, Smithers Rapra
Paper 14 Drilling Fluid Influence on Elastomers

Helmut Benning and Marcus Davidson, Research & Development, Baker Hughes
DEVELOPING AND MODELLING MATERIALS FOR THE FUTURE OF THE INDUSTRY
Paper 15 Modeling and Design of Reinforced Elastomeric Products

Dr Stuart Brown, Managing Partner, Veryst Engineering
Paper 16 Development of Materials for Sealing Solutions in HPHT conditions

Mathilde Leboeuf, R&D manager Europe, St-Gobain Seals
Paper 17 HNBR Rubber in CO2

Daniel L Hertz III, President, Seals Eastern, Inc.
High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
KEYNOTE PRESENTATION
OPERATOR PERSPECTIVE: CHALLENGES AND POLYMERIC
REQUIREMENTS - CURRENT AND FUTURE
John Lawson, Senior Technology Advisor
Chevron
Email: john.lawson@chevron.com;
John Lawson was enticed into the Oil and Gas Industry in 1975, gaining experience
in construction, maintenance, project management, diving and subsea engineering.
Eventually specialised in pipeline design engineering, construction and installation.
John was Texaco’s Operations Pipeline Engineer in Aberdeen from 1998, progressing,
post merger, to Chevron’s Head of Subsea Engineering, leading a team of engineers
working on design and installation, operations, integrity management of subsea
systems whilst concurrently pursuing a strong interest in research & development.
Now Senior Technology Advisor with Chevron Energy Technology Company. Member of BSI, DNV and ISO
Code Committees. Chairs Aberdeen Pipeline Users Group (PLUG) and a variety of JIPs. John is also a
Chartered Engineer, IMarEST Council Member and Membership Committee Member, part time university
lecturer, contributing to undergraduate and post graduate teaching. He has a First class honours in
Mechanical Engineering and a Doctorate in Engineering Design Methodology.
PAPER UNAVAILABLE AT TIME OF PRINT
Page 1 of 2 pages Paper 1 - Lawson

15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014
Paper 1 - Lawson Page 2 of 2 pages

OFFSHORE DRILLING: USE OF ELASTOMERIC MATERIALS

AND FUTURE CHALLENGES
Nicolas Arteaga
Cameron
4601 Westway Park, Houston, TX 77040
Tel: +1 (281) 606-6190 email: nicolas.arteaga@c-a-m.com
BIOGRAPHICAL NOTE
Nic Arteaga joined Cameron in 2005. During his time at Cameron, he has been in riser
engineering, subsea blowout preventer engineering and is currently the engineering
manager with responsibility for Cameron’s Townsend and Guiberson product lines,
Elastomer R&D lab and Elastomer engineering groups. Nic graduated from Texas A&M
University with a degree in Mechanical Engineering and holds a Professional Engineering
License in Texas. On his time off, Nic is a Lieutenant with the Jersey Village Fire
Department, volunteering his time as a firefighter and Emergency Medical Technician.
ABSTRACT
The oil field is a constantly evolving industry, and with new methods come new challenges. In order to
properly design a product and choose the best materials, it is imperative that the designer has a grasp of
the performance requirements of the product. From the rig floor to the sea floor, elastomeric seals are a
critical part of maintaining well control. The seal materials used are predominantly NBR (acrylonitrile-
butadiene rubber), HNBR (hydrogenated acrylonitrile-butadiene rubber) and fluoroelastomers due to their
excellent mechanical properties, as well as their temperature and chemical resistance. These materials can
also be formulated to increase their explosive decompression resistance. Our challenge for the future is to
develop elastomers that 1) are resistant to the effects of new chemicals and chemical combinations; 2) are
resistant to higher temperatures; 3) are resistant to higher pressures; and 4) have improved abrasion
resistance.
INTRODUCTION
The oil field is a constantly evolving industry, and with new methods come new challenges. Of particular
interest here will be the elastomeric materials used in offshore drilling applications. In order to properly
design a product and choose the best materials, it is imperative that the designer has a grasp of the
performance requirements of the product.
Offshore drilling evolved from land drilling, and due to the proliferation of deep water drilling, it has
developed its own specific equipment and methods. Drilling involves tapping into natural resources in the
earth and bringing them up in a controlled manner. A critical element to controlling the fluid is having the
right elastomers in the right places. In this paper, I will cover elastomers used from the drill floor down to
the sea floor, with a particular focus on the blowout preventer (BOP) elastomers.
HISTORY
Drilling for oil began in the 1800’s. As we discovered more uses for this natural resource, drilling became
more common and evolved based on experiences. Soon enough, a common experience came to be
uncontrollable pressure that was allowed to vent to the environment until it was within manageable limits.
Today, this is referred to as a blowout.
In 1922, the first blowout preventer was developed by Cameron Iron Works. This device was essentially a
valve placed on the well to shut it off in case of rising pressure which could lead to a blowout.
Page 1 of 8 pages Paper 2 - Arteaga

Figure 1 – Schematic of BOP submitted by Cameron Iron Works for patent
As offshore drilling developed from fixed structures near land to today’s floating vessels far out at sea, the
need for supplemental methods of control evolved. Now, complicated systems exist that all use elastomers
in some manner.
SCOPE
The scope covered here will begin on the drill floor, which is where the wellbore begins. The wellbore
encloses the column of circulating mud, cuttings and formation fluids. A generic diagram of subsea drilling
equipment is shown in Figure 2, which highlights some particular elements of this scope.
Figure 2 – Generic layout of subsea drilling equipment
Paper 1 - Arteaga Page 2 of 8 pages

The diagram in Figure 2 is not intended to represent a typical configuration, but rather depict many of the
different components used offshore today. For purposes of this paper, these will be further broken into
surface equipment, subsea equipment and blowout preventers. The seal materials used are predominantly
NBR (acrylonitrile-butadiene rubber), HNBR (hydrogenated acrylonitrile-butadiene rubber) and
fluoroelastomers due to their excellent mechanical properties, as well as their temperature and chemical
resistance. These materials can also be formulated to increase their explosive decompression resistance.
SURFACE EQUIPMENT
The spider and gimbal work in conjunction to support the riser as the entire subsea system is slowly
lowered to the sea floor. These components are used while raising and lowering the system and are
removed before drilling occurs, thus they are not exposed to, nor do they ever contain, wellbore fluids. The
spider operates with hydraulic cylinders to move support dogs in and out to hold the weight of the
suspended system. Standard hydraulic cylinders are employed, generally using typical O-rings. The
gimbal’s role is to support the weight of the system and accommodate for pitch and roll of the floating
vessel. These loads may be up to and exceeding one million pounds. To achieve this task, the gimbal is
fitted with multiple elastomeric shock mounts. These shocks include multiple plates molded and bonded
inside the rubber, as well as an upper and lower mount.
Figure 3 – Gimbal with six shock mounts
The rotary table acts as a means to rotate the drill string, although on a modern rig, it is primarily a backup
system to a rotary top drive. The rotary table does not contain wellbore fluids and has no sealing
capabilities. The diverter is the top of the wellbore fluid column and has outlets to direct returning fluids.
The diverter has a critical sealing function in order to redirect the wellbore fluid. A large elastomer packer is
used to seal off the wellbore, similar to an annular BOP. This packer will be exposed to wellbore fluids. For
this reason the elastomer must be compatible with the chemistry of the wellbore fluids. Wellbore fluids
include formation fluids and drilling mud, which are commonly oil-based. The diverter packer must be able
to both open to the full diameter of the bore and close off the bore. This requires very large elongation
properties of the packer—in excess of 200%, sometimes even greater than 400%. This packer is actuated
by a piston that is hydraulically operated, as shown in Figure 4. The diverter must also seal off the outlet
flow lines. This is done using a custom elastomer seal that is energized by pressure applied behind it.
Again, fluid compatibility is critical for this seal that contacts wellbore fluid.
Depending on the rig preference, there may or may not be an upper flex joint. The upper flex joint is
designed to allow minor angular displacement of the riser. The flex joint relies on an internal elastomer
component to take load through it while also sealing during angular offset. Chemical compatibility and
elastomer/metallic bonding are critical factors in the design of the upper flex joint.
During drilling operations, there is up and down heave of the rig. The telescoping joint plays a critical role
by adjusting for this heave by stroking an inner barrel through the outer barrel. The telescoping joint lock is
used during running and retrieving operations and is not exposed to wellbore fluids. The tension ring pulls
tension on the lower half of the telescoping joint (the outer barrel) to keep the system taut. In order to seal
the wellbore fluids between the inner and outer barrel, a sealing system known as the double seal
assembly is provided with a primary seal and a redundant secondary seal. This is a custom elastomeric
sealing element that must have good chemical compatibility with the wellbore fluids, elastomer/metallic
bonding and abrasion resistance for longevity. The seals are energized by applying pressure behind the
seal.
Wellbore
Actuation
Flowline seals
Figure 4 – Diverter cross section
Sealing
elements
Figure 5 – Double Seal Assembly
The goosenecks on the telescoping joint are the beginning of each auxiliary line, which will travel all the
way to the stack on the sea floor. These auxiliary lines each have different functions, with some typical
functions including a choke line, a kill line, a hydraulic supply line, a mud boost line and/or a glycol line. The
functions of the auxiliary lines provided and their orientation relative to the rig are specific to each rig and
its capabilities. Each of these lines, however, will have many connections between the gooseneck and the
stack, with each one requiring elastomeric seals. The seals are generally O-ring energized lip seals with

chemical compatibility and extrusion resistance as critical parameters. For safety, there are two seals at
each connection for redundancy. A complete system could require up to 800 or more of these seals.
Consideration should be given to the operating temperature range in all elastomeric components provided
above the water level. While high temperature resistance is usually considered when choosing a wellbore
seal, low temperatures may also be a factor in surface elastomer applications, as some drilling locations
have temperatures well below freezing.
SUBSEA EQUIPMENT
Equipment below the water level is protected from dramatically cold temperatures, although at extreme
depths the temperature may drop just below freezing. The key characteristics of elastomers in subsea
equipment is obviously chemical compatibility, and for some wells, high temperature.
A rotary control device (RCD) allows the driller to maintain wellbore pressure while drilling, without relying
solely on the weight of the mud column. This is achieved by using a custom elastomeric seal, which seals
around the drill string while rotating on a bearing in the RCD. Abrasion resistance is a critical property of
the seal, as it can allow stripping of the pipe as it runs in and out of the wellbore. Rotary control devices are
becoming an increasingly popular option as they allow the use of managed pressure drilling (MPD) for well
control.
The riser gas handling (RGH) system is designed to allow circulation of a gas pocket out of the wellbore.
This equipment uses a modified annular BOP and packer to seal off the wellbore and redirect the gas. This
packer must have good elongation properties to be able to open fully to clear the wellbore, while also
closing completely to shut off the wellbore. Typical annulus diameters of the riser for deep water rigs are
approximately twenty inches.
The riser joints make up the majority of the length of the complete system, with connections at the ends of
each one that require seals. A seal sub is used, which employs redundant O-ring energized lip seals on
both ends.
The subsea flex joint is similar to the upper flex joint, in that its purpose is to accommodate angular offset
without damaging equipment. Like the upper flex joint, this component uses a large custom elastomer
component to withstand the loading while maintaining a pressure seal.
BLOWOUT PREVENTERS
The subsea stack is comprised of both annular and ram type blowout preventers. The number of BOPs and
the configurations vary from rig to rig based on the end user’s requirements.
The function of a BOP—whether it is annular or ram style—is to control the wellbore pressure. In order to
control the wellbore pressure, BOPs use elastomer elements to seal off the bore.
An annular BOP has a specially designed packer with metal inserts molded into and bonded to the
elastomer. When the BOP is open, the full wellbore diameter is clear for downhole tools to pass through.
When functioned, the annular BOP can either close around pipe in the bore or close completely if nothing
is passing through the bore. In order to achieve this feat, the packer elastomer must have very large
elongation properties—up to 400%—yet it must be able to retract completely to clear the bore. The design
of the metallic inserts molded into the packer are also of particular interest as they are critical for reducing
the extrusion gap once closed in order to hold wellbore pressure. The annular packer is also used for
stripping operations, in which the packer is closed around a pipe, holding wellbore pressure and the pipe is
then moved either into or out of the wellbore. The durability of an annular packer during stripping is related
to its abrasion resistance, the wellbore pressures being maintained and the closing pressure being applied
to the packer. The Cameron DL annular BOP uses two elastomer elements as shown in Figure 6. The
Donut is a solid elastomer without inserts, whose function is to translate the vertical movement of the piston
into a radially inward motion, pushing the packer. The packer inserts of the DL are in the shape of an iris to
close evenly around tubulars in the wellbore and keep the extrusion gap at a minimum.
The ram type BOP is available in many different configurations and options. The ram style BOP uses rams
designed for a specific function. The rams are actuated by hydraulic pistons and have hydraulically
actuated locks to maintain the ram position in case of hydraulic failure of the piston. In a typical stack, there
are multiple ram BOPs, often five or six, in order to accommodate all the different ram types required.
Figure 7 shows a few common ram styles. The blind shear rams (BSR) are designed to shear pipe in the
wellbore, and subsequently seal, and hold wellbore pressure. They are referred to as “blind” shear rams
because they may also be used to seal off the wellbore if there is nothing in the bore. A pipe ram is used to
seal around a fixed pipe size and only that size. It is common for pipe size to change during the course of
the drilling program, which is why a variable bore ram (VBR) is a popular product. The VBR has inserts in
an iris pattern similar to the annular packer, which allows it to seal on a range of pipe sizes. The flexpacker
is similar to the VBR, with the exception that is can only seal on certain pipe sizes within its range.
Figure 6 – Cameron DL Annular BOP
The design of the ram packers requires several inserts bonded and molded into the elastomer. When the
ram blocks are functioned by the BOP, they bring the packers face to face to seal across the bore and then
also have side packers and a top seal to seal around the wellbore. Ram BOPs are the first physical line of
defense to shut off the wellbore against rising pressure that needs to be controlled. High temperature
packers are becoming more common, along with the need for resistance to higher Hydrogen Sulfide (H2S)
concentrations.

Figure 7 – Cameron BOP Rams
FUTURE CHALLENGES
Although we cannot be certain what the future will bring, we do know that ideas and drilling methods are
constantly evolving. In order to keep up with technological advances, there are some areas where
elastomeric improvement could be of great benefit.
Chemical compatibility is an ever changing and constant battle. While it may be simple to test samples
against a new chemical, we cannot know the combined effect of the mixture of chemicals until they are
tested, which most often only happens in the field. In particular, the level of Hydrogen Sulfide found in
wellbore fluids continues to be a challenge due to the fact that it can degrade the performance of the
elastomer.
What is considered “high temperature” is constantly being redefined to higher and higher limits. Steam
injection pushes the current boundaries, reaching over 500°F. For most current seals, this is already
beyond their limits. While high temperature extrusion resistance is a major goal for future elastomers,
equipment manufacturers must also consider the design of the seal and the sealing cavity, which may need
to change to reduce extrusion gaps.
Abrasion resistance continues to be a challenge for those particular seals in operations where there is high
abrasion. As the cost of downtime for seal replacement increases, the research to develop more resistant
elastomers must continue.
Similarly, working pressures requirements are slowly increasing. Typical current equipment is designed for
15,000 psi maximum wellbore pressure, but requests for as high as 25,000 psi have been made. The
tensile strengths of sealing elastomers will need to follow suit, keeping in mind that test pressures are
typically 1.5 times the working pressure.
These four functional limits—chemical compatibility, temperature limits, abrasion resistance and pressure
limits—represent some of the most visible areas in which technological advancements will have a major
impact on drilling operations. The goal is to, first and foremost, improve safety—protecting lives and the
environment—and then improve efficiency. I believe there is always a better way and we must continue to
be relentless in our research for the sake of the future.

REFERENCES
Bommer, P. (2001). A Primer of Oilwell Drilling. Austin: The University of Texas at Austin
Salem, H. & Zonoz, R. (Oct. 2013). Explosive Decompression of Elastomeric Materials in Oil & Gas Sealing
Applications. Fall 184th Technical Meeting. Lecture conducted from American Chemical Society, Cleveland,
OH.

IFL™ - A NOVEL APPROACH TO THE REHABILITATION OF

SUB-SEA HYDROCARBON PIPELINES USING HIGH
PERFORMANCE SOLEF PVDF FLEXIBLE KEVLAR
REINFORCED LINERS
Robert A Walters, IFL Global Project Director
APS Dubai
Email: robertwalters@apsdubai.com
BIOGRAPHICAL NOTE
Robert Walters is the founder and Chairman of Anticorrosion Protective Systems, a

global group of companies which specialises in pipeline corrosion engineering
services, particularly focussed on pipeline coatings, linings and rehabilitation systems.
His company, APS, now owns, licenses and operates one of the largest portfolio of
pipeline inspection and rehabilitation technologies in the Middle East and Asian
regions and offers services ranging from the turnkey installation of PE pipelines to
large diameter spiral wound UPVC, GRP liners and cured in place systems.
ABSTRACT
In common with many offshore operators, PETRONAS Carigali (PCSB), own and operate an extensive
network of sub-sea pipelines which are situated in the vast offshore oil-fields which span the South China
Seas.
Many pipelines run from platform to platform and platform to onshore facilities over distances of between
several hundred meters to several kilometers, in varying water depths.
Internal corrosion, due in large part to sulfate reducing bacteria (SRB) has historically caused aging pipelines
to have a relatively short life, resulting in regular and expensive replacement cycles, requiring the
deployment of significant lay-barge and marine spreads.
In an effort to reduce long term expenditure, PCSB have invested heavily in the development of a reinforced
high-performance liner system which can be easily and rapidly deployed platform to platform within their sub-
sea pipelines in-situ, over long distances thereby providing a viable corrosion resistant & rehabilitation
system without the need for conventional lay barges.
The system, known as InField Lining, or IFL™ is the culmination of one of the largest research and
development projects undertaken in recent years in the pipelining industry, representing over three years of
effort, which has produced a Kevlar reinforced flexible hose with an inner Solef PVDF layer providing high
end temperature, pressure and hydrocarbon performance.
This paper will describe the research, development and subsequent application of the IFL™ System
including its deployment during 2013, platform to platform in subsea flowlines situated within the Samarang
field for Petronas.
THE DEVELOPMENT OF A VIABLE REHABILITATION SYSTEM FOR DEPLOYMENT IN NEW AND

EXISTING SUB-SEA PIPELINES BY PETRONAS CARAGALI SDN BHD
INTRODUCTION
PETRONAS Carigali (PCSB) is the owner and

operator of an extensive network of sub-sea
pipelines which are situated offshore from main
land Malaysia, in the South China Seas.
Figure 1. Onshore Prototype
Testing of IFL™ Liner
Page 1 of 6 pages Paper 3 - Walters

Many of these pipelines run from platform to platform and platform to onshore facilities over distances of
between several hundred meters to several kilometers, in varying water depths.
It is recognized that internal corrosion, due in large part to sulfate reducing bacteria (SRB) can cause the
pipelines to have a relatively short life cycle which has historically resulted in the replacement of pipelines
becoming necessary within a time period as short as four years.
The relatively short life cycle and frequent replacement requirements are representative of substantial capital
expenditure for PCSB and PCSB therefore desired that the means and mechanisms for the in-situ placement
of a corrosion barrier to be developed that could then subsequently be successfully deployed for use in
existing and new pipelines, thereby providing substantial reductions in capital expenditure on new lay
replacement pipelines & the ability to substantially elongate the life expectancy of their existing pipelines.
Historically, there has not been a viable methodology that could be utilized to install such a corrosion barrier
to within a sub-sea pipeline, thus this project for the Design and Development of Infield Liners (IFL™) was
instigated.
THE MISSION STATEMENT:
To develop, implement and make globally available, a practical resolution to the economic and
environmental risks created by the internal corrosion of sub-sea pipelines and to do so in a manner
that constitutes a significant technical and commercial advancement for the benefit of the
international off-shore pipeline industry, by henceforth providing an effective and practical option to
that of pipeline replacement.
The project began in April of 2011 and operating under the joint management of Petronas and Anticorrosion
Protective Systems, who are globally recognized pipeline rehabilitation specialist engineers and contractors,
it has been possible for the project team to deliver a substantially market-ready product within a two year
time-frame and in line with the original estimates and budgets.
IFL™PROJECT OBJECTIVES
The IFL™ research and development project has been squarely aimed at realizing the primary objective of
developing the materials and technologies necessary to successfully implement the installation of plastic
liners to existing and new sub-sea carbon steel pipelines being operated by PCSB and other Petronas
Companies, for the conveyance of corrosive hydrocarbon media, where SRB is one of the principal sources
of corrosion activity. The IFL™ liner will protect the internal pipe bore from corrosion of any kind and will also
offer a secondary containment capability in the event of a rupture or damage to the outer steel pipeline.
The project start point has been the testing and qualification of an existing nominal eight inch Kevlar
reinforced plastic liner product, which is produced and manufactured for the utility market by a German
company, Raedlinger. This material was selected as a good starting point for the project development work
as it was recognized that although it not previously used for the purpose of lining sub-sea pipelines, the
general liner matrix does demonstrate many of the physical attributes that are perceived as being necessary
to contribute toward the likely requirements for success, such as:
- High tensile and good physical properties.

- Moderate chemical resistance.
- A high degree of flexibility.
- The ability to be manufactured and spooled in long lengths.
LINER QUALIFICATION AND DEVELOPMENT PROCESS
The qualification of the liner has been generally undertaken in accordance with the API Recommended
Practice 15S (First Edition March 2006) “Qualification of Spoolable Reinforced Plastic Line Pipe”, with further
reference to the applicable ASTM test standards, API 17 series and Nace standards. The testing and
qualification procedures have been undertaken in a number of locations including Germany, Norway and the
UAE.
Investigations have been undertaken to product having both single and double layers of Kevlar reinforcing in
the liner matrix.
Paper 3 - Walters Page 2 of 6 pages

During the course of the project, the investigation and development process has been successfully
undertaken whereby various reconfigurations of the liner manufacturing process have been implemented
comprising of the following:
- Alteration of the type of plastics liner and/or outer jacketing, so as to provide improved performance
capabilities (permeability, chemical resistance, temperature resistance etc).
- Alteration of the fiber reinforcement, so as to provide improved tensile capabilities.
- The production of a non-standard diameter, so as to provide a liner that will provide a close fit inside
of the host carbon steel pipeline into which it will be inserted.
The final enhanced IFL™ Liner matrix comprises of a

solvay Solexis PVDF inner liner, a tightly woven Aramid
core, using Dupont Kevlar fabric, with an our layer of
abrasive resistant Thermoplastic Polyurethane from
BASF. Other versions of the liner are also available for
less aggressive service conditions, such as water
reinjection and gas transmission.
Figure 2. IFL™ Liner matrix
As at the date of this publication, it would be true to say that all the principal objectives and milestones of the
project have been totally fulfilled. A new enhanced version of the IFL™ liner has been developed.
Performance testing has been undertaken which has been able to completely justify the utilization of IFL™ in
very aggressive, hot, sour hydrocarbon service conditions of up to 120 degrees centigrade, with IFL™ liners
exhibiting a stand-alone burst capability of up to 120 Bar.
The IFL™ liner design and development process has also encompassed the investigation and
implementation of changes to the existing methods and mechanisms for terminating the liner at a flange
interface as well as investigating the methodology by which long lengths of liner coils could potentially be
joined together.
This could be of significant interest to the global market as a basic and enhanced IFL™ product versions are
effectively now available for use which between them cover a wide range of operating conditions.
Predictive Installation software has also been developed that will

enable accurate calculations of the tonnages required for the
installation of the IFL™ liner and hence determine if the lining of any
given pipeline length and configuration is in fact viable. This software
model has been extensively tested and calibrated during over 60 full
scale prototype trials when actual towing loads have been checked
during IFL™ liner installations to full scale mock-up pipelines running
from platform to platform.
The IFL™ Liner system is initially available in diameters suitable for

the rehabilitation of pipelines from 6 inch to 20 inch.
INDUSTRY MOTIVATION TO UTILIZE IFL™ LINERS
The majority of sub-sea pipelines are constructed from carbon steel,

laid by barge lay, during which single or double random joints
of steel pipe are welded together on the deck of the barge and
gravity laid onto the seabed. After completion of the welding process,
crews on the barge can “make-up” the external corrosion protection
and “infill” the missing concrete protection because this is easily
accessible. It is not however possible to “make-up” any damage that
may be caused to any internal coating by the welding process, or
Figure 3. IFL™ Riser Flange Liner “infill” any cut-back to the internal coating that would be necessitated
Termination Connector
so as to facilitate the steel weld.

To compensate for this it is common for most sub-sea pipelines to be laid without an internal coating and an
additional wall thickness of sacrificial steel to be added to the design so as to compensate for the calculated
rate of corrosion throughout the design life of the pipeline.
Unfortunately however, corrosion is rarely a linear phenomenon and certain types of corrosion can cause
damage to the pipe wall much more quickly than was allowed for at the design stage. Pitting, grooving,
cracking or crevicing to the interior pipeline wall can occur in a remarkably short period of time, such that, for
instance, a pipeline installed with a twenty-year design life, may experience failure after as little as four years
in service.
In summary, IFL offers the pipeline industry a viable, fast, economical option to new-lay pipeline
replacement.
IFL™ can be utilized for the rehabilitation of an existing sub-sea pipeline

where:
- It is desirable to extend the service life of the pipeline beyond the
period of operation for which it was originally designed.
- Unforeseen operational parameters such as CO2 or SRB corrosion

have caused the pipeline to reach the end of its useful life ahead of
the originally intended schedule. The pipeline may or may not have
at that point, already been shut down and abandoned for safety
and/or environmental reasons.
- Routine inspection of the pipeline has shown that greater than

Figure 4. IFL™ Liner in
anticipated corrosion is taking place and that unless the corrosion
tight fit configuration is arrested, the pipeline will fail at some predictable point in the
future, at a time which is less than the design life.
- In instances where pipelines have been decommissioned or

abandoned due to integrity related issues.
The IFL™ predictive software is utilized to determine if any specific pipeline is a suitable candidate for an
IFL™ liner installation. Overall pipeline liner lengths that can be achieved are dependent upon the pipeline
diameter, configuration and number of short radius bends, but trials would indicate that the rehabilitation of a
typical 6 or 8 inch diameter hydrocarbon flow line could be feasible over distances of up to 10 kms.
The replacement of a pipeline by the process of designing and laying of a new one and the abandonment
and/or removal of the old one is normally representative of a major engineering, procurement and installation
campaign and an equally major capital expense.
The insertion of an IFL™ liner into a defective pipeline is perceived as being a process of a far lesser
magnitude in terms of planning, implementation and expense. It may well be possible (in terms of project
turn-around) to achieve with IFL™ in weeks, what may otherwise take months or even years, with
conventional pipe lay replacement, especially if the necessary pipe-lay barges for conventional barge lay are
not located in the region.
OUTLINE IFL™ INSTALLATION PROCEDURE
A thorough inspection of the existing sub-sea pipeline prior to the detailed planning of any IFL™ Liner
rehabilitation project is a mandatory pre-requisite, as is the collation of all data relative to the prevailing
operating parameters and conditions. Inspections can be carried out by intelligent pigs or other external
remote inspection tools such as the MTM Aqua.
This data is used to assess the general condition and remaining wall thickness of the existing pipeline and to
verify the IFL™ Liner size requirements in the event that an enhanced tight-fit high pressure liner is required.
At lower pressure and less arduous conditions, where flow capacity allows, the IFL™ system may be utilized
in a reduced diameter loose fit format.
Prior to the offshore deployment of the IFL™ Liner installation marine spread, the host pipeline will have
been decommissioned, cleaned and finally gauged ready for the liner insertion.

The IFL™ Liner material, although manufactured in a circular profile, is able to be temporarily flattened for
transportation and reeled onto a transportation drum that can be sized so as to fit into conventional shipping
containers. Each drum can be loaded with up to 5 kilometers of IFL™ liner, dependent upon the liner
diameter.
These drums are then shipped to an onshore location within the destination country, usually a marine supply
base, where they are further processed into a folded liner format prior to finally being sent offshore to the
platform location for installation.
Figure 5. Liner Shipping Reel Loaded with Liner. Figure 6. IFL™ Liner in Folded Form Ready for Offshore Deployment.
The actual IFL™ Liner installation process is extremely fast, being operated at speeds of approximately 10
meters per minute, hence providing for the insertion of a typical two kilometer liner in a period of no more
than 3.5 hours.
The IFL™ Liner drum is, wherever practical positioned on the offshore platform structure, or when
necessary, on the deck of a work boat from where the liner can be unspooled using the Exd. powered drive
mechanism equipped on the liner reel.
A feeder cable will have been fired through the pipeline during the final cleaning and gauging procedure and
this is used to pull back through the liner installation winch cable for connection to a towing head which is
located on the leading end of the liner.
During the engineering Phase of the pipeline rehabilitation

project, the specific winching loads necessary for the liner
insertion are carefully analyzed using the proprietary predictive
IFL™ software. The winch packs used for the actual
installation process are equipped with load cells and over-ride
devices so that in the event of greater than predicted loads
being experienced during the winching, the operator is alert to
the situation and the devices can be set so as to automatically
cut out at a given load if the engineered safety factor relative
to the liner yield strength is approached.
In reality, for most liner insertion situations in the 0.5 to 5 km

range, the insertion forces are no more than one tenth of the
liner tensile yield strength.
Figure 7. Winch Being Positioned on Platform at

Receiving End of Host Pipeline

Figures 8 & 9. IFL™ Liner Entering Platform Riser.
Prior to the liner installation, the IFL™ end termination coupling devices are installed at the riser flange
locations. Once the IFL™ Liner has been drawn through the entire pipeline length, it is then re-rounded by
filling with either air or water. The liner, which is manufactured to the same diameter as that of the host
pipeline bore, then expands to form an intimate fit with the inner wall of the host pipe.
Figure 10. IFL™ Liner Inflation Procedure.
With the liner then fully re-rounded against the wall of the host pipeline, the last task is the installation of the
end termination inserts which ensure reliable compression seals and restraint at the liner ends. The re-lined
pipeline can then be hydrotested in the conventional manner and the all top-side pipe work reconnected,
following which the pipeline is then ready for re-commissioning and for its new, extended life of operation.
CONCLUSION
The IFL™ Development Project has successfully achieved its primary mission goal, having delivered
Petronas a viable alternative to the replacement of deteriorated offshore pipelines.
It is anticipated that Petronas will henceforth favour the option of pipeline rehabilitation over that of new-lay
pipeline replacement and in so doing will be able to drastically reduce their offshore operational cost base.
It is further anticipated that the IFL System will become globally available by way of a Global
Commercialization and Franchising arrangement during the course of 2014.

NONMETALLIC PROGRAM FOR THE OIL AND GAS INDUSTRY

Abdullah Al-Dossary
Nonmetallic Engineer, Saudi Aramco
Email: abdullahdosary@yahoo.com
BIOGRAPHICAL NOTE
Abdullah Al-Dossary is a Nonmetallic materials engineer from Saudi Aramco

since 2004. Certified by the National Association of Corrosion of Engineers
(NACE) as a Corrosion Technologist. Abdullah holds a Msc degree in polymer
engineering from the Institute of materials engineering and polymer technology
in UK. Holds a Bachelor of science degree in Mechanical engineering from
Pennsylvania State University.
ABSTRACT
The presentation will cover the practical challenges/barriers against the utilization of nonmetallic and Saudi
Aramco solutions to successful applications. The nonmetallic program is a cornerstone for a comprehensive
and leading corrosion management program.
PAPER UNAVAILABLE AT TIME OF PRINT
Page 1 of 2 pages Paper 4 - Al-Dossary

Paper 4 - Al-Dossary Page 2 of 2 pages

RAPID GAS DECOMPRESSION RESISTANCE OF

ELASTOMERIC O-RINGS TO SUPERCRITICAL CO2
Peter Warren, Steve Winterbottom and Andrew Douglas
James Walker & Co. Ltd
Cockermouth, Cumbria, CA13 0NH
Tel: +44 (0)1900 898277 email peter.warren@jameswalker.biz
BIOGRAPHICAL NOTES
Peter Warren is Head of Materials Engineering at James Walker & Co Ltd and has 38
years of experience in the industry. He has a broad knowledge of sealing materials,
though his specialism is elastomer technology and its relationship to applications.
Peter is a Fellow of the Institute of Materials, Minerals and Mining, a Chartered
Engineer and a Chartered Scientist
Steve Winterbottom is Senior Development Technologist at James Walker & Co Ltd and has 45 years
experience in the industry. He is an elastomer expert whose current speciality is compounding and materials
development. Steve is also a Fellow of the Institute of Materials, Minerals and Mining and is a Chartered
Scientist
Andrew Douglas, a graduate of Heriot - Watt University is the Laboratory Manager at James Walker & Co
Ltd and has 20 years experience within the company. His expertise is materials testing in relation to
applications and, in particular, with Rapid Gas Decompression testing.
ABSTRACT
Enhanced Oil Recovery (EOR) can use CO2 to increase well pressure and reduce the viscosity of the crude.
When combined, the yield can increase considerably. Using this process the level of CO2 in the recovered
crude can be greater than 60%, and high levels of CO2 are known to be damaging to elastomeric seals in
high pressure systems that are subjected to rapid depressurisation. This paper describes a programme of
work that evaluated the rapid gas decompression (RGD) resistance of HNBR and FKM elastomers to CO2,
and varying levels of CO2/methane mixtures at a range of temperatures. The paper compares and contrasts
the effects of the gas on the polymers, and the values obtained from the RGD testing. It concludes with an
assessment of the possible limits for each material in terms of temperature and CO2 level when subjected to
rapid depressurisation.
CO2 Properties
The surface tension of saturated CO2 decreases with increasing temperature and becomes zero at the
critical point. In the majority of applications this means that the surface tension is zero and the viscosity is
close to zero. This explains why it is so invasive.
CO2 is a very efficient solvent and this property becomes more pronounced as the gas becomes a
supercritical liquid. This characteristic is of paramount importance when considering elastomers for use in
CO2. Under decompression there is a step change in enthalpy and density as pressure reduces to the liquid
vapour line. For the CO2 to transform from liquid to gas, heat must be added in the same way as heat must
be added to convert liquid water into steam.
Above the critical temperature there is no noticeable phase change, hence when the pressure is reduced
from above to below the critical pressure, a smooth enthalpy change occurs from super critical fluid to gas.
Pure CO2 has a triple point at - 56.6 ºC and 5.18 bar, which determines the point where CO2 may co-exist in
gas, liquid and solid state. At the right combination of pressure and temperature CO2 may turn into the solid
state commonly known as dry ice1.
Page 1 of 10 pages Paper 5 - Warren

CO2 has been used as a refrigerant, commercially, since 1869 and has been piped successfully for about
forty years. There is therefore a lot of data available on the problems associated with CO2 but there is only
limited detail on the use of elastomers. Where information is available this tends to be in relation to the
refrigerant use.
Fig. 1. Carbon dioxide pressure – temperature phase diagram (Finney and Jacobs)
Effects of CO2 Immersion on Elastomers
There are two main consequences of immersion in CO2 both of which are intensified where the CO2 is in a
supercritical state. The first is one of swell. This occurs in the low pressure state as well as in the
supercritical situation. This swell may also coincide with plasticisation with the resultant softening. The
second effect is that of susceptibility to rapid gas decompression damage. CO2 is far more likely to cause
RGD than the other gases which are common in the oil and gas industry.
Swell
Swell has been measured in subcritical, critical and supercritical phases by numerous people over the years
as well as ourselves. All the reported testing to date however is in an unconstrained situation as with a
restrained situation measurement is not practical.
Danny Hertz III reported his evaluations of swell in CO2 at the 2012 Smithers High Performance Elastomers
& Polymers for Oil & Gas Applications conference2. The pressure he employed was 750psi and testing was
at room temperature (testing being under subcritical conditions only). He found that there could be high
levels of swell under these conditions and that it was polymer dependant primarily and formulation
dependent secondly.
The 2010 paper by Dr Hans Maag, Achim Welle, Dr Matthias Soddemann and Dr Kevin Kulbaba reported
their findings for CO2 immersion3. They also measured swell on a number of compounds which were
exclusively HNBR based compounds. Their initial observations were using 7.5MPa and 20°C. They also
observed swelling at this subcritical phase. They then increased the temperature to 31°C (Critical) and
100°C (supercritical). When white fillers were used in the compounds it was noticeable that as the
temperature increased so did the swelling. These differences due to the conditions however were not
dramatic and, with the black compound, swell was worse at the subcritical level. Differences were however
minor. Again it was shown to be compound dependent and, in this case, dependent on the ACN level.
Paper 5 - Warren Page 2 of 10 pages
The 2012 presentation by Bjørn Melve is in relation to work carried out by Statoil to evaluate swelling and the
consequences of such4. The first assessment was using HNBR in CO2 at 150 bar and 72°C. This showed
minimal swelling and no blistering. The material used for shear rams was also evaluated and in this case
swelling was far more pronounced. After extensive evaluations of swell in cycle testing it was decided that
the majority of swell was during decompression. In application the swelling was occurring when components
were brought to the surface. In essence the swell may be limited by the pressure.
These papers and presentations were measuring linear change using fairly crude but effective methods. The
evaluations by MERL (now Element) presented by Sabine Munch used a very sophisticated sapphire
window and cameras5. This enabled accurate measurement of changes in section. It did however use a thin
section giving a large surface area in comparison to thickness. This may mean that times of swelling and
recovery are not typical of application. The effects however would be comparable for the majority of
applications.
Using a FKM type 1 material the section was subjected to 100°C and the pressure was ultimately 300 bar.
The expansion as the gas was introduced under pressure gradually reached 26% after 18 minutes and then
stabilised. This time would be increased somewhat with a complete o-ring of this section. Upon release of
the pressure the sample did not expand initially unlike the samples in the previous papers. It did however
contract back to the original size after the pressure had reduced to ‘zero’. A second test using less than 100
bar did however expand prior to final contraction.
The final paper which gives significant information in relation to swell in CO2 is that of Zoltan Major et al6.
Although primarily looking at test methodology for rapid gas decompression it has a lot of data on swell of
both HNBR and FKM. The measurement on this occasion was by CCD camera in an autoclave. Although
testing was undertaken on both restrained and unconstrained samples the measurement was only
conducted on the latter. The testing is interesting as it in effect compares CO2 with methane for the FKM
material. All the tests were completed using the 100°C temperature associated with the Norsok M-710
methodology. Pressure was at 140 and 280 bar. The 140 bar CO2 results for the FKM compound show initial
expansions of 8% in height. The peak expansion under decompression was 65%. For the 280 bar testing it
was 16% and 160% respectively. As with earlier work the maximum expansion was at minimal pressure
following decompression and contraction occurred after this peak. This peak will be the point at which the
supercritical liquid reverts to the vapour phase. The time to contract was surprisingly short taking about one
minute in each case.
When the gas was changed to methane the expansion was much lower at just over 2% and the peak on
decompression was 60%. The decay from the peak expansion was slightly slower. The HNBR was only
tested using CO2 at 280 bar but showed a significantly lower swell in comparison with the FKM. Figures were
6% and 30% for initial and peak expansions.
Rapid Gas Decompression.
Each of the papers we have quoted was undertaken with rapid decompression in mind. The swell during and
after decompression, however, indicates the effects when unconstrained. It may be that expansion can be
related to RGD performance but our interest was in the actual RGD testing which we could then potentially
relate back to these and our studies.
The evaluations mentioned earlier did consider RGD performance but only the Major et. al. work performed
tests on O-rings. The other tests were performed on cut samples. It is difficult to assess the results using this
type of test sample and, in particular, materials cut from test sheets which may contain voids.
The paper by Morgan, Sully and Davies (1997) evaluated cycle tests in CO27. Although this pre-dates
NORSOK/ISO tests it is nevertheless applicable to our studies. This testing was completed at 40 bar as that
was the application requirement. The majority of the work was undertaken using silicone rubber seals but
nitrile rubber and FKM were also evaluated. The paper discusses two phenomena, which are temperature
cycling and pressure drop. They found that temperature cycling at a given pressure could create what they
called thermal decompression. In their particular tests the nitrile rubber had better performance than the FKM
and the two silicone rubbers tested produced markedly different results. This is likely to be related to the
hardness as the best results were from the 80 hard (IRHD) version and the worst from the 60 hardness
version.
The previously mentioned work by Major et al included O-ring testing as part of the evaluation. Testing was
constrained and unconstrained. The testing was in two modes, single cycle and multi cycle. It would be
difficult however to draw any conclusions about RGD performance in CO2 from this work although there was
a tendency for unrestrained samples to have a greater extent of damage. There were no instances of
undamaged seals and there was evidence of initiation points which may have affected the comparative
results.
RGD and Swell Testing – James Walker Materials
It was obvious that a lot of work had been undertaken over the years using CO2 as the medium. We,
however, decided that we needed to fully understand the effects of CO2 on our materials compared to those
when using orthodox NORSOK/ISO media. This may also permit a comparison of polymer types but that
comparison may only be applicable to our materials and may not apply universally. For our investigations we
tested five of our materials which have already been qualified against NORSOK M710/ ISO 23936-2 RGD
criteria.
We do have comprehensive test rigs dedicated to RGD testing. This includes 4 off 8 port flange rigs for
conventional testing and a Bomb test rig for development purposes. This testing was undertaken using the
flange rigs with the bomb rig being used for comparative unrestrained expansion data only. The 8 cycle ISO
test was performed in preference to the 10 cycle NORSOK test for reasons of rig availability and the high
number of tests involved. In order to find the limits of each material testing was undertaken at 23°C, 50°C,
75°C, 100°C and 150°C. Pressure was 150 bar and decompression rates and soak times were defined in the
ISO Standard.
Figure 2, The eight port flange rigs
This testing has taken place over a three year period using a variety of materials. The detail given refers to
examples of material types by selection of RGD resistant grades. Although a greater number of materials
have been tested, including different hardness, our initial interest has been to document our RGD resistant
compounds, primarily for prediction purposes. All testing was at nominally 14% compression and 84%
groove fill. This testing is ongoing.
As expected the damage was more severe than with the standard 90:10 CH4/CO2 testing. The results were
also fairly well in line with the proportional capabilities of the individual compounds when using the standard
90:10 CH4/CO2 mixture. This suggests that the resistance in CO2 could be reasonably predicted based on
the results obtained during NORSOK/ISO tests. This is not just the published NORSOK passes but the full
testing undertaken on a large variety of sizes, temperatures and pressures. In reality it may not be as simple
as that as the swelling may also need to be considered. Evidence so far however suggests that the degree
of swelling may not be a good indicator of RGD behavior in CO2. A swollen material will be plasticised to
some extent which may affect the performance for better or worse.

Figure 3, The Bomb Rig which accepts a flange assembly or unconstrained samples
For more information about the comparative swells of the materials we had been testing for RGD resistance
we organised additional swelling tests performed by Element using the sapphire window technique. The
HNBR and peroxide cured terpolymer based materials were chosen for this evaluation. Samples were cut
from test sheets into discs and were then measured. They were placed, unconstrained, in the rig and then
heated to the test temperature of 100°C. The volume change due to thermal expansion was noted at this
point. Pressure was increased to 150 bar using CO2 and further measurements taken. After 24 hours the rig
was decompressed and the size changes plotted in comparison with pressure decay.
Material Thermal expansion Pressurized sample Maximum size during

decompression
Thickness Volume Thickness Volume Thickness Volume
Change % Change % Change % Change % Change % Change %
FKM 1.2 3.6 6.7 21.4 9 29.7
HNBR 1.0 3.1 4.3 13.3 6.6 21.1
Table 1, Summary of comparative swelling behavior
These volume changes seem quite high at first glance but were tested unconstrained. The differences
between the two polymer types was as predicted but the differences are less than expected based on
previously reported testing. As with the earlier work we referred to the maximum expansion as following the
decompression. The increase in size however did occur earlier with the FKM in comparison with the HNBR.
As most o-rings would have an 85% or lower groove fill the expansion from thermal and CO2 swelling may
not be sufficient to completely fill the housing or would be slightly restricted in expansion by the housing.
There would therefore not be a significant increase in force. Under decompression there will be a greater
force generated as the housing will restrict the expansion to a higher degree. This is however a very short
term expansion and decay is rapid taking less than five minutes. With weaker materials this could damage
the seals. In an unconstrained situation a 30% swell is not sufficient to damage a tough material such as an
RGD resistant O-ring. Again we must stress that these results are from a limited number of our materials and
it cannot be assumed that other materials will behave in the same manner.
RGD Test Results
The material which has the best RGD resistance (test results of 10mm+ section passing NORSOK M710)
with conventional gases is undamaged at 150°C in pure CO2. This seems a remarkable result but is
proportional to the results in 90:10 CH4/CO2. Table 2 indicates the “safe” temperatures for individual
materials as they currently stand for 329 size O-rings where NORSOK ratings are ‘0000’. Further testing is
underway to fill the gaps in our knowledge. We also will be considering the effect of groove fill and
compression on the results.

Figure 3, Typical failure mode in CO2 of splits but no blistering.
10% CO2 50% CO2 90% CO2 100% CO2

Bisphenol cured FKM
Terpolymer 100ºC (A) 50ºC 50ºC 50ºC
Peroxide cured FKM
Terpolymer 150ºC (A) 150ºC (A) 150ºC (A) 150ºC (A)
Low Temperature Type 3
FKM 100ºC (A) 75ºC 50ºC 50ºC
Low Temperature Higher
Fluorine Type 3 FKM 100ºC (A) 100ºC (A) Under Test Under Test
HNBR 100ºC (A) 50ºC 50ºC 50ºC
A - Tested up to this temperature only, limit unknown
Table 2, Summary of temperature limits with various gas mixtures for O-rings of 5.33mm cross-section
Having established the performance at 5.33mm our next evaluations involved testing materials NORSOK
M710 / ISO 23936-2 qualified to 6.99mm section. As the section increases it becomes more difficult to
withstand RGD damage. We were interested to see if the results followed the same trend the same when
employing CO2 as the gas. Seal O/D as 329 size.
Material 10% CO2 50% CO2 100% CO2

Peroxide cured FKM 125°C (A) 100°C (A) 100°C (A)
Terpolymer
Low Temperature Type 100°C (A) 75°C 50°C
3 FKM
High Temperature 100°C (A) Not yet tested Not yet tested
Higher Fluorine Type 3
FKM
HNBR 100°C (A) 50°C 50°C
A – Tested up to this temperature only, limit unknown
Table 3, Summary of temperature limits with various gas mixtures for O-rings of 6.99mm cross-section
Interestingly the results for the HNBR and low temperature type 3 FKM performed as well at 6.99 as they
had at 5.33mm. This is only a small evaluation and will need extending to get the full picture. This situation,
where section size is not as related to failure, may be due to the high explosive energy as supercritical fluid
changes to vapour. The following refers to decompression due to storage containment failure but may well
apply to deliberate decompression. “The explosion energy can be estimated as the energy released by
expansion of the refrigerant contained in a component or system In case of a component rupture, however,
the explosion energy (stored energy) may characterize the extent of potential damage. The expansion
process will be very rapid, with little or no time for heat transfer between the ambient air and the expanding
gas, and the explosion energy can therefore be estimated as the reversible adiabatic work of expansion”9.
The test programme is ongoing, and through this we are gaining a better understanding of the factors
influencing RGD performance, which will allow improvements leading to increased capabilities for given
materials over time.
Summary
How does RGD behaviour correlate with the swell results? Based on the test data the higher the swell the
better the RGD performance for at least one of the FKM’s versus the HNBR. This however is not really a
true picture, as the performance in CO2 is still related to the overall ability of the material to withstand RGD
damage. There is no doubt that CO2 is far more likely to damage seals at any given temperature than
methane, as the CO2 rapidly diffuses into the elastomers under even moderate pressures and exhibits more
aggressive behaviour. Some materials swell severely as the gas is absorbed, though the level of swelling
will be restricted by the housing dimensions. This will also be the case during decompression where it has
been seen that the level of swelling whilst unconstrained is much higher. Even when the increase in volume
is restricted by the housing, the stored energy released during decompression can be a destructive force
which creates the damage to the seals. It is important to note that we did not experience any sign of
extrusion damage during our testing, and that any swell was effectively constrained within the housings.
There is also a possibility that the plasticising effect of the CO2 may improve the RGD resistance of some
materials by making them more ductile.
The question has often been raised of permeability of the materials versus their RGD performance. The high
speed of uptake and decay of CO2 shown by the quoted papers suggest that the differences between
polymers would have minor influence on RGD performance. If the CO2 uptake was slowed dramatically that
may also mean the diffusion out of the seal would also be reduced. That simply means that the forces that
create damage would be more sustained.
FKM is generally swollen more by CO2 than HNBR as solubility is higher. As a result of the increased
solubility the time taken for the CO2 to fully dissipate is longer. This has been ably demonstrated by one of
our customers by immersing CO2 swollen seals in water and observing the decay in bubble formation.
Figure 4, FKM moulding immediately after 5 days immersion in CO2

Figure 5, HNBR moulding immediately after 5 days immersion in CO2
All our failures were due to splits perpendicular to the force applied by gas pressure. We did not see any
evidence of blistering on tested seals. All these seals however were designed for RGD resistance. This may
be a characteristic of our materials and other materials may have different modes of failure. It has been said
that HNBR is more suitable for RGD resistance than FKM with high levels of CO28. Our evidence does not
support that theory. HNBR is however successfully used in applications involving high levels of CO2 within
its operating limits.
Conclusions
CO2 does create problems for elastomers. Materials may swell and be more prone to RGD failure. This
however is a somewhat simplistic view. In most cases damage caused by extreme swelling doesn’t apply
due to the use of housings. Although there is a plasticisation effect as a result of CO2 uptake, the severe
softening effect that would be associated with fluid uptake to this degree is not present. The integrity of the
seals are therefore less likely to be compromised. The usual consideration of greater than 10% swell in a
housing being problematical does not apply. Unlike fluid swell where it takes a long time for expansion to
decay it contracts more rapidly with gas induced swell. There does not appear to be any chemical effects
and any loss of properties is due to physical damage or plasticisation. It would appear that softening due to
gas plasticisation is minimal in comparison with that from fluids.
Selection of materials for CO2 service would seem to suggest using the same materials as for the other gas
mixtures but the temperature and seal size capability of these materials is far more limited. We can now
predict material performance, to some extent, to suit our customer’s requirements, though still need to
complete this work by increasing the number of compounds tested and to try other temperatures and ring
section sizes.

REFERENCES
1. Recommended Practice - Design and Operation of CO2 Pipelines - Det Norske Veritas April 2010
2. Danny Hertz III – Elastomers in CO2 -2012 Smithers High Performance Elastomers & Polymers for
Oil & Gas Applications conference
3. Elastomeric Materials based on Hydrogenated Nitrile Rubber for Seals in Carbon Dioxide (R 744)
High Pressure Service. Improving the Resistance against Explosive Decompression. Dr Hans Maag,
Achim Welle, Dr Matthias Soddemann, Dr Kevin Kulbaba. Merl Oilfield Engineering with Polymers
2010
4. Bjorn Melve – Effect of CO2 on Elastomers for Snøhvit CO2 Injection Well – Rubbercon 2012
5. Dr Sabine Munch, Glyn Morgan and Dr Barry Thomson – Observing Rapid Gas Decompression: A
Novel Technique. 2012 Smithers High Performance Elastomers & Polymers for Oil & Gas
Applications conference
6. Z Major, K Lederer, M Moitzi, T Schwarz and RW Lang – Development of a Test and Failure
Analysis for Elastomeric Seals Exposed to Explosive Decompression. Oilfield Engineering with
polymers 2006
7. AF George, S Sully and OM Davies – Carbon Dioxide Saturated Elastomers: The loss of Tensile
Properties and the Effects of Temperature Rise and pressure Cycling. Fluid sealing – BHR Group
1997
8. BP Technical Bulletin – Avoiding Gas Decompression Damage of Rubber Seals
9. Fundamental process and system design issues in CO2 vapor compression systems
Man-Hoe Kim, Jostein Pettersen, Clark W. Bullard - Progress in Energy and Combustion Science
30 (2004)
Acknowledgement
The authors gratefully acknowledge the additional help given by their colleagues John Rogers and John
Gray.


INDUSTRY STANDARDS - A BLESSING OR A CURSE?

Alexandra Torgersen
FMC Technologies
Email: alexandra.torgersen@fmcti
BIOGRAPHICAL NOTE
Alexandra Torgersen have a PhD in materials science from University of Oslo (1998)
and worked for several years at General Motors before moving into the oil-business.
The oil-related career has taken her from FMC Technologies to DNV, Statoil and RWE
Dea before she now have come full circle and rejoined FMC Technologies. All these
positions in different companies have circled around materials engineering and
polymer/elastomer materials. She also runs her own consultancy firm.
In addition, in many of the positions she has had over the years, she has been responsible for qualifying new
technical solutions for soft seals in subsea equipment. Through years of working within elastomer/polymer
materials, she has become member of both Norsok M710 working group as well as several ISO working
groups. During the writing and revision work on different industry standards, she has become increasing
familiar with both how such standards are made and how they were intended to be read
Background
Much of the subsea equipment used today is in some way linked to industry requirements through
standards. These standards, ISO, API, NORSOK or others are all geared towards ensuring that the
equipment is fit for purpose. These standards impose regulations either through requirements to equipment
and system, or through requirements on materials.
When new sealing systems are designed, or new equipment is designed, there is always a question of how
to qualify these new technologies, and industry standards are the basis for any such qualification. The main
aim for qualifications is to prove fitness for service and application. It is always important to make sure
standards are used correctly and actually assist in the final goal of ensuring fitness for service.
Fitness for service is a wide term that covers leakage of seals, service life, dynamic performance (if relevant)
and a multitude of potential failure mechanisms that needs to be addressed. Norsok M710 is coming out
relatively soon with a new revision, and ISO23936-2 has been out only a few years.
Standards
There are two main types of industry standards available for qualifying and verification testing sealing
solutions, namely the materials standards and the applications standards. A materials standard concerns
itself primarily or even pure with testing material properties in a specific environment, while the application
standards cover tests proving functionality of the sealing solution. Below are some main examples of
industry standards used widely for subsea sealing applications.
Norsok M710 and ISO 23936-2 are standards that covers materials properties of elastomers and polymers
in subsea applications. There are many types of applications that can be covered by Norsok M710 and also
ISO 23936-2, but common for all applications is that the methodology only covers materials properties, not
application specific properties. The main properties covered are chemical ageing and RGD. The
methodology is generic, and can be adopted to suit any type of fluid.
ISO 1817 is a pure materials property standard and describes in some detail methodology for assessing long
term properties of materials after exposure to a specific fluid. It is a generic standard that can be used for
any material and any fluid.
ISO 13628-6 covers subsea control systems, and within this standard there is an annex C that covers
qualification of hydraulic fluids for use in control systems. This annex can also be used to test new materials
Page 1 of 2 pages Paper 6 - Torgersen

for suitable material properties during exposure to hydraulic fluids. It is a materials property standard (when
considering sealing materials).
ISO10423 is written for christmas trees and vertical systems, and is an equipment standard. There are 2
annexes that cover testing of elastomers and polymers, namely Annexes F1.11 and F1.13. The first covers
temperature and pressure cycle testing of seals in grooves, and the last covers chemical ageing followed by
pressure testing to verify functionality of sealing system. These annexes should be used to verify sealing
design.
ISO 13533 covers testing and qualification of all types of blow-out preventers. It is an applications-based
standard, and only covers testing relevant for proving that the seal functions for blow-out preventer
purposes.
ISO 14310 covers down-hole packers and bridge-plugs. Again, this is a typical applications-based standard
and does not consider materials properties. Its only concern is the application and all test methodology is
written to cover the functionality needed to perform as bridge-plug or packer.
Discussion
As shown above, there are several standards covering materials properties and application specifics. The
key element when using standards is to fully understand the scope that each covers, and also how to best
utilize the information gathered from each of the relevant standards.
The main concept of standards is that they are written based on past experience and joint industry
understanding. Furthermore, such standards are always written as a minimum requirement agreed between
all participating persons/companies that help develop each standard. As such, they will only ever fully work
for new applications that are identical or similar to prior types of applications. These applications must also
be using types of elastomers that behave the same way as those previously used for the same applications.
This means that all new technology development will inherently be placed outside the main scope of the
standards, or at the very best on the side of current standards. Thus, using these standards are a mixed
blessing. They may not fully cover the application that is intended. They may also not have fully relevant
acceptance criteria based on current qualification needs.
The key to successful qualification of new sealing solutions is to be able to use the standards for what they
cover and introduce relevant acceptance criteria, and possibly additional tests to be able to cover the entire
scope of the new application within its qualification program.
The industry standards are certainly a blessing, they standardize how the industry approaches common
sealing qualifications. There is a common understanding of how seals should be tested, a minimum
requirement that all seals must meet. However, industry standards are also a curse, since many use them as
the whole truth about qualifying new sealing technology. The main focus must always be the fitness for
service. If the standards are not fully relevant or do not fully cover the application, then the application must
be the basis for testing. Not simply adhering to standards.
Paper 6 - Torgersen Page 2 of 2 pages

NEXT GENERATION CNT/RUBBER SOLUTIONS FOR OIL AND

GAS INDUSTRIES
Michael CLAES, Alicia RUL
Nanocyl SA
Rue de l’Essor, 4 - 5060 SAMBREVILLE - Belgium
michael.claes@nanocyl.com
BIOGRAPHICAL NOTE
Michael Claes is Global Technical Director at Nanocyl SA (Sambreville,

Belgium). He got his Master degree in chemistry in 2000 and studied polymer
chemistry in the frame of a Ph D. thesis both from University of Liege (ULg),
Belgium. He joined the staff of Nanocyl, one of the world leaders in Carbon
Nanotubes, as researcher in 2004 and is managing global R&D and Technical
Service since 2009. To promote creation and development of new market
applications for carbon nanotubes in the field of composite materials, he
supervises a team of 15+ trained people. He is the author or co-author of
several papers and patents.
ABSTRACT
The rubber industry is always demanding properties improvement for its applications. Carbon black and
silicas, most common used fillers mainly for rubber reinforcement, do not allow anymore reaching high level
specifications requested for novel applications. One example is the pressure and temperature that are more
and more high in fluid transfer systems, for which rubber compounds need to show very good properties
either at low temperature and frequency, either at high temperature and frequency.
The rubber formulations required high technology and strong technological development. In the aim to face
these specification issues, other fillers have been studied to reach these excepted specifications. Among all
existing organic and mineral particles, nanoparticles show the higher potential thanks to their small size, in
particular for non-spherical fillers. However, these nanoparticles are often agglomerated powder form that
requires a work on dispersion in the aim to reach optimized properties. Carbon nanotubes (CNT) are
nanosized particles with tubular shape that allow the creation of well-defined 3D network at low loading.
This study will show the high potential of use of CNT in rubbers in the goal to improve several common
properties such as mechanical, dynamical, thermal properties, electrical conductivity, abrasion, etc.
Furthermore, we will show that carbon nanotubes can be used in synergy with other fillers to reduce the total
loading of filler in the rubber formulation, and then improve mechanical properties of rubber by recovering
neat rubber behavior.
Rubber nanocomposites attract many researchers as well as industrially for decades due to their high
potential of unique properties. In practical applications, fillers as carbon black and silica are generally used to
improve mechanical and physical properties of a rubber matrix. Considerable interest on nanoparticules is
explained by the potentiality to increase mechanical and physical properties using a low filler loading
because of the small particle size leading to the increase of its surface area. The typical reinforcing fillers
include clays (layered silicate), carbon or silica fibers, expanded graphite, POSS (polyhedral oligomeric
silsesquioxane), and single-wall and multiwall carbon nanotubes (MWCNT) with a range of diameter and
lengths.
The extent of properties improvement using fillers mainly depends on interactions between the matrix and
the filler. It is crucial to improve the filler dispersion as well as the adhesion of the matrix polymer to the filler
surface, in order to increase the effective filler volume efficiency. These composites are realized through
different processes involving melt mixing. Several non-conventional techniques such as latex coagulation
and solution mixing are also developed to reach the best level of dispersion of nanofiller in the rubber matrix.
Indeed, a fine dispersion of filler is required to reach the strong impact of nanofiller on rubber properties. For
Page 1 of 6 pages Paper 7 - Claes

this reason, rubber nanocomposites are expected to exhibit higher modulus, hardness, abrasion resistance
and barrier properties in comparison with other micro-sized fillers.
Recently it has been shown that the use of carbon nanotubes (CNT) lead to a more efficient reinforcement1
due to the very high aspect ratio of CNT2. Indeed, spherical shape of carbon blacks and silicas require a high
quantity of filler to create a 3D percolation network, while lower loading of CNT is requested to reach the
same level of results.
Carbon nanotubes were identified by Endo in 19763, and described in 1991 by Iijima4 while studying the
surfaces of graphite electrodes used in an electric arc discharge. A Carbon Nanotube is a tube-shaped
material, made of carbon, having a diameter measuring on the nanometer scale. A nanometer is one-
billionth of a meter, or about one ten-thousandth of the thickness of a human hair. The graphite layer
appears somewhat like a rolled-up chicken wire with a continuous unbroken hexagonal mesh and carbon
molecules at the apexes of the hexagons. Interest in this area is immense and had increased exponentially
since 1994 when Ajayan et al.5 published the first introduction of PM/nanotubes composites. CNT are usually
made by carbon arc discharge, laser ablation of carbon or chemical vapor deposition (typically on catalytic
particles). Possible large scale production6 makes multi-wall carbon nanotubes extremely attractive because
their production is less complex, more cost effective and can produce high yields of CNT products (chemical
vapor deposition synthesis route).
On the opposite with carbon fibers, structure of carbon nanotubes is entirely known at the atomic level. The
helicity in the arrangement of the carbon atoms in hexagonal arrays on their honeycomb surface lattices
introduces significant charges in the electronic density of state, and hence provides a unique electronic
character.
TEM image of Pure NC7000 7 MWCNT schematic representation7
Due to their nature, carbon nanotubes can potentially improve mechanical, electrical, and thermal polymer
properties. In fact, the perfect arrangement of structural bonds oriented along the axis of nanotubes linked
with the carbon-carbon covalent bond strength, which is one of the strongest in the nature, would produce an
exceedingly strong material with huge physical properties. Practically, improvements of these properties are
observed using CNT in comparison of the use of carbon black, but these improvements are not as good as
these we can expect working at nanometer-scale. That could be explained by a poor dispersion and a lack of
interfacial adhesion between CNT and the polymer8. The nature of dispersion for CNT is very different from
other conventional fillers such as spherical particles or fibers because of the high aspect ratio of CNT (>100).
In order to increase mechanical and physical polymer properties by added reinforcing CNT, it is essential to
optimize the dispersion of filler agglomerates to reach the nanometer-scale. Indeed, the load transfer of the
properties is highly dependent on the extent of both the distribution and the dispersion of anisotropic nano-
fillers. The desire within the advanced composite research community is to seek a CNT/polymer composite
with physical properties that approach the theoretical maximum value of an individual nanotube. Although
CNT can be incorporated into polymers via solution blending to improve dispersion of CNTs in polymer
1
Gerspacher M., Sid Richardson Carbon Company, Fort Worth, Texas, ACS, Rubber Division, 2002
2
Coleman et al., Carbon, 44 (9) 2006, Pages 1624-1652
3
Oberlin, A. & M. Endo, J. Cryst. Growth 32, 335–349 (1976)
4
S. Ijima, Nature 354, 56 (1991), 391
5
Ajayan et al., 1994
6
Perez et al., Polymer Engineering and Science (2009) 866-874
7
Nanocyl S.A.
8
Bokobza L., Polymer 48 (2007) 4907-4920
Paper 7 - Claes Page 2 of 6 pages
matrices, melt blending is more suitable for industrial approach due to its higher environmental and
economic cost9. Melt blending is the most efficient and convenient process for preparation of CNT reinforced
polymers, but there are only few studies reporting impact of these fillers on mechanical and physical
properties of these composites.
Practically, mechanical properties of a rubber compound could be reached with low loading of CNT in
comparison of carbon black. The charts below show that a lower quantity of Carbon nanotubes is requested
to reach high modulus value in comparison with conventional carbon black in a NBR rubber.
Stress / Strain Evolution of NBR rubbers filled with CNT (NC7000™) – left chart and CB (N550) – right chart.
A reinforcing factor could be calculated comparing different fillers such as carbon blacks and Carbon
Nanotubes, showing that CNT reinforcing factor is higher for CNT and in particularly NC7000™, the
commercial product of Nanocyl, in comparison with standard Carbon black (N550). These results are
summarized on the following chart.
Reinforcing factor of NC7000™, the commercial CNT product of Nanocyl,

in comparison with a conventional Carbon black N550.
Once, it has been proved that same mechanical properties can be reached with low loading of CNT vs.
carbon black, other properties could be evaluated. Electrical resistivity is the most evident property that is
provided by CNT to rubber. In fact, few percent of NC7000™ allows reaching a dissipative level of electrical
conductivity while 10 to 20 % of conductive carbon black where needed. The charts below show the
electrical conductivity of NC7000™ vs. N550 carbon black in NBR.
9
Perez, Polymer Engineering and Science (2009) 866-874
Electrical conductivity of NC7000™, the commercial CNT product of Nanocyl,

The important fact with Carbon Nanotubes is that, not only they allow to improved mechanical and electrical
properties, but also other properties like abrasion and gas permeation thanks to the peculiar network created
with this tubular filler. The following chart shows that CNTs reduces absorption rate of solvent that is
measured for CB/rubber compound.
Degree of swelling of NC7000™, the commercial CNT product of Nanocyl,

The 3D network created with Carbon Nanotubes could also be improved using Carbon nanotube and Carbon
black in combination in the rubber. First of all, Carbon nanotube and Carbon black show very good affinity
that is shown in the following TEM picture.
TEM picture of NC7000™ in combination with conventional Carbon black in a NBR rubber.

The aim of this talk is to show the good synergistic effect between carbon nanotubes and carbon black. The
innovative idea is to replace an important part of carbon black used in a rubber formulation with a small
amount of Carbon Nanotubes. In that case, mechanical properties of materials would be improved thanks to
CNT’s higher surface area than CB, and synergistic effect of both fillers would have an impressive effect on
mechanical, dynamical and thermal properties. Furthermore, combining use of CB and CNT would also
provide electrical conductivity to the new material.
TEM picture of NC7000™ in combination with conventional Carbon black in a NBR rubber.
These “hybrid systems” enhance the material performance including for oil and gas applications when high
performances are requested. One example is the stator of power sections that need good mechanical and
dynamical properties. Combination of CNT/CB in NBR formulation lead to a better elongation modulus that
reduce abrasion resistance and improved wear resistance, and an increase of power density of 30 %. This
homogeneous dispersion of CNT and CB in rubber matrix allow an increase of operating time from 100
hours for the reference to 130 hours. Use of NC7000™ into the downhole industry helps to drill faster,
stronger, deeper and longer.
Electrical conductivity, mechanical and dynamical properties, abrasion and friction resistance, heat
dissipation, and barrier properties are all properties that would enhance performances such as dissipative
systems, anti-vibration systems and moving parts, work in temperature or pressure parts. Structural parts,
pipes, sealing systems, motor suspension, cable and wires, rolls, wheels, tyres, conveyor belts could contain
Carbon nanotubes to be used in several industries like automotive, aeronautic, railway, tyres, downhole, fluid
transfers …


HIGHLY EFFECTIVE ANTI-CORROSION EPOXY SPRAY

COATINGS CONTAINING SELF-ASSEMBLED SMECTIC CLAY
Peng Li and Hung-Jue Sue
Polymer Technology Center, Texas A&M University,
College Station, TX 77843, USA
Tel: +1 979 845-5024 Fax: +1 979 845-3081 email: hjsue@tamu.edu
BIOGRAPHICAL NOTE
Prof. Hung-Jue Sue received his Ph.D in the Macromolecular Science and Engineering
Program in 1988 at the University of Michigan, Ann Arbor. After working at Dow
Chemical for 7 years, he joined Texas A&M University in 1995. He currently serves as
the Director of the Polymer Technology Center at the Texas Engineering Experimental
Station. Dr. Sue has published over 200 peer-reviewed journal articles and book
chapters and has presented at over 200 international conferences. He has trained over
40 Ph.D. students and postdocs. His research interests include nanomaterial synthesis,
dispersion, and assembly for functional and structural applications, mechanical
properties of polymeric thin films, and scratch behavior of polymers and coatings.
ABSTRACT
Epoxy nanocomposite coatings containing self-assembled 2D colloidal α-zirconium phosphate nanoplatelets

(ZrP) in smectic order have been prepared via a simple, energy-efficient fabrication process that is favorable
to industrial practices. These smectic epoxy/ZrP coatings are highly effective against metal corrosion. The
usefulness of the above epoxy/ZrP coatings for a vast variety of engineering applications will be presented
and discussed.
Introduction
1
Metal corrosion is estimated to cost the U.S. $300 billion dollars annually. A wide variety of anti-corrosion
coating technologies have been developed to prevent or delay metal corrosion. However, the technologies
that are known to be effective tend to cause undesirable side effects. For example, chromate-based
coatings, which exhibit excellent corrosion resistance, are banned from usage in many applications because
2
of their toxicity and carcinogenic nature. Zinc-based coatings are undesirable due to their lack of ductility,
3 4, 5 6, 7 8, 9
high cost, and shortage of raw materials. Zeolites, ceramics, and graphene have also been explored
as corrosion-resistant coating materials, but show only limited success. Recently, a new generation of
organic coatings has attracted significant attention due to their facile and eco-friendly nature in fabrication
10-13
and functionalization. Strategies employed to prepare these new organic-based coatings include, but are
14-18
not limited to, hydrophobicity-induced reduction in water accessibility, passive oxidation-enabled metal
19-21 22-24
protection, and nanofiller-integrated corrosion inhibition. Unfortunately, these methodologies demand
complex chemical processes, making them difficult for large-scale commercial implementation.
Consequently, new anti-corrosion organic coatings that utilize existing industrial practices, such as spray
coating, are rigorously sought after.
Plate-like nanostructures, such as graphene and its derivatives and clay, are impermeable to gases and
25, 26
moisture. Therefore, nanocomposite coatings containing 2D plate-like nanostructures have been
reported to improve the corrosion resistance of metals. Among these protective coatings, it has been found
that their barrier properties strongly depend on the nanoplatelet aspect ratio, volume fraction, dispersion
27, 28
level, and particularly the degree of filler alignment. Highly aligned platelet-based lamellar structures are
29
readily observed in natural materials, such as nacre. Several assembly techniques have been developed to
30
fabricate lamellar-like polymer/clay nanocomposites, e.g., Layer-by-Layer (LbL) assembly, ice templating
31 32 33
and sintering of ceramics, vacuum-assisted self-assembly, electrophoretic deposition, and air/water
34, 35
interface assembly. Again, most of the above approaches are based on time-consuming sequential
depositions or require extensive energy consumption, which severely limit their large-scale practical
Page 1 of 4 pages Paper 8 - Sue
applications.
In the present study, a facile and scalable spray-coating approach has been developed to prepare anti-
corrosion epoxy coatings that contain self-assembled zirconium phosphate (ZrP) nanoplatelets in smectic
order. These ZrP-containing epoxy coatings exhibit long-range order with platelet orientation parallel to the
metal surface, which acts as highly effective barrier layers to prevent electroactive species, such as water
and oxygen, from reaching the metal surface. Electrochemical analyses reveal an improvement against
corrosion by as much as an order of magnitude when compared with the neat epoxy coating counterpart.
Results and Discussion
From an industrial manufacturing perspective, the application of corrosion-resistant coatings must be fully
scalable and allow for a high-throughput application without complicated procedures or expensive setups.
Our simple, yet effective method is capable of achieving the formation of 2D lamellar-like nanostructures on
a metal substrate as shown in Figure 1. The spray-coating technique allows for the fast and efficient
assembly of functionalized smectic structures under ambient conditions. Layered ZrP, Zr(HPO4)2·H2O, was
36
synthesized using a hydrothermal method and exfoliated by a proton exchange reaction in acetone. ZrP
nanoplatelets have strong covalent bonding along the primary plane but interact with neighboring platelets
37
through out-of-plane van der Waals (vdW) forces and hydrogen bonding. The P-OH functional groups on
38
ZrP surfaces cause the nanoplatelets to be functionalized with proton donors (i.e., amines). As a result, the
individually exfoliated ZrP with a high aspect ratio of 160 nm was prepared in organic solvent.
Figure 1. Scheme of the preparation process of smectic epoxy/ZrP coating.
The mesoscale structure of the smectic epoxy/ZrP coating on a metal substrate was investigated using TEM
and GISAXS. TEM images of the epoxy/ZrP (11 wt.%) coating show that the ZrP nanoplatelets are self-
assembled into a well-aligned mesoscopic structure (Figure 2(a)). Lamellar structures that are aligned
parallel to the substrate display Bragg peaks along the vertical axis (qz-axis). Peaks present along the
horizontal plane (qx-axis) indicate a perpendicular alignment to the substrate. The Bragg peaks obtained
from GISAXS 2D pattern (Figure 2(b)) appear exclusively in the qz-axis, indicating that the smectic ZrP
nanoplatelets are aligned parallel to the metal substrate. GISAXS 1D spectrum of smectic epoxy/ZrP (11
wt.%) films displays the characteristic peaks for the lamellar phase with a d-spacing of 6.1 nm (Figure 2(c)).
The creation of such smectic structures on a metal surface is likely to be the main cause of the greatly
enhanced anti-corrosion performance.
To investigate the corrosion resistance of the coatings, potentiodynamic measurements were conducted
(Figure 3). Two important parameters, corrosion potential (Ecorr) and corrosion current density (Icorr), were
39
measured to determine corrosion resistance. Ecorr is the measure of corrosion susceptibility, and a positive
shift in Ecorr indicates increased corrosion resistance. Icorr represents the intensity of the cathodic oxygen
40
reduction and anodic dissolution of metal ions. It was observed that the aluminum alloy (Al) substrate had
an Ecorr of -0.825 V. Coating the Al substrate with neat epoxy and smectic epoxy/ZrP increased the Ecorr to
Paper 8 - Sue Page 2 of 4 pages
2
0.102 V and 1.068 V, respectively. At the same time, Icorr decreased from 29.41 µA/cm for the bare Al, to
2
0.102 V and 1.068 V, respectively. At the same time, Icorr decreased from 29.41 µA/cm for the bare Al, to
2 2
1.98 µA/cm for the neat epoxy-coated Al, and finally to 0.36 µA/cm for the smectic epoxy/ZrP-coated Al.
Table 1 summarizes Ecorr, Icorr, and corrosion rate (CR) for various sample systems. Compared to the neat
epoxy-coated Al, the Ecorr of the smectic epoxy/ZrP-coated Al increased by 950%, and the Icorr was reduced
by 80%. This demonstrates that smectic epoxy/ZrP coating is promising as an anti-corrosion coating.
Figure 2. (a) TEM of a cross-section of smectic epoxy/ZrP coating (11 wt.%). (b) 2D and (c) 1D
diffractograms of GISAXS, suggesting ZrP nanoplatelets are aligned parallel to the Al substrate.
Figure 3. Potentiodynamic polarization of bare Al alloy, neat epoxy, and smectic epoxy/ZrP-coated Al.
Table 1. Electrochemical corrosion properties of bare Al, neat epoxy, and smectic epoxy/ZrP-coated Al.
Conclusion
We have demonstrated the excellent metal anti-corrosion performance of a sprayable epoxy coating
containing ZrP in long-range smectic order for the first time. These coatings effectively prevent the
permeation of oxygen to disrupt the corrosion process. The potentiodynamic polarization experiments
quantitatively indicate that these smectic epoxy/ZrP coatings can remarkably improve the corrosion
resistance of Al substrate. Such high-performance epoxy nanocomposites are suitable for large-scale anti-
Page 3 of 4 pages Paper 8 - Sue
corrosion and barrier film applications.
References
1 G. H. Koch, M. P. Brongers, N. G. Thompson, Y. P. Virmani and J. H. Payer, U. S. Department of

Transportation, Federal Highway Administration, Washington D.C., 2001.
2 M. Rohwerder and A. Michalik, Electrochim. Acta, 2007, 53. 1300-1313.
3 A. R. Marder, Prog. Mater. Sci., 2000, 45, 191–271.
4 A. Mitra, Z. Wang, T. Cao, H. Wang, L. Huang and Y. Yan, J. Electrochem. Soc., 2002, 149, 472-478.
5 X. Cheng, Z. Wang and Y. Yan, Solid-State Lett. , 2001, 4, 23-26.
6 F. Zhang, L. Zhao, H. Chen, S. Xu, D. G. Evans and X. Duan, Angew. Chem. Int. Ed., 2008, 47, 2466-2469.
7 H. Jensen, G. Surf. Coat. Technol. 1996, 84, 500−505.
8 F. Zhou, Z. Li, G. Shenoy, L. Li and H. Liu, ACS Nano, 2013, 7, 6939-6947.
9 S. Chen, L. Brown, M. Levendorf, W. Cai, X. Li, C. Magnuson, R. Piner, J. Kang, J. Park and R. S. Ruoff,
ACS Nano, 2011, 5, 1321-1327.
10 X. He, C. Chiu, M. Esmacherb and H. Liang, Surface and Coatings Technology, 2013, 237, 320-327.
11 K. L. Choy, Prog. Mater. Sci., 2003, 48, 57-170.
12 T. K. Rout, G. Jha, A. Singh, N. Bandyopadhyay and O. N. Mohanty, Surf. Coat. Technol., 2003, 167, 16-
24.
13 M. L. Zheludkevich, I. Salvado and M. Ferreira, J. Mater. Chem., 2005, 15, 5099-5111.
14 M. Liu, S. Wang, Z. Wei, Z. Song and L. Jiang, Adv. Mater., 2009, 21, 665–669.
15 X. Huang, J. D. Chrisman and N. S. Zacharia, ACS Macro Lett., 2013, 2, 826-829.
16 A. Leon, R. Pernites and R. Advincula, ACS Appl. Mater. Interfaces, 2012, 3169-3176.
17 Z. Chen, Z. Dong, D. Yang and H. Lu, Adv. Mater., 2013, 25, 5352–5359.
18 R. Naderia, M. Fedelb, F. Deflorianb, M. Poelmanc and M. Olivierc, Surface & Coatings Technology, 2013,
224, 93-100.
19 F. Chen and P. Liu, ACS Appl. Mater. Interfaces, 2011, 3, 2694-2702.
20 T. Yeh, T. Huang and J. Yeh, Polym. Chem., 2012, 3, 2209-2216.
21 A. Yağan, N. Özçiçek and P. Yıldız, Electrochimica Acta, 2008, 30, 5242-5251.
22 T. H. Hanga, T. A. Truca, V. K. Oanha, J. B. Jorcinb and N. Pébère, Surface & Coatings Technology, 2007,
201, 7408-7415.
23 M. Behzadnasaba, S. M. Mirabedinia, K. Kabiria and S. Jamalib, Corrosion Science 2011, 89-98.
24 M. Behzadnasaba, S. M. Mirabedinia and M. Esfandehb, Corrosion Science, 2013, 75, 134-141.
25 L. Sun, W. Boo, A. Clearfield and H. J. Sue, Journal of Membrane Science, 318, 129-136
26 R. Nair, H. Wu, P. Jayaram, I. Grigorieva and A. Geim, Science, 2012, 335, 442-444
27 N. Lape, E. Nuxoll and E. L. Cussler, Journal of Membrane Science, 2004, 236, 29-37
28 C. Lu and Y. Mai, Composites science and technology, 2007, 67, 2895-2902.
29 P. Podsiadlo, A. Kaushik, E. Arruda, B. Shim, J. Xu, H. Nandivada, A. Ramamoorthy and N. A. Kotov,
Science, 2007, 318, 80-83.
30 P. Podsiadlo, M. Michel, K. Critchley, S. Srivastava, M. Qin, J. W. Lee, E. Verploegen, A. J. Hart, Y. Qi and
N. A. Kotov, Angew.Chem., Int. Ed. , 2009, 38, 7073-7077.
31 E. L. Munch, M. E. Alsem, D. H. Saiz, E. Tomsia, and A. P. Ritchie, Science, 2008, 322, 1516–1520.
32 K. Putz, O. C. Compton, M. Palmeri, S. Nguyen and C. Brinson, Adv. Funct. Mater., 2010, 20, 3322-3329.
33 P. Sarkar and P. S. Nicholson, J. AM. Cerem. Soc., 1996, 79, 1987-2002.
34 B. R. Heywood and S. Mann, Adv. Mater., 1994, 6, 9-20
35 L. J. Bonderer, A. R. Studart and L. J. Gauckler, Science, 2008, 319, 1069-1073.
36 W. Boo, L. Sun, J. Liu and H. J. Sue, Journal of Polymer Science Part B, 2007, 45, 1459-1469.
37 L. Sun, J. Y. O'reilly, D. Kong, J. Su, W. Boo, H. J. Sue and A. Clearfield, J Colloid Interface Sci., 2009, 15,
503-509
38 L. Sun, Boo, W. J., Browning, H.J. Sue and A. Clearfield, Chem. Mater., 2005, 17, 5606-5609.
39 R. K. Singh Raman and P. M. Ajayan, Carbon, 2012, 50, 4040-4045
40 G. Kear, B. D. Barker and F. C. Walsh, Corrosion Science, 2004, 46, 109-135.
41 E. L. Cussler, S. E. Hughes, W. J. Ward and R. Aris, J. Membr. Sci., 1988, 38, 161–174.
Paper 8 - Sue Page 4 of 4 pages

POLYMER MATERIALS AND NANOTECHNOLOGY

FOR OIL AND GAS
Rigoberto C. Advincula
Case Western Reserve University, Department of Macromolecular Science and Engineering
Cleveland, OH, 44106 USA
Tel: 216-368-4566 email: rca41@case.edu
BIOGRAPHICAL NOTE
Dr. Rigoberto C. Advincula is Professor at Case Western Reserve University,

Department of Macromolecular Science and Engineering. He is a Fellow of the
American Chemical Society (ACS) and has recently received the Mark Scholar
award of the Polymer Chemistry Division, ACS. He has published over 400
papers and is highly cited with an H-index = 41. He is Editor of the Journals
Reactive and Functional Polymers and Polymer Reviews. He has been plenary
and invited speaker to a number of international conferences and has held
visiting professor positions at MPI-P, NUS, TUAT, and AIT. He has mentored 26
Ph.D. students and continues to mentor undergraduate and high school students
as well. He obtained his Ph.D. at the University of Florida and did Post-doctoral
work at the Max Planck Institute for Polymer Research and Stanford University.
He consults and collaborates with a number of companies.
ABSTRACT
This paper summarizes the opportunity to investigate potential applications of new polymers and
nanomaterials in the oil and gas energy industry. For the last 10 years there has been an increase in interest
and research for new materials useful for upstream, midstream, and downstream processes to effectively
find function in demanding environments including directional drilling and hydraulic fracturing. High
temperature high pressure (HT/HP) and brine conditions pose a challenge for emulsification, demulsification,
and viscosity of drilling fluids. This talk will give a short overview of the polymer materials requirements in the
oil and gas industry, the opportunities and function for new structure and compositions, and the use of
graphene, graphene polymer coatings - nancomposites in high performance materials, drilling, well-logging,
anti-corrosion, anti-scaling, and other additive materials.
Polymers in Oil and Gas
Polymers are large molecules or macromolecules that derive their property based on size, composition, and
ability to interact with neighbouring chains or solvents (non-covalent interactions and cross-linking). Plastics
is a most familiar term.1 Although, these days many everyday things from clothing to car parts to packaging
are synonymous with material comfort. As a barrier or coating material, its properties are important in
preventing chemical and physical degradation of the bulk phase that needs protection. On the other hand, as
a packaging material, it is important to protect the outside environment from the contents of the bulk phase.
A high number of these materials are classified as engineering materials and has to be fabricated through a
number of methods from their corresponding resins. A number of abbreviations are as follows: polystyrene
(PS), polymethylmethacrylate (PMMA), polyetheretherketone (PEEK), polyvinylchloride (PVC),
polycarbonate (PC), etc. (Figure 1)2,3
Polymers (thermoplastics, thermosets, and elastomers) play an important role in many phases and stages of
the oil and gas energy production. They can be divided into upstream, midstream, and downstream
applications. In the upstream applications they include: drilling mud viscosity modifiers, dispersants, anti-
corrosion, and anti-scaling agents, polymeric cement, elastomeric seals, thermoset coatings, thermoset parts
- replacement of corrosion prone parts. The requirements are more demanding when going to high
temperature/high pressure (HT/HP) conditions where degradation and creep is faster. This requires high
performance polymers.4 For offshore technology and sub-sea condition demands, there is also the
requirement for marine environment durability. That means preventing plasticization and redox degradation
mechanisms due to high salt conditions. Often, protective coatings have to approach almost hermitic sealing
conditions to enable long-term durability to prevent creation of salt-bridges. This is also observed with
requirements in the upstream applications for geothermal energy harvesting where HT/HP demands can be
Page 1 of 6 pages Paper 9 Advincula
combined with high brine conditions and silication scaling. In drilling applications, they find uses that enable
stability or viscosity control at various stages of drilling and completion. Polymeric cement is another
interesting applications since setting (curing time) and viscosity control is important at various stages of
casing development especially with directional drilling. In midstream applications, the use of pipes, coatings,
and other mechanical applications in housing, parts replacement, is present. However, polymers play an
important role in demulsification and as additives in controlling the transport of complex gas/liquid
compositions of produced oil/gas. Often, water emulsion is a problem and lack of viscosity control can have
enormous consequences in production rate and cost. Otherwise, the same demands for external
environment applications apply. In the downstream role, polymers even play a more important role as
additives with applications such as: anti-corrosion, anti-scaling, dispersants, emulsifiers, flocculants, viscosity
modifiers, asphaltene control, etc. As additives they have solubility ranges from hydrophilic, lipophilic, and
amphiphilic or surfactant properties. Hence solubility parameters and hydrophobic-lipophilic balance (HLB)
behavior is important. The use of polysaccharides, polyelectrolytes and hydrogels are numerous. It should
be noted that small molecule additives play an important role as well in many applications from upstream to
downstream especially as additives fulfilling the same role as polymers but with a different phase and
chemical behaviour.
Figure 1. Hierarchy of polymer materials and their stability (Tg and Tm) and possible degradation
correlation. (images obtained from Ref. 2 and 3)
One important role worth mentioning in detail is corrosion mitigation. A large amount or resources is often
devoted to mitigating corrosion. Thermodynamically, corrosion is a very energetically favorable process
converting a high-energy metal or metal alloy into its low-energy oxide form. Corrosion not only poses
serious problems economically and industrially, but also endangers human life, i.e. structural failure and
biofilm formation. It is very difficult to stop corrosion from happening completely. The only option is to slow
down its rate by preventing aggressive oxidizing species from reaching the metal surface, or by having a
sacrificial material that preferentially reacts in place of the metal. Once corrosion is observed, it is important
to investigate the different pathways and sources of failure to prevent further degradation. Thus it is
important to utilize effective mitigation strategies with new materials and coatings. Essentially, the use of
nanomaterials can be in the form of inhibitors, inhibitor reservoirs, cross-linking agents, and nanofillers for
improved coating durability.
Thus the role of polymeric materials as durable high performance materials and protective coatings is
obvious and from these, one can add the use of nanomaterials as additives or as blended agents in
enhancing the role of polymers in oil and gas energy.
Nanotechnology in Oil and Gas
Nanotechnology involves the applications of nanomaterials, nanostructuring (nanoengineering), and

nanoscale phenomenon towards practical solutions. In the oil and gas energy services, there are a number
of opportunities primarily related to interfacial phenomena and colloidal science. Another possibility is in the
spectroscopic or catalytic signature of these materials.
Paper 9 Advincula Page 2 of 6 pages

Here are several applications and research areas for oil and gas:
1) Drilling: Application in drilling and completions will involve materials and methods to drill and
complete our wells with increasing strength, durability and provide completion design options not
possible with existing technologies. Nanotechnology for example has been used to for formulating
novel guar fracturing fluids. For example, the use of stabilizing nanoparticle dispersions in very
demanding high salinity, high temperature downhole environments – since nanoparticles can act as
very stable surfactants and rheology modifiers. Current drilling and completion methods are often
faced by challenges and demands of increasingly hostile environments of new oil and gas finds both
onshore and offshore. Applications of advanced drilling and completion methods will require
materials performance beyond current technologies. Nanotechnology may provide unique solutions
to these challenges. Applications of nanotechnology to cement, drill bit, and drilling fluid design is
possible.
2) Reservoir engineering and production – From upstream to midstream, the viscosity of oil is very
important. Achieving low viscosity is essential for higher production in heavy oil wells. Increasing the
temperature of heavy oil is a usual practice in lowering its viscosity and consequently, increasing
productivity – but at a cost. Investigating methods of modifying the reservoir characteristics in
particular with hydration control, helps release more oil from the formation. Can nanoparticles be
used to induced more localized heating or delivery of higher heat capacity particles further in the
formation? There is also interest in controlling the demulsification process to control oil/water
mixture. Another interest is in the use of reservoirs for sequestering CO2. Is it possible to use
nanomaterials and nanoscale phenomena to increase the efficiency of CO2 injection and enable
predictions on the kinetics of mixing.
3) Flow management issues – This can be based on controlling scaling, corrosion, and paraffin
formation. Can nanotechnology be used to mitigate this problems and therefore assure flow
conditions well into the production phase? It is desired to have increased recovery, efficient reservoir
sweep water management and use of chemically modified nanoparticles to achieve reservoir
illumination for monitoring. For scaling control, this could involve the use of nanoparticle stabilized
emulsion that are also sources of scaling and corrosion inhibitors. Stable amphiphilic Janus particles
or anisotropic clay or graphene particles can be particular intriguing. The use of nanoparticle
materials in enhanced oil well cement hydration and improved mechanical properties is also
important.
4) Inhibitors and Monitoring agents – nanoparticles due to their nanoscale sizes are able to
penetrate deeper and farther into the formation and therefore, mitigate problems at early stages or
provide a better resolution of the reservoir formation. Nanoparticle inhibitors can be utilized to
prevent corrosion, scaling - they can be effective nanoparticle sealants. Fluorescent nanoparticles
can be used for real time reservoir monitoring. Depending on the nature of the particles, they can
also aggregate at the surface and prevent fluid loss. Examples include applications in enhanced oil
recovery (EOR) where the optimization of the well require both preservation of the formation (for high
yield) and mapping of the interconnectedness of wells and formation with sensors. They can also be
particularly useful in HTHP environments.
5) Scaling control in downhole conditions is important as delays or stops in production can be very
expensive. The effect is not only on the downhole structure but also on the equipment. A main
motivation is of course maintenance cost reduction. Can nanotechnology be used for unconventional
scale control? Several examples of applications can include, the use of nanoemulsion anti-scaling
agents - scale inhibitors based on reservoir and formation character and proper use of scale
indexing and nanomaterial coating on the formation to prevent adhesion.
6) Devices and Tools. Nanoscale phenomena can also be used to produce better alloys, coatings,
and composites. For example, it can be used for high strength in-situ self-corrodible metal
composites and alloys for completions tools. A main driver is retention of function even at HTHP and
high brine conditions. New materials and sensors at this highly corrosive and demanding
environments are needed. Knowledge gain on nanoscale phenomena in crystallization, annealing,
dispersion of materials can be used for fabrication. Nanofabrication has been shown to create high
resolution chemical sensors and robust sensors.

7) Sensor and tracers - Nanoparticles and other nanomaterials can be used to enhance various
aspects of oil and gas exploration and production. Metals, metal oxides, chalcogenides, carbon
based nanomaterials – their unique spectroscopic and catalytic properties, for example, can be
viewed as the next generation of efficient tracers or agents. This can be done by introducing these
materials as substitutes and alternatives for radio-isotope tracers. They can also modify surfaces,
interfaces and fluid properties, such as wettability and rheology. Because of their size and shape, the
deep penetration of nanoparticles in oilfield formations can be a determining factor in designing
successful nanotechnology-based treatments in the oil and gas industry.
Graphene Materials in Oil and Gas
Carbon polymorphs and nanomaterials include carbon nanotubes (single walled or multiwalled), fullerenes,
and graphenes. Graphene is the simplest of all graphitic forms - it can be stacked to form 3D graphite, rolled
to form 1D carbon nanotubes and wrapped to form 0D fullerenes. It is essentially a single-atom thick,
honeycomb network of two-dimensional sheets of sp2-hybridized carbon. Discovered in 2004, graphene with
its long-range -conjugation, has attracted tremendous attention for its exceptional structural, chemical,
mechanical, thermal and electrical properties. These remarkable attributes have stimulated escalating
interest for applications as field-effect transistors (FET), photovoltaics, biosensors and electrodes. Due to its
outstanding transparency (97.7%), graphene is envisioned as the future of transparent, touch-screen and
foldable electronic displays.5
Figure 1. Graphene structure of 2-D fused aromatic systems and actual TEM image of isolated
nanomaterial.
Some of the possible uses of introducing graphene into any material system includes: increased thermal
conductance, reduced friction, reduced wear & tear, stable viscosity, higher load bearing capability and
sensing or well-logging. These improvements would drive the applications of polymer nanocomposites and
nanomaterials in the oil & gas industry to even higher performance. Here are a few examples of the
applications of graphene:
1) Sensing. Graphene can possibly be used for well logging. Well logging protocols provide data and
evaluation on the geological properties and map of reservoirs of interest. This is employed in the production
phase (upstream). A commonly used logging technique to provides information downhole by the use of
wirelines. Wirelines are long wires with sensors attached at the end. These are lowered into an exploration
hole to provide information about the hole, its contents, depth profile, and connectivity (or conductivity). An
extension of wireline logging protocol is logging-while-drilling (LWD), which relies on sensors at the end of
the drill itself. However, both methods utilize oil-based fluids for drilling and lubrication. The difficulty is that
oil-based fluids are not very good conductors of electricity. This is ideal for the use of conducting graphene
materials. The Tour group developed magnetic graphene nanoribbons (MGNRs). The MGNRs can form part
of a conductive coating in oil-based drilling fluids, improving the reliability of the information relayed back up
the hole by the sensors.6 Furthermore, the magnetic properties of the ribbons could also be exploited as
sensors. Because of the small size and the aspect ratio, MGNRs can be made small enough to pass into
smaller fractures and crevices of the rock which can still hold extractable oil. As sensors, they can perhaps
send wireless data which contains information on oil location and concentration. Tiny sensors coated with
graphene could expedite the discovery of oil and natural gas reserves, As sensors in fluids injected down
exploration wells, they can then move more efficiently through cracks and crevices in search of
hydrocarbons. The interaction with graphene in particularly can be recorded as history or transmitted
wirelessly to a receiver. Most oil wells are drilled vertically - drillers know what's happening at well A and well
B, but do not know what is happening in the spaces between them. Such a technology will enable
geomapping and explore deposits laterally. It can have interesting using with directional drilling and hydraulic
fracturing.
2) Drilling Mud. The Tour group at Rice University first showed that adding graphene oxide (GO) platelets to
the drilling operation has several advantages. Adding GO to a water-based drilling fluid blend decreased the
losses of the fluid to the surrounding rock, as compared to a standard mixture of bentonite clays and
polymers commonly used in the drilling industry. When the drill bit is removed and drilling fluid displaced, the
formation oil forces remnants of the filter cake out of the pores - as the well begins its production phase.
However, some clays remain, reducing the portent productivity. Pliable and “starfish like” flakes of GO can
form a thinner, lighter filter cake - the flakes fold in upon themselves and look something like starfish sucked
into a hole. When the well pressure is relieved, because of the shape, the flakes can still be pushed back out
by the oil with the latent pressure. The thinner graphene layers can then be extracted more easily than the
layers from traditional clay-enhanced liquids. Thus with extraction, the GOs are rendered many times smaller
than the flakes’ original diameter by folding.
3) Hydraulic Fracturing. In hydraulic fracturing, proppants or propping agents are used to keep or “prop up”
fractures in the formation after a drilling hydraulic fracturing phase. Often, this involves several stages.
Although the bulk of this composition is water (99.5%), drillers need to add other chemicals and additives to
the high volume of water. It will be of high interest to use produced water (water already in the ground) or
seawater into the wells. However, high salt content (existing ions) and other hydrocarbons and aromatics
can result in complex viscosity and fluid conditions that are hard to control once the drilling and completion
phases are put into stage. As proppants, they can be used to fill gaps or go into very narrow fractures.
Graphene and graphene oxide can be used as sensors and additives to relay data up to the surfaces – and
army of sensors that can relay the data based on time (stages), distance, and depth profiling. As sensors
they could detect gases or changes in the water/ oil chemistry as well as the interface of emulsions. As
proppants and additives in the traditional sense, they can be used as surfactants and amphiphilic agents
capable of controlling the pH, viscosity, and emulsification process. By derivatizing GO with complementary
hydrophilic groups, ligands, and ionic groups, they can augment the performance of existing additives and
have the sensor reporting function. Like any additives, they have to be evaluated by their cost-effective ratio.
4) Coatings. Graphene oxide based nanocomposites and other polymer-filler materials can have superior
coating performance based on controlled wetting and conductivity. The preparation of graphene-polymer
electrodeposited films have been shown to have controlled wetting properties, conductivity, patternability,
and even anti-microbial properties has been demonstrated by our group. The latter is important for
preventing biofilm formation and preventing MIC corrosion. Patternabiliy and template deposition has been
demonstrated with graphene oxide – polyvinylcarbazole composites.8,9,10
CONCLUSION
Numerous opportunities are possible with the use of new polymer systems and nanomaterials like graphene
in the oil and gas production. This has been enumerated for upstream, midstream, and downstream
applications. The classification of polymer materials into thermosets, elastomers, and thermoplastics can
easily categorize their use as coating and engineering materials. However as additives, their solubility and
viscosity is of high importance. Nanomaterials include metals, inorganic oxides, semiconductors, organics,
and carbon based materials will find increasing use in the oil and gas industry as their salient properties are
reported in specific applications. Graphene in particular has some interesting applications based on its size,
shape- aspect ratio, and conductivity. In the future, such convergence of the use of polymers and
nanomaterials in the industry will be a matter of cost-effective ratio studies, where even a small amount of
the latter can make a difference in high performance and efficiency of operation.

REFERENCES
1. Stevens, M. “Polymer Chemistry: An Introduction” Oxford University Press, USA; 3 edition, 1998,
576 pages.
2. http://www.imt.kit.edu/english/243.php
3. http://www.chinaarray.com/resincompanies.html
4. Fink, J. “High Performance Polymers” :CHIP Publishers, Weimar, TX, 2008, 1- 609 pages
5. Warner, J. et. Al. “Graphene: Fundamentals and Emergent Applications”, Elsevier, 2013, 450 pages.
6. Genorio, B., Peng, Z., Lu, W., Hoelscher, B. K. P., Novosel, B., Tour, J. M. Synthesis of Dispersible
Ferromagnetic Graphene Nanoribbon Stacks with Enhanced Electrical Percolation Properties in a
Magnetic Field. ACS Nano 2012, 6(11), 10396-10404.
7. Tour, J. et. al. “Graphene Oxide as a High-Performance Fluid-Loss-Control Additive in Water-Based

Drilling Fluids” ACS Appl. Mater. Interfaces, 2012, 4 (1), pp 222–227.
8. Santos, C.; Mangadlao, J.; Ahmed, F.; Leon, A.; Advincula, R.; Rodrigues, D. “Graphene
nanocomposite for biomedical applications: fabrication, antimicrobial and cytotoxic investigations”
Nanotechnology 2012, 23, 395101.
9. Santos, C.; Tria, C.; Vergara, A.; Cui, K.; Pernites, R.; Advincula, R. “Fabrication and
characterization of electrodeposited thin films from highly dispersed poly(N-vinylcarbazole) (PVK)-
graphene oxide (GO) nanocomposites” Macromol. Chem. Phys., 2011, 212, 2371-2377.
10. Pernites, R.; Vergara, A.; Yago, A.; Cui, K.; Advincula, R. "Facile Approach to Graphene Oxide and
Poly(N-vinylcarbazole) Electro-Patterned Films" Chem. Comm. 2011, 47, 9810-9812.

CHALLENGES OF TEMPERATURE EXTREMES FOR

ELASTOMER MATERIALS
Glyn Morgan, Philip Clarke, Dr Salim Mirza, Dr Nickie Smith
Element Materials Technology Ltd
Wilbury Way, Hitchin, Herts, SG4 OTW, UK
Tel: +44 (0) 1462 427850 Fax: +44 (0) 1462 427851 email: glyn.morgan@element.com
BIOGRAPHICAL NOTE
Glyn Morgan is involved with behaviour of thermoplastics, elastomers and composites in

liquids and gases relevant to oil and gas production and exploration, with particular
reference to diffusion, permeation, rapid gas decompression resistance and chemical
ageing.
Use of polymers in seals, pipes, liners, packers etc. and material property relevance to
service. Current areas of interest include supercritical carbon dioxide, permeation of fluid
mixtures and testing for service life.
Member of ISO WG7/TC67 committee implementing ISO 23936 Non-metallic materials in contact with media
related to oil and gas production.
ABSTRACT
Elastomers function as seals, packers, barriers etc. by deforming against surfaces to prevent passage of
fluids. Their elasticity allows them to accommodate changes in temperature, pressure and movement in
ways that are impractical for ‘harder’ materials. However, at very low and very high temperatures this
elasticity may be compromised particularly when also under pressurised conditions causing the component
to lose its rubbery capabilities and cease to function as expected. The understanding and evaluation of
elastomers under HP and LT/HT conditions is still developing; this paper reveals some test methods,
observations and interpretation which should further this knowledge and provide insight into material and
component performance under these demanding thermal conditions. The following are discussed:
Elastomers at low temperature when pressurised; is Tg an effective measure of seal performance? The
effect of high pressure on the glass transition temperature of elastomers – are rubber-like properties lost?
Visual examination of seals as they experience energisation, pressure, low temperature; swelling,
contraction, movement, leakage.
Elastomers at high temperature; extrusion, gas decompression, chemistry.
Are test regimes such as API 6A (Appendix F.1.11: PR2) or ISO 10423 robust enough to capture all possible
HPHT failure modes possible during large extremes of temperature and pressure cycling?
How finite element modelling of polymeric components using appropriate material properties with
subsequent validation through functional testing provides added value engineering in critical thermal
applications.
INTRODUCTION
Elastomers have a long history in critical applications in the oil and gas industry with many instances of
successful operation for decades at high and low temperatures, high pressures and in potentially aggressive
media. But there are also examples of failure, leakage and disappointment that an elastomeric component
has not appeared to reach its potential.
Some of these setbacks can be linked to inadequate materials, poor design or misunderstanding of the
operational requirements or capabilities of the elastomer involved and with the increasing insistence on
reliability in the Oil and Gas sector and ways of measuring, monitoring and predicting the condition of all
Page 1 of 16 pages Paper 10 - Morgan

safety-critical hardware in the industry, it is becoming ever important that the polymer community is seen to
be actively addressing these issues.
One such issue is the behaviour of elastomers at extremes of temperature, both high and low. More and
more oil and gas fields are being developed where low temperatures are common (topside Arctic conditions,
for example) or high temperature production exists (deep water regions). That some of these also involve
high pressures can make any polymeric solutions even more challenging.
Elastomer sealing is one function where extremes of temperature can lead to problems which have not been
fully understood and sealing issues form a focus of this paper. For example, a typical ISO 10423i F.1.11 (API
6A) test might involve subjecting a sealing arrangement to pressure and temperature cycling between -18
and +121°C with pressure differentials of 690 bar; the seals must not leak during a set of pre-determined
hold points of pressure and temperature. Seals made of elastomers with theoretical low temperature
capabilities well below -18°C have been found to fail at the final test phase when pressurised.
Although not as mature as the metals industry in terms of understanding material degradation processes to
predict operational limits, there are areas where new insight is increasing knowledge and help avoid
problems and build confidence.
Element Hitchin have instigated some studies which have indirectly provided some insight into these
observations as well as the behaviour of elastomers at high temperature and pressure.
COLD TEMPERATURE
High Pressure
As noted above, several seal tests have been run in the laboratory to standards such as ISO 10423 or API
6A which have resulted in seals made of elastomers with reasonable low temperature properties failing the
low temperature requirement in the standard because of leakage at temperatures well above what the
elastomer should be capable of. This highlights one area of concern with this type of test (and more
importantly in service) which is that pressurising an elastomer raises its glass transition temperature by way
of reducing internal free volume and hence increasing stiffness. What this means in practice is that,
particularly at low temperatures, there is potential for seals to lose elasticity when at high pressure due to Tg
shift.
For years there has been a widely held belief that raising pressure approximately 50 bar causes an increase
in Tg of elastomers of 1°C but searching the usual sources and querying contacts has failed to reveal any
references, papers or test results to confirm this for relevant applications in the oil and gas field, although
other unrelated, studies have been made many years agoii,iii. Although scientists were very innovative in the
past, there is the distinct possibility that this relationship is based on theoretical calculations rather than
actual test results.
Therefore, Element Hitchin decided to investigate whether such testing was possible and where this could
lead. Fortunately, recent developments have led to DSC being used at high pressures in the oil and gas
industry for gas hydrate research and wax appearance temperatures in oil, as well as in the life science and
food industries and more generally throughout the polymers industry. We decided to run one set of tests to
determine whether Tg shift with pressure could be detected for a common elastomer type used in the oil and
gas industry with a view to performing more detailed studies depending on its success and potential.
An FKM Type 3 was chosen based on Viton® GLT which has a relatively low Tg. This was tested in nitrogen
at 1 bar, 250 bar, 500 bar, 750 bar and 1,000 bar using an HP-MicroDSC. Figure 1 shows the resultant Tg
versus pressure plot and its remarkable linear trend with the equation relating the two variables confirming a
rough 50 bar ≡ 1°C shift in Tg.
In practice this means that 1,000 bar pressure is capable of increasing Tg by 20°C which when combined
with contraction and housing/seal design could explain why seal leakage has been observed at temperatures
well above what Tg alone would suggest is expected. This begs the question – is Tg an effective measure or
predictor of seal performance at low temperatures? Two Joint Industry Projects (COLD and COLDX JIPsiv)
have investigated this whole area and a further iteration is planned looking at the influence of high pressure
on cold operations (COLD HP).
Paper 10 - Morgan Page 2 of 16 pages

-5
-10
y = 0.0198x - 33.372
-15 R² = 0.9973
Tg (°C)
-20
-25
-30
-35
0 200 400 600 800 1000 1200
applied pressure (bar)
Figure 1: Shift in Tg of an FKM Type 3 with pressure (in nitrogen)
Based on the successful high pressure test reported above, further testing is planned substituting methane
or carbon dioxide for nitrogen and choosing elastomers which will then swell as well as be compacted by the
pressurised environment. Is the 50 bar ≡ 1°C relationship universal for ‘all’ elastomers and gases?
One further point of interest is that there could be a limit to this relationship if all free volume is eliminated
from an elastomer; does a situation analogous to that shown in Figure 2 exist?
forced close
packing
permeation rate
limit of proportionality
pressure
Figure 2: Schematic plot of gas permeation rate versus applied pressure showing suggested
compaction effect of elastomers
Figure 2 was derived from gas permeation considerations, which is another process linked to internal free
volume and is reproduced from a paperv by Bob Campion which describes a ‘leathery’ state an elastomer
may assume when pressurised above its Tg. Work elsewherevi showed that gas diffusion through nitrile
decreased as pressure was increased. Presumably such a limit to compaction is dependent on elastomer
type, but what else, and would a similar relationship eventually occur for Tg versus pressure at high enough
pressures?

As more oil and gas fields are developed under high pressure, high temperature (HPHT) conditions, so these
considerations will become ever important – 1,400 bar testing of materials has already been undertaken and
this may increase to 2,000 bar in the near future. Similarly for temperatures; whilst Arctic exploration has hit
the headlines, lower temperatures are experienced by blow down situations and LNG applications and upper
temperatures have reached 315°C (600F) in the test lab. Even at high temperatures, if coexisting with high
pressures then shifts in Tg will occur and elasticity reduces.
HIGH TEMPERATURE
At higher temperatures, several elastomer failure mechanisms become more prevalent, especially when in
conjunction with high pressures and aggressive media. Amongst these is extrusion, where elevated
temperature makes materials weaker and softer and easier to tear. Also affected by temperature is gas
decompression resistance; higher temperatures make the diffusion process faster and solubility lower which
are compensated for by the weaker material which cannot resist bubble formation. And thirdly, chemical
attack is accelerated at higher temperatures for such species as hydrogen sulphide, corrosion inhibitors and
other treatment chemicals resulting in hardening (or softening), increased stiffness, loss of elongating
properties and loss of sealing properties through compression set and sealing force issues.
Extrusion
Seals function in fluid containment by bridging gaps caused by manufacture and assembly constraints and
serve to stop fluid leaking away. The manufacturing process defines the size of the gaps (extrusion gaps)
and the process conditions (pressure) determine whether additional features such as back-up rings or ant-
extrusion devices are required. Elastomers are often used as seals because they can accommodate
considerable misalignment and manufacturing compromises which lead to relatively large extrusion gaps,
eccentricity and surface finish qualities that are not acceptable for thermoplastic or metal seals.
However, at high temperatures elastomer seals need special treatment to function satisfactorily. The
following steps show what happens when things go wrong and what remedies are available.
A test at Element Hitchin used O-rings as face seals with pressurised treatment chemical on the inner
diameter with the whole assembly heated to test temperature. The purpose of the test was to measure the
sealing force acting on the seal using a load cell arrangement to measure the total force generated by the
seal and the environment and to monitor how this changed with ageing of the seal. To do this, a constant
dimension of extrusion gap was required which could be reproduced successfully every test. Unfortunately,
as can be seen from Figure 3, the extrusion gap was too large for the applied pressure and although the seal
force was seen to change, any extrusion was unwelcome as it compromised the intention of the test. By
reducing the gap and lowering the pressure, tests were subsequently run successfully. As an exercise,
further work was undertaken to determine whether finite element analysis (FEA) could have helped avoid
this problem from the start by predicting that excessive extrusion would occur. The project included a whole
host of mechanical tests on the O-ring materials such as double shear, tensile etc. so plenty of alternative
data was available for FEA modelling.
Figure 3: Badly extruded O-ring
Figure 4 shows an FEA representation of the O-ring in question with temperature and pressure applied and
the resultant strain levels experienced – 143% maximum. Extrusion has begun and continues with time.

Figure 4: FEA strain map of pressurised O-ring extruding into gap
This prediction was then compared to the stress strain data obtained for the test material as shown in Figure
5 for double shear, which was felt most closely represented the O-ring compressed-pressurised arrangement
(compared to tensile). Obtaining results at a range of temperatures allowed closer comparison with the test
situation and it was found that at 100°C the failure strain was about 100% which is sufficiently below that
predicted in Figure 4 to say confidently that failure and extrusion at the gap would occur, just as seen in
practice.
12
23C
10 50C
100C
150C
8
Stress (MPa)
2
Strain to failure
0 at 100C =
0 50 100 150 200 250
Strain (%)
Figure 5: Stress strain plot of seal material at various temperatures
A second test was performed at a lower temperature which the FEA analysis predicted would show less
extrusion with the result seen in Figure 6.

Figure 6: Less extrusion seen for test at lower temperature, as predicted by FEA
This shows the value of using FEA as a tool to give early warning of problems, predict material responses to
various inputs and to help optimise designs; appropriate, valid input data is essential to benefit from the
potential of this technique, as is the ability to verify the prediction by appropriate component testing.
For the above example, back-up rings would have eliminated the problem completely and are highly
recommended in service applications at high pressure but in this case would have compromised what we
were trying to measure.
Gas decompression
Damage to seals caused by rapid gas decompression (RGD) events has been a problem for many years but
several compounds have been introduced in that time which are resistant to the phenomenon and have
established service histories. However, as the pressures and temperatures of exploration and production
continue to increase, so the number of effective compounds will decrease. Once again, design can come to
our assistance because a well designed seal housing can reduce the effects of RGD, for example high levels
of groove fill will enable a moderately performing material to remain undamaged by RGD. Even so, there are
very few materials which are resistant to RGD at temperatures above 180°C and pressures in excess of 690
bar.
Testing for RGD resistance often takes the form of seals in custom-built hardware (Figure 7) designed so
that several materials can be tested simultaneously, with the best material progressing to further service-
specific testing or marketed as qualified to certain standards such as NORSOK M-710vii or ISO 23936-2viii.
Figure 7: Elastomer O-rings ready to be assembled into RGD test fixtures

Another result of this testing might take the form of a service envelope as shown in Table 1 where two
elastomers have been RGD tested across a range of pressures and temperatures to find the boundary at
which they can safely operate.
Table 1: Example of material test matrix and results
Test Temperature Pressure (bar)

Material
(°C) 350 200 150 35
180 FAIL FAIL PASS
150 FAIL FAIL
A
100 PASS PASS
75 PASS PASS
180 FAIL FAIL FAIL
150 FAIL PASS/FAIL
B
100 FAIL FAIL
75 PASS PASS
The usual inspection routine for RGD damage involves disassembling the fixture and observing internal and
external damage features such as shown in Figure 8, which shows a range of splits, blisters and cracks, all
of which become more prevalent as temperature is increased as confirmed in Table 1. The lower photograph
in Figure 8 shows one internal split with typical ‘rings’ each of which signifies a single decompression event
and an initiation point in the middle of the circle where the bubble first grew.
Figure 8: Examples of rapid gas decompression damage; worse as temperature increases
In a development of the inspection process, Element have taken a still-assembled metal fixture (right hand
side of Figure 7) containing a pair of tested O-rings and subjected it to X-ray CT scanning. Figure 9 shows
the inside of one of the pair of O-rings inside the fixture (scanning from the top of the fixture following the
length of the bolt in Figure 7), where the dark patches are damaged areas (confirmed on disassembly). The
second O-ring of this pair was an RGD resistant material - it had no such internal damage visible on the CT
scan. The resolution of the image is about 0.5mm (O-ring section 5.33mm) so small splits and inclusions are
all visible. This technique has great potential for inspecting valves etc. in-situ without the need for
disassembly, so could be a viable NDT tool in service.
sleev
O-
bolt
spigo damage
– dark
Figure 9: CT scan image of inside O-ring assembled in its metal RGD fixture
Chemical ageing
Chemical reaction rate is bound to temperature by the laws of chemistry and physics, so if reaction between
an elastomer and its environment is possible then it will accelerate at higher temperatures until service life
threatens to become unacceptably short.
Examples of chemical degradation are seen in Figure 10 where two elastomers respond differently to two
corrosion inhibitors; one hardens and cracks, the other softens and dissolves.
Figure 10: Different degradation modes of two elastomers in corrosion inhibitors
An example of deliberate accelerated ageing is shown in Figures 11 and 12 which demonstrate how an
HNBR deteriorates in a chemically hostile fluid. Both modulus and elongation at break change significantly
as the material embrittles at the exposure temperatures which are well above those expected in service.
36
32
28
24
50% Modulus (MPa)
20
160°C
16 170°C
180°C
12
0
0 5 10 15 20 25 30 35 40 45
Time (days)
Figure 11: Change in modulus of an HNBR at high temperature in a hostile environment
180
160
140
120
Elongation at Break (%)
100
160°C
80 170°C
180°C
60
40
20
0
0 5 10 15 20 25 30 35 40 45 50 55
Time (days)
Figure 12: Change in elongation at break of an HNBR at high temperature in a hostile environment
In order to estimate how quickly the modulus of the material changes by a set amount (here 50%), the
Arrhenius approach has been used as there is a clear relationship between temperature and property. Figure
13 shows the outcome and the quoted equation can be used to estimate material property ‘life’ at service
temperatures i.e. the time required for the modulus to increase 50%. The R2 value confirms a good
relationship exists across the test temperature range (the closer this number is to 1 the better the
relationship).

-2.6
ln(1/time to 50% change in modulus)

-2.8
-3
-3.2
-3.4 y = -7325.4x + 13.379
R² = 0.9991
-3.6
-3.8
-4
0.0022 0.00222 0.00224 0.00226 0.00228 0.0023 0.00232
1/Temperature (K)
Figure 13: Arrhenius plot for 50% increase in modulus of an HNBR
Of course there are many elastomers much more resistant to chemical change than HNBRs and for these
there is often very little change measurable at the test temperatures required for 150-180°C operations, see
Figure 14. Arrhenius plots are not feasible from such data so the conclusion is that the material has very long
life (in terms of property retention) at the anticipated service temperature. However, where operational
temperatures are very high there are challenges for materials, obviously, but also testing, where even higher
temperatures need to be used for acceleration purposes. Recently-introduced, premium grades of elastomer
are capable of operating at these very high temperatures without short term degradation.
5.0
4.5
4.0
3.5
3.0
Modulus (MPa)
2.5 200°C
215°C
2.0 230°C
1.5
1.0
0.5
0.0
0 5 10 15 20 25 30 35 40 45 50
Time (days)
Figure 14: Chemical stability of a fluoroelastomer in hostile chemical environment

TEST PROTOCOLS INVOLVING TEMPERATURE DEVIATIONS
ISO 10423 and API 6A use standard temperature classifications for equipment to cover their design and
operating conditions such as shown in Table 2 with the right hand column added in ISO 23936-2. The
functional testing of seals contained within the standards requires leak testing at the temperature extremes
using realistic representations of the actual equipment, or at least the seal system element with regards
housing dimensions (maximum tolerances), surface finish etc.
Table 2: Temperature limits for ISO 10423 testing
Temperature Operating minimum Operating maximum Elevated test

classification (°C) (°C) temperatures (°C)
K -60 82 97, 112, 127
L -46 82 97, 112, 127
P -29 82 97, 112, 127
R RT RT 36, 51, 66
S -18 66 81, 96, 111
T -18 82 97, 112, 127
U -18 121 136, 151, 166
V 2 121 136, 151, 166
X -18 180 195, 210, 235
Y -18 345 Not possible
Non-ISO/API 0 150 165, 180, 195
Bespoke As shall be agreed between interested parties
This involves following a pressure/temperature profile (part of the F.1.11 procedure) such as shown in Figure
15. If a seal is going to leak it most often does at the final low temperature excursion when the pressure is
applied (-18°C and 10,000 psi in Figure 15).
16000 140
14000 120
12000 100
temperature ('C)
10000 80
pressure (psi)
8000 60
6000 40
4000 20
2000 0
0 -20
0 20 40 60 80 100 120 140 160 180
pressure
time (hours)
temperature
Figure 15: ISO 10423 F.1.11 temperature/pressure schedule
Why this should be such a problem area for seals made of materials with Tg sufficiently low to expect
flawless operation at -18°C is still being investigated but three themes are being pursued.

i. COLD JIP experience has led to development of a robust low temperature functional seal test and a
much better understanding of how seals behave when cooled and pressurised and whether Tg is
relevant to functional performance. The sequence of pressure-temperature cycling is important, as is
the rate at which pressure is applied (a seal can be made to function at a much lower temperature if
pressure is applied quickly enough). Figures 16, 17 and 18 show test sequences for 1 material
where pressure is applied to a seal after it is cooled or before it is cooled, leading to 3 different
leakage/resealing temperature measures.
Figure 16: Application of pressure to a seal after it is cooled - resealing
Figure 17: Application of pressure to a seal before it is cooled – leakage

Figure 18: Application of pressure to a seal before it is cooled – subsequent resealing
A strong correlation was found between the low temperature functional limits and the Tg of the seal
material in the ‘unaged’ state of 15 elastomer types. After liquid swelling, the low temperature
functional limits and the Tg of the material both decreased but not by the same amounts. However,
with thermal ageing, the low temperature functional limits increased (reduced performance) although
the Tg of the material was unchanged. Thus, the Tg of the sealing material alone is not sufficient to
explain all of the observed behaviour of seals in service environments and efforts are continuing in
this area with a proposed JIP: COLD HP.
ii. Visual observation of seals and materials has enabled direct monitoring of the effects of heating or
cooling, as well as what happens when pressure is applied and removed. Direct observation of
pressurised seals squeezed against a sapphire window has provided the following evidence that
seal movement under the influence of applied pressure is a primary influence on sealing capability;
o Seals cooled with applied pressure remain locked in their pressure-energised shape when
leakage occurs – no movement or loss of contact width is observed
o Seals reseal after cooling (with or without applied pressure) at a temperature corresponding to
the onset of low level seal movement at the seal ID
o Constrained face seals when cooled without applied pressure show a reduction in contact width
but complete loss of seal contact does not occur
The ability of the seal to move and change shape under applied gas pressure appears to be more
critical to low temperature sealing than the attainment of a particular level of contact force in the
unpressurised state resulting from stress relaxation of the material, even if sealing force reaches
zero.
o Low temperature functional limits are governed primarily by the ability of the seal material to
deform when pressurised; that is, they depend on seal stiffness and hence why the low
temperature functional limits have been found to correlate well with Tg. Reseal temperature
limits are generally 0-20C lower than the Tg of the material, as measured by DMA according to
the project test procedures. Hence Tg is a reliable and conservative basis for seal selection,
provided that significant thermal ageing does not occur during service.

iii. Use of FEA incorporating suitable mechanical and thermal data into models to mimic the behaviour
of seals undergoing both laboratory testing and service loadings such as the simple thermal
expansion shown in Figure 19 forms a powerful tool when combined with the schematics of what can
happen in the sealing cycle expressed in Figure 20 and already seen in Figure 4.
a) squeeze
b) temperature rise to
100°C
Figure 19: Thermal expansion of O-ring
(a) Initial State (e) Pressure
(b) Assembly (f) Viscoelastic Analysis

Creep under constant P
(c) Squeeze (g) Return to ambient

conditions
(d) Temperature (h) Unload

rise 100C
Figure 20: Seal Life Cycle: Modelling of Mechanical, Thermal and Viscoelastic Response
The intention is to model the ISO 10423 F.1.11 test cycle using material property data including
stress relaxation rate to determine what parameter is most influential in the loss of performance after
the third low temperature excursion; stress relaxation, thermal contraction, sealing force, stiffness,
proximity to Tg etc. Predictions will be validated by testing.
CONCLUSIONS
In summary, temperature extremes can have a significant impact on elastomer performance of seals,
bonded hoses, flexible joints etc. and the following points have been found:
o Cold temperatures when combined with high pressures can result in raised Tg and contribute to
unexpected seal leakage.
o High temperatures facilitate extrusion, RGD and chemical ageing. Avoidance measures can be
taken at the design stage and using new, highly thermally resistant materials.
o Standard test methods incorporating temperature cycling aspects of seal applications may need
investigation to determine whether they are too severe or are in some way unrepresentative of
service by causing premature failure of otherwise good sealing systems.
o Combinations of testing, visual observation and FEA will continue to advance our understanding
of temperature (and pressure) effects on seals and related elastomeric products.

REFERENCES
i
ISO 10423:2009 Petroleum and natural gas industries — Drilling and production equipment —
Wellhead and christmas tree equipment.
ii
Gee, G., The thermodynamic analysis of the effect of pressure on the glass temperature of
polystyrene, Polymer, 7 1966, 177.
iii
Pae, K.D., Tang, C.L. & Shin, E.S., Pressure dependence of glass transition temperature of
elastomeric glasses, Journal Applied Physics, 56 (9) (1984) 2426.
iv
COLD JIP run by MERL and Phil Clarke defined a test method for measuring the temperature at
which pressurised seals actually leaked; concentrated on correlating Tg with leakage and using
unaged seals. The follow on COLDX JIP used aged or swollen seals to represent service to
investigate how low temperature performance changed.
v
Campion, R.P & Morgan G.J., High pressure permeation and diffusion of gases in polymers of
different structures, Plastics, Rubber & Composites Processing & Applications, 17 (1992) 51-58
vi
Briscoe, B.J., Liatsis, D. & Mahgerefteh, H., The pressure dependent diffusion of carbon dioxide in a
nitrile rubber, Proc Conf (Reading) Diffusion in Polymers, Plastics & Rubber Institute, London, 1988,
paper 17.
vii
NORSOK M-710 Qualification of non-metallic sealing materials and manufacturers Rev 2 2001.
viii
ISO 23936-2:2011 Petroleum, petrochemical and natural gas industries — Non-metallic materials in
contact with media related to oil and gas production Part 2: Elastomers

COLD TEMPERATURE EFFECTS ON POLYMERS -

CRYOGENIC SPILL PROTECTION
Sebastien Vialea*, Laurent Pomiea, Romain Legentb
a
Advanced Subsea Architecture, Technip Innovation and Technological Center 43-45 Bd Franklin
Roosevelt, Rueil-Malmaison, 92500 France.
b
AETECH/Cybernetix Rue les Rives de L’Oise, Compiegne, 60201 France
sviale@technip.com
BIOGRAPHICAL NOTE
Sebastien Viale, 38 years old, Hold a Ph. D. in Polymer Chemistry from Technical
University Delft, The Netherlands. Has been working more than 15 years in different
fields from Liquid Crystal synthesis to coating production in various universities and
companies around the world. Join Technip two years ago, in charge of the polymer
issues at corporate level in the newly founded Innovation & Technological Center.
Focus on these following topics: Electrical Isolation, Insulation for subsea pipe, High
efficiency insulation materials for offhsore application and finally in charge of the
material screening for the cryogenic spillage protection.
ABSTRACT
UNAVAILABLE AT TIME OF PRINT
Introduction
Carbon steel has a tendency to become brittle when temperature rapidly decreasesi. Even special grade
Carbon Steel such as Low Temperature Carbon Steel (LTCS) used for LNG decks displays a ductile-brittle
transition temperature (DBTT)ii. Operating Liquid Natural Gas (LNG) liquefaction units in offshore conditions
has raised new safety considerations. LNG being a cryogenic medium, the protection of hulls and topsides
assets against cryogenic spillage is among the most critical concerns, as the load bearing structures cannot
be design against these accidental cases.
Figure 1, Effect of accidental release of LN2 on Carbon Steel
Experiences learnt from LNG carriers indicate that cryogenic spillage is a real factor to take into account
while designing itiii. Cryogenic spillage is a common risk in offshore facilities where sprayable materials are
used to protect steel items avoiding them to reach their DBTT. Combining those feedbacks, wet applied
products are the current materials of choice when it comes to cryogenic spillage protection.
Page 1 of 8 pages Paper 11 - Viale
It is important to mention that today there is no international standard or recommended practice to test the
performance of such materials in case of rapid decrease of temperature. Therefore we propose a more
quantified approach to the different stake holders. Technip has started to coordinate a Joint Industrial
Project, involving operators, OEM and engineering contractors.
The objectives of this project are multiple:
 To determine accidental exposure criteria to rank the performance of protection materials (i.e. type
of cryogenic threat – liquid or vapour, protection duration and threshold temperature, to maintain
stability of the exposed item)
 To screen and to assess performances of different others technologies of protection materials. The
only solution existing, to date, is intumescent epoxy originally developed to protect substrate from
high temperatureiv.
 To establish a testing set-up and protocol representative of real conditions and quantifying
performances against accidental exposure criteria.
CAPEX and OPEX value engineering of alternate solutions have been scrutinized and new product design
easing IMR (Inspection/Maintenance/Repair) during service life have been encouraged.
The philosophy of this project is to offer to the market a wide range of materials technologies for cryogenic
spillage protection, including new possibilities of integration other than wet-applied one.
The integration of CSP materials being design and project dependent, Technip has decided to carry out a
baseline survey on testing a very conservative set-up: this test has been named Technip Proof Test (TPT).
This test as well as the lessons learnt during the JIP has been currently used as foundation for an ISO
standard and this paper will introduce the long journey from internal protocol to international standard.
In the following, we will present Technip’s product design when it comes to horizontal surface, the TPT and
finally we will introduce the standard derived from the TPT.
Background
As mentioned above, materials used today for cryogenic spillage protection are usually wet applied epoxy
resins. Such materials are directly sprayed onto the substrate via special pumps and spray guns.
Figure 2, Typical application of wet applied epoxy. (Photo courtesy of PPG)
The main advantage of these materials is the perfect adhesion onto substrate. But this great feature is also
their major drawback when it comes to inspection, maintenance and repair (IMR). It is almost impossible to
remove easily materials slabs without damaging a large surface. Moreover, in terms of safety, applying
materials on the yard requires full personal protection equipment for operators and handling large quantity of
chemicals.v To be complete, those systems need at least 24 hours to fully cure.vi For Technip, safety is a key
parameter; therefore; we decided that prefabricated panel should be the product design for horizontal
application.
Paper 11 - Viale Page 2 of 8 pages

As can be seen on Figure 3, casting will be performed at supplier’s factory. This approach has several
advantages and can be summarized as follow:




 Reduce installation time
 Improve quality
 Highly tunable (possible to apply anti-skid coating at paint shop stage)
 Reduce HSE impact at the yard
 Improve IMR
Figure 3, Prefabricated Panel for wet applied technology
By promoting prefabricated panels for wet applied epoxies, it becomes obvious that installation time can be
tremendously reduced since normally 24 hours are necessary for a fully cured system. In the other hand, out
of the shelves epoxy glues only require less than 4 hours to cure, since glue application procedure is similar
to the wet applied one, same applicators can perform similar job without extended training.
It is also important to mention that other technologies interesting for CSP application are usually supplied as
free standing panels, for example: sandwich panels, woods. By giving the opportunity to the coating industry
to develop casted products from wet applied formulations, we will test all technology in the fair approach
meaning in exact configuration.
Technip Proof Test
We decided to test only free standing panels independently of raw materials forms i.e., solid or liquid.
Moreover, for samples adhesion issues, gluing has been selected and we decided to use bi-components
epoxy glue Araldite 2015 due to its track record in terms of cryogenic application.vii In order to mimic the hull
to protect, a square meter of P275NL1 carbon steel (8 mm thick) has been used as substrate. Planarity of
such sample holder has been carefully tuned and requirement was 1mm deviation on 1m length.
Figure 4, Schematic of test set up of first version of TPT.
As seen in Figure 4, two exposure conditions have been tested in parallel during the TPT: a liquid cryogenic
pool exposure for the bottom part and vapor exposure for the top part. The material to be tested should have
a minimum 1m by 1m size. Thickness is fixed by sample owner. The side walls should have a height of 15

mm and a maximum width of 10 cm. The bottom panel is exposed to liquid nitrogen (LN2) at -196°C during
120 minutes with a minimum of 3cm of liquid for complete test duration.
For cryogenic liquid spillage (pool), the panel is directly glued with Araldite 2015 to the CS panel of 8mm
thickness and 5 thermocouples are attached to the back-face surface of the CS panel as shown in Figure 5.
Figure 5, Localization of Thermocouples underneath CS panel
For the cryogenic vapor exposure, the setup is closed with a 1*1m sample equipped with thermocouples
inside and outside, in the same position as the TC2 (Figure 5). This sample will be exposed to the vapor of
LN2 generated by the lower pool. All thermocouples of type K will record the overall temperature evolution
every 2 seconds. Lower Thermocouples are glued onto the CS plate using cryogenic cyanoacrylate glue,
CC-33A from Kyowa.viii. Upper thermocouples are attached with heavy duty tape.
The embrittlement temperature (TE) of the metallic material to be protected is conservatively chosen as -
20°C (hull structure) and -40°C (structural steel and process equipment in first approach).
As illustrated in the figure below on the right, this temperature (TE) is defined at the interface between
protecting and protected materials.
Figure 6, Definition of the Temperature embrittlement for the TPT

To select CSP protecting materials, criteria were fixed according to what was just explained about
embrittlement temperatures. The average temperature on the whole 1*1 m steel plate backface surface
should not go under -40 °C. At least three of the thermocouples should have an individual temperature
higher than -40 °C for both pool and vapor condition.
After the 2 hour exposure to liquid and vapor nitrogen, visual aspect of the exposed materials should be
done according ISO 4628-4 and crack density should not exceed rating 3. Size of the crack should not
exceed rating 3.ix
TPT, test procedure and sample preparation have been fully scrutinized and analyzed to validate our
technical choice by the Cryogenic Laboratory at Kennedy Space Center.
During the JIP test campaign, more than 30 different samples were tested. Since CSP is a major safety
related subject, Technip decided that no intellectual property whatsoever will be claimed and further than
that, TPT will be the foundation for an ISO standard.
From internal test to international standard
TPT is a Technip requirement in term of product design and integration. Therefore an intense work of retro
engineering was necessary to generate an international standard.
In the JIP, we focused our studies only on a cryogenic event while an accidental cryogenic spillage of LNG
could rapidly develop into a fire event. We seriously took this sequence in consideration. However fire
standards already exist to estimate materials performance under those conditions.x-xi We did not want to
include such test in our standard, therefore based on learnt lessons, we decided that the only technical
solution to do a fire test after cryogenic exposure was to ensure that sample size was exactly the same as
per ISO 22899-1.9 By doing so, we ensure that operators will have enough time to swap samples from one
test set-up to another. In our study, it took an average of 60 minutes to 120 minutes for samples to return at
room temperature after cryogenic exposure. We would like to stress the point that not only LNG can be
found in large inventory on board of FLNG but LN2 as well with no subsequent risk of fire.
With this parameter fixed, standard characteristics can be discussed. The main driver is the universality of
the test, so we did rework on the four main components of the TPT namely:
 Test set-up
 Sample preparation
 Test Procedure
 Report
Worldwide results should be similar. Since temperature is the major parameter: external temperature and
sample temperature should be controlled. So outdoor conditions are therefore rejected and indoor tests
should be done in a thermostated room at 20~25°C and average sample temperature will set at 25~30°C.
We will perform liquid and vapor with separate set-ups in order to follow continuously crack generation in
pool configuration (by camera recorder namely).

Pool walls will be constructed from CS plates and sample holder will be different for each technology. For the
wet applied one, the pool will be built with 10 mm thickness CS as shown in the figure below.
Figure 7, Sample holder for wet applied technology
As you can see from Figure 7, wet applied product will be prepared according to supplier’s procedure and
not anymore cast and glued onto a carbon steel plate. Coating industry was really interested to have a
sample holder where their products will be tested in a configuration close to field application.
One of the most stringent conditions to maintain during the test is the thermal gradient on sample. One face
is rapidly cooled down while, in the meantime, the opposite side can remain at room temperature for a long
period of time. This thermal gradient induces severe sample contraction and leads to cracks. In our TPT, this
phenomenon was clearly underlined thanks to sample gluing, therefore sample holder need to impair
specimen movement.

Figure 8, Sample holder for pre-fabricated panel (LHS, bottom part, RHS top part)
You can notice diamond cut jaws all around the frame top and bottom parts; this will reduce the movement of
the sample. In order to achieve liquid tightness, cryogenic silicon based sealant can be applied all around the
frame.
Sample size is bigger (1.5m*1.5m), therefore number of thermocouples has been increased and their
location under the sample is shown in the figure below.
Figure 9, Thermocouples location (LHS for free standing panel, RHS for wet applied)
Thermocouples are K type due to their quick response time as well as temperature range. Each
thermocouple need to be tested in cryogenic fluid prior to use and need to be new for each run.
Samples are either free panels and therefore squeezed between jaws or they are directly wet applied into a
pool shape sample holder. Sample size as discussed above will be the exact same as per ISO 22899-1.9
Test procedure will be decided by specimen owner mainly: test duration, TE. LN2 is injected directly in the
center of the sample and normal to the surface. The test is stopped when catastrophic failure is observed
(time recorded), when the TC average temperature reaches the predetermined temperature limit (time
recorded) and finally, at the end of the predetermined maximum test duration (temperature recorded). LN2
level is kept constant around 3~5 cm during the test.
Conclusion
Standard creation is necessary when it is related to safety. The purpose of such standard is to give all
players in the FLNG market a tool to study material performance.
This standard will also give new technologies the opportunity to compare with the actual ones.
ACKNOWLEDGEMENT:
SV and LP would like to thank all members of the JIP for funding this research program as well as the
Technip's Offshore Division especially J-M Letournel and J. Arjona. Without the trust and the support of all
the materials suppliers, this JIP will never be a success, thanks again. The authors would like also to thank
all the team at Cybernetix in Compiegne for their patience. HSE Design department did an outstanding job
during this work and also to review this manuscript.
REFERENCES
i
Degarmo, E. Paul; Black, J T.; Kohser, Ronald A. (2003), Materials and Processes in Manufacturing
(9th ed.)
ii
MacGregor, C.W; Grossman, N.; Shepler, P.R. Weld. J. Res. Suppl (1947), 50.
iii
M. Foss, LNG Safety and Security, Center for Energy Economics, November 2006.
iv
NASA Technical Note D-4713, pp. 8, 1968
v
http://www.cdph.ca.gov/programs/hesis/Documents/epoxy.pdf
vi
http://www.international-pc.com/PDS/2045-P-eng-usa-LTR.pdf
vii
http://www.huntsman.com/portal/pls/portal/docs/49849641.PDF
viii
http://www.kyowa-
ei.co.jp/eng/product/strain_gages/gages/adhesives_leadwirecables/adhesives.html
ix
ISO 4628-2:2003, Paints and varnishes -- Evaluation of degradation of coatings -- Designation of
quantity and size of defects, and of intensity of uniform changes in appearance -- Part 2:
Assessment of degree of blistering
x
ISO 22899-1:2007 Determination of the resistance to jet fires of passive fire protection materials --
Part 1: General requirements
xi
ASTM E1529 - 13 Standard Test Methods for Determining Effects of Large Hydrocarbon Pool Fires
on Structural Members and Assemblies

ARLON 3000XT: A NEW HIGH PERFORMANCE MATERIAL

DESIGNED FOR EXTREME ENVIRONMENTS
Kerry Drake, Senior Scientist, and Burak Bekisli, Scientist
Greene, Tweed & Co
Email: KDrake@Gtweed.com and BBekisli@Gtweed.com
BIOGRAPHICAL NOTES
Kerry Drake has a PhD in Polymer Chemistry. He is a Senior Scientist at Greene, Tweed
and Co. He leads a team of researchers tasked with the development of new materials and
new technologies. His research focus includes polymer chemistry and materials engineering
of high performance polymers.
Dr. Burak Bekisli recevied his PhD degree in Mechanical Engineering from Lehigh
University in 2010 and has been employed as a Scientist at Greene Tweed since 2011. His
expertise is in testing, numerical analysis and structure-property relationships of advanced
polymers for oilfield and other applications.
ABSTRACT
Arlon ® 3000 XT is a patent-pending, high performance polymer system that has been developed by
Greene, Tweed specifically for demanding oilfield applications. The material has been engineered for
enhanced high temperature properties, while maintaining excellent chemical compatibility for most common
oilfield fluids.
Arlon 3000 XT has been optimized through the use of experimental methods and predictive models
developed by Greene, Tweed to effectively and efficiently simulate performance in harsh high pressure high
temperature (HPHT) environments. Arlon 3000 XT tested with this new methodology yielded superior
extrusion resistance over glass and carbon-filled polyketones at conditions ranging up to 35,000 psi and
550°F(288°C). Product tests performed on back-up rings and electrical connectors under conditions similar
to the laboratory tests have shown that Arlon 3000 XT significantly outperformed current best-in-class
unfilled materials. Test results showed Arlon 3000 XT’s high temperature stability and creep resistance from
350°F(177°C) to 600°F(316°C) exceeded all other polyketones tested without sacrificing chemical resistance
or any other key properties.
Use of these new test regimes and predictive models along with validation through real world product testing
show enhanced performance and reliability of Arlon 3000 XT relative to other polyketones over a wide range
of conditions. Depending on product requirements, the advantages of Arlon 3000 XT can be seen at
temperatures starting at 300°F(149°C) to 350°F(177°C), with increasingly greater performance advantages
seen at HPHT conditions.
1. INTRODUCTION
1.1. BACKGROUND
The modern world is incredibly reliant on fossil fuels and hydrocarbons for energy. 55% of the world’s energy
is derived from oil and gas, and over the next decade annual consumption is predicted to increase by 10%
for liquid hydrocarbons and 20% for hydrocarbon gas [ Energy Information Administration].
Hydrocarbon reserves are located throughout the world both on land and under the sea. When one analyzes
global reserves relative to depths, the following general trends are observed. The bulk of oil reserves fall
within a window between 7,000 and 18,000 feet and a temperature range of 150F(66C) to 300F(149C).
Page 1 of 22 pages Paper 12 - Drake

At temperatures above 300F(149C) reserves tend more towards gas than liquid. A clear correlation
between depths of wells and temperature is also seen (Figure 1.1.1).
Figure 1.1.1 Relationships between depth of reserves vs. temperature and reserve type, oil vs.gas [Hyne].
Typically reserves at higher temperatures and greater depths were not pursued actively in the past due to
production difficulties (technology gaps) [Baird et Al] and the belief that reserves at greater depths were of
poor quality for profitable extraction [Total]. However, over the last several decades, improved surveying,
predictive modeling, and exploratory drilling has identified more high quality reserves at greater depths
(Figure 1.1.2) Some estimates now predict that almost one third of all probable reserves are located in these
deeply buried reservoirs.
Figure 1.1.2. Percentage of world’s estimated reserves in reservoirs with depths greater than 4,000 meters
(13,000 feet) [Total].
The increased demand for energy has led to increased exploration and development of more reliable
technology for production in these deep reserves. Many reserves that were once too difficult to tap are now
being actively drilled. Large gas reserves under thousands of feet of salt, such as the Davy Jones reserve,
are now being considered for production. This reserve in particular has an estimated bottom hole
temperature of 440F(227°C) and pressure of 27,000 psi [Beims].
The improved economics of petroleum and gas production in more extreme environments is now driving
technology innovation.New technology is needed to be able to produce from these reserves. Technology
needs range from new cementing technology to improved metallurgy for corrosion resistance at higher
temperatures, to new materials for seals. Seals in particular were identified as the major technology gap.
23% of respondents at the 2012 HPHT Wells Summit in London listed seals it as the major gap, up from
10% of respondents at the 2010 summit [Shadravan, et al].
Paper 12 - Drake Page 2 of 22 pages

Clearly the industry is moving in this direction, and some of the largest gaps in servicing these reserves are
in materials, particularly polymeric and sealing materials and seals performance in extreme environments. In
order to better serve the industry, Greene, Tweed has implemented an integrated approach to developing
and substantiating new materials solutions that incorporates cutting edge materials chemistry development
as well as extensive testing and modeling capabilities to simulate extreme conditions.
1.2. MATERIALS CHEMISTRY RESEARCH
Greene, Tweed has decades of experience in successful development of new elastomer and thermoplastic
compounds to address the needs of the oil and gas industry. However, given the specialized nature of
polymer enhancement needed to address higher temperature oil and gas applications coupled with
aggressive chemistries, fundamentally new polymer solutions were needed.
Over the last several years a concentrated effort was initiated to build a deeper understanding of the
structure/property relationships of oilfield polymers, and then target development of new materials that
address the weak points and failure modes of current best in class materials. Details of the initial
structure/property analysis and resulting material property targets will be discussed in greater detail in
Section 2. Results of material property tests on Arlon 3000 XT will be presented in Section 3, and product
tests will be presented in Section 4.
1.3. HPHT SEALS RESEARCH
In response to the increasing need for reliability in seals and other polymer applications at HPHT conditions,
Greene, Tweed has initiated a multi-year, comprehensive research program. The HPHT Seals program aims
at understanding the limits of current materials for these challenging conditions and defining the required key
properties for the next generation materials and products. In the materials side of the project, effects of a
large number of factors; including temperature, pressure, time and fluid compatibility, have been studied
extensively. In addition to more standard experiments to observe the effects of each factor on polymers
independently, specialized equipment and test procedures have also been developed to combine the effects
of multiple factors in single tests. One such test is used to measure creep or extrusion resistance of
thermoplastics at high temperature, high pressure and elongated durations [Bekisli, et al. 2012, Drake and
Bekisli, 2013]. This test was re-visited during the development of Arlon 3000 XT and will be discussed in
more detail in later sections. In another custom test, stress relaxation properties of candidate polymers have
been investigated in representative fluid media (such as H2S containing ISO mix) by immersing a custom
design test fixture seen in Figure 1.3.1 (a). HPHT experimental capabilities have been greatly increased by
investing on HPHT-proper tools like a 100 kN servo-hydraulic universal tester with a temperature rating up
600°F (315°C) with video extensometry (Figure 1.3.1 (b)), and digital pressure vessels appropriate for HPHT
fluid aging studies (Figure 1.3.1 (c)) and product tests at HPHT.
(a) (b)
Figure 1.3.1 (a) Custom test fixture to be used at HPHT stress relaxation tests in fluid media, (b) Servo-
hydraulic universal tester with video extensometry and high temperature testing accessories and (c)
pressure vessels and venting hood for HPHT fluid aging studies.
Investigation of materials in the HPHT program does not only involve experimental methods but also
development of proper numerical techniques; such as non-linear, time-dependent finite element analyses
(FEA). Although it is currently in the developmental stages for reliable prediction of polymer behavior at
HPHT, employment of FEA and similar tools seem to be an inevitable step for obtaining a time and cost-
efficient life-time prediction capability. As an example, pioneering studies on the extrusion prediction of
thermoplastics, [Bekisli et al. 2012], have shown promising accuracy and encouragement for further work.
Although the modeling aspects are still in developmental stages, the actual material tests and correlations to
product performance have been proven robust. This new testing regime has been used extensively in an
iterative fashion to develop and optimize Arlon 3000 XT. Multiple variants were tested and screened via
coupon testing, and then validated through testing of material in final product form. Advanced HPHT coupon
testing and predictive tools were integral in the development of the Arlon 3000 XT material platform. Results
of coupon testing will be presented in Section 3.
2. ARLON® 3000 XT: DEVELOPMENT
2.1. MATERIAL TARGETS FOR DEVELOPMENT
In order to develop new materials, one must first determine the important properties needed for the
applications of interest, and then test and benchmark incumbent materials to understand performance limits.
Once this information is collected, and correlated with polymer structures, materials with modified structures
can be designed to improve performance in the desired areas.
One way to approach this problem is through the analysis of failures of materials in applications of interest.
As a baseline, a review of common failure modes of plastics can be used as a starting point (Figure 2.1.1).
Common Failure Modes of Plastics

120%
100%
80%
60%
40%
20%
0%
Chemical Creep Others Fatigue UV attack Thermal
Resistance Degradation
Failure Mode Cumulative Total
Figure 2.1.1. Common failure modes of plastics (adapted from [Scheirs]).
As can be seen from Figure 2.1.1, chemical resistance and creep related failures combined account for over
50% of total plastics failures. When thermal degradation is added, these 3 areas combined make up almost
two thirds of all plastics failures in the field.
For oilfield materials, especially in aggressive environments, these failure modes are even more critical
(higher likelihood of failure through these modes). Chemical resistance, thermal properties and creep
resistance are intimately related to each other; weakness in one area often carries over to other areas. For
example poor chemical resistance can reduce creep resistance, lower thermal properties relate to increased
creep at temperature [Menard] and lower Tg vs. operating temperature(higher polymer molecular mobility)
results in increased chemical attack due to increased permeation/diffusion of chemicals into the
polymers[Duda].
For most effective targeting of new materials development for oil and gas service, those attributes which
have been identified and validated through field use and wide industry acceptance of materials are as
follows:
1) Chemical resistance: polymers must exhibit broad chemical resistance to common oilfield
chemistries, otherwise they will severely limited in scope of applications.
2) Thermal properties: current best in class thermoplastics materials have glass transitions Tg in the
300°F(150°C) range or higher. Best in class thermoplastics are usually semicrystalline, as this allows
for use at temperatures above the glass transition, albeit mechanical properties are usually
significantly lower than what is obtained at temperatures below Tg[Cogswell].
3) Mechanical properties at high temperatures: properties at application temperatures are often the
most critical weak points of materials in aggressive service environments, especially related to
creep. In addition, creep becomes progressively worse at higher temperatures, i.e. higher strain is
obtained for equivalent stress at equivalent times when a material is exposed to higher service
temperatures (Figure 2.1.2).
4)
Figure 2.1.2. Effect of temperature on creep. Note the increase in strain and strain/unit time at higher
temperatures under the same loading conditions[Menard].
It is the combination of these three attributes that is required for new polymers targeted for oilfield service.
Some materials have one or two of the required attributes, but still see limited overall acceptance. For
example, polyimides have high glass transition and continuous use temperatures but very poor chemical
resistance, particularly hydrolysis [Campbell]. Thermosets such as BMIs and epoxies also have good
thermal and creep properties with lower continuous use temperatures than polyimides, but very poor
chemical resistance as well. Due to the chemical resistance limitations, they are not used as often in
applications with potential exposure to aggressive chemistries such as seals or sealing components.
3. ARLON® 3000 XT: MATERIAL PROPERTIES
Arlon 3000 XT is a modified version of PEEK with highly improved mechanical properties at elevated
temperatures over 320°F(160°C). It was designed specifically to maintain best in class chemical resistance
of PEEK, while significantly enhancing high temperature mechanical properties and performance. Electrical
properties, which are important for connectivity applications, were also maintained at comparable levels to
other polyketones.
In this section, a review of the extensive chemical, thermal, mechanical, and electrical tests will be
presented and the potential advantages of this new material over current best-in-class polyketones will be
demonstrated.
3.1 CHEMICAL COMPATIBILITY
Hundreds of chemicals are used in our industry; testing in each particular formulation would be extremely
difficult and would require extensive re-testing each time a fluid composition was slightly changed. The
strategy Greene, Tweed developed for materials screening was to survey the most commonly accepted
categories of fluids and solvents, and then select representative oilfield fluids from each major category for
in-depth testing. Testing in this manner provides a broad probe of chemical resistance to most commonly
used classes of oilfield fluids. Many formulations are based on aqueous chemistries, so testing in hot water,
steam, brines/completion fluids and aqueous control line fluids covers this area. Hydrocarbon fluids are
covered with oil based control line fluid and ISO hydrocarbon mixture tests (exposure to a mixed
hydrocarbon with H2S, the ISO/NORSOK standard test fluid, was included to simulate exposure to
production hydrocarbons).
This chemical screening strategy is detailed graphically in Figure 3.1.1.
Figure 3.1.1. Common categories of fluids/solvents most commonly seen in the oilfield industry.
The final compositions of test fluids used for initial screening were as follows:
Media Test Temperature Category

Steam 450°F (232°C) Aqueous
Hot water 450°F (232°C) Aqueous

ISO-23936/NORSOK M-710 Hydrocarbon aromatic,
450°F (232°C)
(Sour aging with gas, high H2S) high H2S
Oil based control line fluid* 400°F (205°C) Non-aqueous
Water based control line fluid* 350°F (177°C) Aqueous
Zinc Bromide 400°F (205°C) Aqueous, acidic
Cesium acetate 400°F (205°C) Aqueous, basic
Table 3.1.1. Chemical compatibility screening matrix, 7-day immersion tests per ASTM D543. *=tested at
maximum manufacturer rated service temperatures.
3.1.1 ISO/NORSOK SCREENING TESTING
The mixture composition for this testing is listed in Table 3.1.1:
Liquid Phase, % Gas Phase, % Mol Composition

Volume Volume
60% 70% Heptane / 20% Cyclohexane / 10% Toluene

30% 5% CO2 / 10% H2S / 85% CH4
10% 10% Sea Water (3% NaCl)
Table 3.1.1 fluid mixture composition for ISO chemical resistance screening tests.
The ISO fluid test showed a small decrease in tensile strength of about 10%, which was well within the
acceptance range for this test ( change less than +/- 50% is acceptable for tensile strength
[NORSOK]).Testing of PEKEKK showed much greater changes in properties, which exceeded the Norsok
cutoff values for acceptable use. Note that the ISO aging temperature performed on Arlon 3000 XT was
18°F(10°C) to 36°F(20C) higher than temperatures typically used for PEEK ISO/Norsok certification.

3.1.2 RESULTS
Outside of the ISO test fluid, no statistical differences were seen between chemical resistance of standard
PEEK and Arlon 3000 XT in any properties in the majority of these tests. No differences larger than the ISO
cutoff were seen in any of the aging tests for Arlon 3000 XT [Drake]. These results demonstrate that
chemical resistance of Arlon 3000 XT in most common classes of oil field chemistries at elevated
temperatures is excellent.
3.2 THERMAL PROPERTIES
Higher glass transition temperatures are preferred for high temperature service. Semicrystalline polyketones
have a well known trend of increasing glass transition temperatures as the ketone or biphenyl () content is
increased (Table 3.2.1). Unfortunately, the melting point increases with higher polyketones or biphenyl
content at a higher rate than the glass transition temperature; the end result is that materials with melting
points well above 752F(400C) that are not readily processable [McGrail].
(PEEK)
(PEK)
(PEKEKK)
O
O
O
=
e
t
h
e
r
C
O
=
k
e
t
o
n
e
a
r
y
l

Table 3.2.1 Glass transition temperatures (Tg) and melting temperatures (Tm) of PAEKs, showing trends of
increasing Tg, Tm with ketone content (adapted from [McGrail]).
In addition, PAEKs with higher ketone to ether ratios have been shown to have lower chemical resistance
than PEEK in very aggressive chemistries [Ren et al]. Clearly there is a need for a higher glass transition
PAEK material without the associated increase in ketone content.
Differential scanning calorimetry (DSC) and dynamic mechanical analysis (DMA) are often used to measure
thermal properties of polymers. Unfortunately, for Arlon 3000XT, DSC and Modulated DSC analysis (not
shown) were not able to readily detect a Tg.
Dynamic mechanical analysis (DMA) Figure 3.2.1 was able to detect a glass transition onset at
325F(163C), which was equivalent to that of PEKEKK.

Figure 3.2.1 DMA data of Arlon 3000 XT in comparison with PEEK and PEKEKK.
The overlapping curves at lower temperatures below Tg indicated a similar mechanical stiffness between the
three materials. However, once the glass transition is reached, significant differences occur and the high
potential of Arlon 3000 XT becomes clearly apparent. Both PEEK and PEKEKK suffer from a significant and
quite similar drop in modulus, almost immediately after their respective glass transitions. Therefore, the only
major advantage offered by PEKEKK over PEEK by this test seems to be the shift of Tg by about
36°F(20°C). The same benefit can also be achieved with Arlon 3000 XT which shows an improved Tg very
similar to that of PEKEKK.
Additionally and more importantly, an increased retention of modulus is also observed with Arlon 3000 XT in
the so-called rubbery plateau region. About an order of magnitude decrease in the mechanical stiffness
gradually occurs almost in a linear fashion at the approximate temperature range of 338°F(170°C) to
662°F(350°C), while such a drop in mechanical properties is immediate for both PEEK and PEKEKK at
around glass transition. In other words, highly improved mechanical properties can be expected from Arlon
3000 XT at this temperature range. Considering the typical polymer limitations on the load bearing
applications of oil and gas industry, such an enhancement in high temperature properties should be quite
desirable.
When samples were examined after testing, it was found that Arlon 3000 XT no longer exhibited
conventional melting behavior (flow at high temperatures). Even after extended testing over several hours at
temperatures above752F(400C), Arlon 3000 XT still maintained its structural and mechanical integrity
(Figure 3.2.2).

Figure 3.2.2 DMA samples before testing (top) and after testing (bottom). Note the maintenance of structural
integrity for Arlon 3000 XT, even after extended temperature excursions of 1-2 hours at752F(400C).
(PEKEKK also melts under these conditions, at slightly higher temperatures) .
3.3. MECHANICAL PROPERTIES AT HIGH TEMPERATURES
In order to validate and determine the level of enhancement in high temperature mechanical properties, a
comprehensive set of tensile, compression, flexural and shear tests (ASTM D638, ASTM D695, ASTM D790
and ASTM D732, respectively) were conducted at various temperature points.
Figure 3.3.1 Typical tensile test data of Arlon® 3000 XT at 392°F(200°C) compared to PEEK and PEKEKK
From the observation of the DMA data, a temperature range near 392°F(200°C) is likely to demonstrate a
significant level of mechanical property differences between Arlon 3000 XT and its more traditional
counterparts. Typical tensile stress-strain curves from the three materials at this temperature are plotted
together in Figure 3.3.1 for comparison. As expected, both modulus and strength of Arlon 3000 XT are
significantly higher than PEEK and PEKEKK at this temperature, reaching over 2-3 times higher modulus
and 1.5-2 times higher strength compared to both PEEK and PEKEKK. In the other deformation modes,
test data also reveals a similar trend; Arlon 3000 XT consistently outperforming its competitors with about
50-200% higher strength and stiffness values. A summary of the resulting data is presented relative to PEEK
in Figure 3.3.2.

Figure 3.3.2 Enhanced mechanical properties of Arlon® 3000 XT at 392°F(200°C) relative to PEEK.
As a second temperature point, 500°F(260°C) is also of great interest since it is generally considered a
temperature point where the capabilities of currently available polyketones start to become very limited and
insufficient for many current and future applications. Figure 3.3.3 shows typical tensile curves of the three
materials considered at500°F(260°C). Similar to the results at 392°F(200°C), mechanical performance of
Arlon 3000 XT is considerably greater than the traditional polyketones, with improvements in modulus and
strength reaching up to 1.5-2 times higher values in most cases, compared to PEEK. A comprehensive
comparison of properties at this temperature is presented in Figure 3.3.4.
Figure 3.3.3 Typical tensile test data of Arlon® 3000 XT at 500°F(260°C) compared to PEEK and PEKEKK.

Figure 3.3.4 Enhanced mechanical properties of Arlon® 3000 XT at 500°F(260°C) relative to PEEK.
3.4 Creep Properties at High Temperature
Tensile and other forms of tests performed at high temperatures verify the dramatically improved mechanical
properties of Arlon 3000 XT that were inferred from DMA results. Higher strength and modulus values are
generally desired properties of a high temperature polymer used at load bearing applications but they are
often not sufficient by themselves. Most polymer applications in the oil and gas industry requires the use of
these materials not only at high temperatures and high pressures but also for long durations, exceeding tens
of years in some cases. For such applications, the time dependent properties of the material tend to be
extremely critical. In particular, time related failure modes such as creep are generally a major concern at
high temperature applications and the ideal polymer is one that has excellent resistance to plastic flow. In
addition to the higher mechanical properties determined by static tests, Arlon 3000 XT is also optimized for
an increased creep resistance at high temperatures. In this section, results from some of the time-dependent
tests are presented.
Figure 3.4.1 Shear creep data for Arlon® 3000 XT, PEKEKK and PEEK at 500°F(260°C) and under a stress
level of 10.5 MPa(1,520 psi).
One of the simplest methods to evaluate the creep properties of materials is to apply a pre-defined load on a
sample and then continuously record the amount of deformation with respect to time while the load is held
constant. For instance, Figure 3.4.1 shows the shear strain vs. time data of PEEK, PEKEKK and Arlon 3000
XT materials when a torsional stress of 10.5 MPa (1,500 psi) is applied on samples at 500°F(260°C) and
held for about an hour. For such a stress level, PEEK material immediately deforms to a shear strain level of
over 16% and continues to flow as time progresses. Despite having a much smaller instantaneous strain
compared to PEEK, PEKEKK material also flows considerably, up to about 10% strain at the end of the 1
hour test. On the other hand, thanks to the enhanced high temperature modulus and creep properties, Arlon
3000 XT deforms to only about 7% strain level and remains relatively constant at this level during the time
frame of the test.
In a more comprehensive set-up, the three materials were tested for creep performance following the
procedure described by ASTM D2990 and using a constant tensile stress of 10 MPa at 500°F(260°C). The
tests continued for about 8 hours for each specimen. The resulting strain vs. time data for each material are
plotted together in Figure 3.4.2, yielding a graph similar to the DMA shear creep results (Figure 3.4.1). In
order to compare the resistance of the material to time dependent deformation, a comparison of modulus
values obtained at four time points (namely; instantaneous, 1 hr, 3 hrs and 8 hrs) are plotted relative to the
values of PEEK in Figure 3.4.3. As can be clearly observed, Arlon 3000 XT is about 1.7 times stiffer than
PEEK at the start of the test but the difference continues to increase as testing progresses and approaches
to about 2.5 times higher after 8 hours. A similar but slightly less dramatic trend is also observed between
the creep moduli of Arlon 3000 XT and PEKEKK. Therefore, Arlon 3000 XT not only has the highest creep
modulus of the three but also has the slowest creep rate, i.e. the highest resistance to plastic flow under
sustained loading.
Figure 3.4.2 Tensile creep response of Arlon® 3000 XT, PEKEKK and PEEK materials at 500°F(260°C) and
under a stress level of 10 MPa (1,450psi).
Figure 3.4.3 Progression of creep moduli differences between Arlon® 3000 XT, PEKEKK and PEEK
materials during the 8hr test. Test is performed at 500°F(260°C) and under a stress level of 10 MPa(1,450
psi).

3.5 TIME DEPENDENT PROPERTIES AT HPHT CONDITIONS: EXTRUSION RESISTANCE
A major difficulty in evaluation of polymers for the challenging conditions of the oil and gas industry is to
combine the effects of many factors such as time, temperature and pressure in a simple, time and cost-
effective manner. Previously, a test method designed to evaluate the time-dependent extrusion resistance of
a polymer at HPHT conditions was presented by our research group [Bekisli et al., 2012]. In this very simple
but extremely effective test, a cylindrical sample is placed in a confined volume and pressed using a piston
of a slightly smaller diameter than the sample. A schematic of the test set-up is given in Figure 3.5.1. The
compressive force applied by the piston generates an almost hydrostatic pressure state around the sample
while pushing the material to extrude through the small clearance between the piston and the housing
(extrusion- or e- gap) at the same time. By controlling the pressure, temperature and duration of the constant
load application, a time-dependent test similar to the creep test can be performed at HPHT conditions. The
amount of extruded material (hext) measured at various time points can be used to evaluate and compare the
creep resistance of the polymers for HPHT applications. Although the test method is particularly designed to
closely simulate the extrusion of a seal back-up ring, it should also be applicable to compare other materials
for other applications as well, including electrical connectors that could see sustained differential loads at
high temperatures.
Figure 3.5.1 Illustration of the HPHT extrusion test developed at Greene, Tweed [Bekisli et al. 2012]
Figure 3.5.2 is a reproduction of our previously published data [Bekisli et al., 2012] with the addition of Arlon
3000 XT extrusion performance. The data is generated after tests at 241 MPa (35,000 psi), 550°F(288°C)
and varying test durations; from instantaneous to 6 hours. An extrusion gap of 0.51 mm (0.020”) was used
for all tests and the amount of extrusion height measured immediately after the completion of the test.
Figure 3.5.2 HPHT extrusion performance of Arlon 3000 XT as compared to unfilled and filled grades of
PEEK and PEKEKK. Tests were performed at 241 MPa (35,000 psi), 550°F(288°C) and using an e-
gap=0.51 mm (0.020”).

Not surprisingly, Arlon 3000 XT easily outperforms the extrusion performance of its unfilled counterparts,
PEEK and PEKEKK. The photograph in Figure 3.5.3 demonstrates the differences in the amount of extrusion
between an Arlon 3000 XT (left) and a PEEK sample (right) after a 1 hour test. Based on the amount of
extruded material, Arlon 3000 XT shows 10 times or more extrusion resistance at this extreme HPHT
condition. Despite a much lower extrusion and a more stable time response than PEEK, PEKEKK also
suffers from a significantly higher amount of extrusion compared to the new material.
Figure 3.5.3 Arlon 3000 XT (left) and PEEK (right) samples after HPHT extrusion tests at 241 241 MPa
(35,000 psi), 550°F(288°C) and for a duration of 1 hour.
Even though these are quite promising results for Arlon 3000 XT, what may really be impressive for most
people is the comparison of Arlon 3000 XT with the extrusion performance of filled PEEK and PEKEKK.
Based on this HPHT extrusion test, we find that Arlon 3000 XT shows very similar or better extrusion
behavior when compared to 30% carbon filled grades of PEEK and PEKEKKs. This is an extremely critical
outcome since it may open the doors for replacement of filled grades with an unfilled polyketone and still
obtain a similar extrusion or creep performance with a tougher and more chemically-resistant material.
3.6 ROOM TEMPERATURE AND OTHER FUNDAMENTAL PROPERTIES
Development of Arlon 3000 XT targeted the improvement of high temperature mechanical and creep
properties of PEEK, while conserving the excellent compatibility and resistance of PEEK to common oilfield
chemicals. However, most HPHT applications may also require a desired level for the room temperature
properties of the polymer. For instance, heavily filled thermoplastics may easily outperform the unfilled
versions at high temperature stiffness, strength, and creep resistance, but this benefit generally comes at the
expense of a greatly reduced ductility and toughness. In many cases, products from these filled grades
(sealing components in particular) are difficult to install and risk catastrophic failure under impact conditions.
Therefore, their use may not be preferred despite their extraordinary high temperature properties.
in order to ensure the appropriateness of room temperature properties, the development of Arlon 3000 XT
has been performed with a strict consideration of common requirements. Table 3.6.1 summarizes some of
the room temperature data along with other key material properties in comparison to PEEK.

®
Property Standard Units PEEK Arlon 3000 XT
Tensile Modulus
ASTM D638 psi 595,000 570,000
(Room Temperature)
Strength at Break
ASTM D638 psi 14,000 16,500
(Room Temperature)
Elongation
ASTM D638 % 25-35 8-15
(Room Temperature)
Compressive Strength
ASTM D695 psi 19,900 22,300
(Room Temperature)
Shear Strength
ASTM D732 psi 14,100 16,100
(Room Temperature)
Impact Toughness (Notched) ASTM D256 ft-lbf/in 1.38 1.64
Impact Toughness
ASTM D4812 ft-lbf/in No break 37.8
(Unnotched)
Hardness ASTM D2240 Shore D 86 87
Specific Gravity (g/cc) ASTM D792 1.31 1.28
Glass Transition Temperature

ASTM D5279 °F (°C) 289 (143) 325 (163)
(DMA onset)
Heat Deflection Temperature ASTM D648 °F (°C) 338 (170) >572 (>300)
-5
CTE (T<Tg) ASTM E831 (mm/m per °F x10 ) 2.6 2.5
-5
CTE (T>Tg) ASTM E831 (mm/m per °F x10 ) 7.5 5.6
Table 3.6.1 Comparison of some basic material properties of Arlon 3000 XT to PEEK (* For electrical
properties, baseline material is PEK due to higher use in electrical applications).

®
Arlon 3000
Property Standard Units PEK
XT
*Dielectric Breakdown
ASTM D618-A kV/mm 15.5 15.2
Voltage
16 17
*Surface Resistivity ASTM D257 Ohm 6.62x10 1.95x10
17 17
*Volume Resistivity ASTM D257 Ohm-cm 4.03 x10 3.49x10
Table 3.6.2 Comparison of some basic electrical properties of Arlon 3000 XT to PEK (*For electrical
properties, baseline material is PEK due to its common use in high temperature electrical applications.)
As can be seen from the data, properties of Arlon 3000 XT at ambient conditions are similar to PEEK. The
modulus is slightly lower while the strength is about 15-20% higher than that of regular PEEK. Due to its
more rigid structure, elongation is slightly reduced to a range of 8-15% but this is still much higher than
typical elongation capabilities of glass and carbon-filled grades [Victrex]. Although the elongation capability
of unfilled PEEK is much higher, it should be noted that most of the applications do not utilize this full range
of elongation, and strains over 10-15% are rarely observed. Therefore, on a practical level similar room
temperature mechanical performance between PEEK and Arlon 3000 XT should be expected based on the
tensile data.
Although tensile elongation is a bit lower than standard PEEK, impact tests showed Arlon 3000 XT has
better notched impact resistance than unfilled, carbon filled and glass filled PEEK [Victrex]. Impact related
failures with PEEK and derivative polymers are generally related to their notch-sensitivity and presence of
design or processing faults like sharp corners, cracks or material impurities. 20% more energy is required to
break notched Arlon 3000 XT samples compared to regular PEEK. Arlon 3000XT had somewhat lower
unnotched impact strength than unfilled PEEK, but significantly higher unnotched impact strength than
carbon or glass filled PEEK [Victrex]. These results indicate Arlon 3000 XT should have a much lower
sensitivity to the presence of notches, sharp corners or already established cracks in the products.
Therefore, it can be a reliable solution where PEEK is already considered acceptable. In addition Arlon 3000
XT can provide broader design freedom for applications where creep resistance is critical, notch effects are
problematic, and filled grades do not provide the required toughness.
As shown in the preceding data, ambient temperature properties (hardness, specific gravity, CTE below Tg,
etc.) are very similar between PEEK and Arlon 3000 XT. However, when properties at elevated temperatures
are considered, the true differentiation of Arlon 3000 XT becomes apparent. For instance, the glass transition
temperature of Arlon 3000 XT is about 36°F(20°C) higher than PEEK and the heat deflection temperature
based on ASTM D648 testing is 338°F(170°C) for regular PEEK and more than 572°F(300°C) (temperature
limit of the test equipment) for Arlon 3000 XT. Also due to the higher modulus and thermal stability at higher
temperatures, the coefficient of thermal expansion (CTE) of Arlon 3000 XT is much lower than that of PEEK
above Tg.
Finally, comparison of electrical properties which may be extremely critical for the insulation reliability of
connectors also shows no significant difference between the modified and traditional versions of PEEK. Note
that listed values for the baseline correspond to PEK which is a more commonly used polymer for electrical
connectors.

4. Arlon® 3000 XT: Product Tests

4.1 Back-up Ring Tests
One of the major target applications for Arlon 3000 XT is the back-up rings used in sealing systems to
support the sealing element and to provide a last line of defense in case of a failure. For such applications,
stiffness and creep resistance of the material is of greatest importance and therefore Arlon 3000 XT is a
perfect candidate for many back-up ring applications at HPHT.
In order to verify and demonstrate the improved product performance with this new material, back-up rings
made from Arlon 3000 XT and PEEK were tested side by side. For this study, back-up rings with an outer
diameter of 31.75 mm (1.25”) were used together with elastomeric seals (FFKM O-rings) to seal a test fixture
as illustrated in Figure. 4.1.1. The system was brought to 450°F(232°C) and pressurized up to 40,000 psi
using a water based control line fluid Transaqua HT2) as the pressurizing medium. For the pressurization,
the profile shown in Figure 4.1.2 was used. After a total test duration of 48 hours, the back-up rings were
removed from the test set-up and the deformation on each sample was evaluated. Particular attention was
given to the extrusion in the axial direction where the pressure differential is applied.
Figure 4.1.1 A schematic showing a back-up ring supporting a sealing element.
Figure 4.1.2 Pressure profile used in the testing of back-up rings.

The tested samples were section-cut at several locations and the amount extrusion in the axial direction was
measured by using a Keyence VHX1000 digital microscope. Typical images of the cut sections before and
after testing are shown for both regular PEEK and Arlon 3000 XT in Figure 4.1.3. In agreement with many of
the previously described tests, PEEK had three times worse extrusion performance than Arlon 3000 XT at
these conditions. The average extrusion values for PEEK and Arlon 3000 XT were 0.53 mm (0.021”) and
0.18 mm (0.007”), respectively. Therefore, full scale product tests in the form of HPHT back-up rings confirm
that the improved material properties of Arlon 3000 XT translated very well to product performance. Carefully
optimized mechanical properties of Arlon 3000 XT increased the stability and deformation resistance of the
back-up rings, resulting in a more reliable support to the full seal assembly.
Figure 4.1.3 Sections of back-up rings seen before and after the HPHT testing. Amount of extrusion is
compared between PEEK and Arlon 3000 XT.
4.2 Electrical Connector Tests
Another common HPHT application where PEEK and PEK based materials are reaching their limitations are
electrical connectors. These products are generally electrical circuits or sensors which are sealed by an
insulating polymer. However, the connector in this case also protects the electrical assembly from the
adverse effects of temperature, and down-hole chemicals and provides the necessary structural rigidity and
stability at pressures reaching 30,000 psi. Coupon testing showed that Arlon 3000 XT has good electrical
properties (equivalent to PEK, see table 3.7.2). Coupling good electrical properties with enhanced creep
resistance should equate to better overall product performance, but testing was needed to confirm.
For the comparative pressure testing, single-pin connectors from both Arlon 3000 XT and PEK material were
injection molded (see Figures 4.2.1 and 4.2.2). PEK was selected for this study as it is currently considered
the best-in-class unfilled polymer for this specific product.
Using a test arrangement as illustrated in Figure 4.2.1, evaluation of the combined effects of temperature
and pressure on electrical connector samples was targeted. Over a test duration of 8 hours, the pressure
and temperature profiles as shown in Figure 4.2.1 (right) were applied on the samples. Combinations of
three maximum temperature —350°F(177°C), 389°F(198°C), and 428°F(220°C) — and three maximum
pressure points —20,000 psi, 25,000 psi, and 30,000 psi —were used in various tests. When the tests were
complete, the deformation on the connectors were first inspected visually and then evaluated by measuring
the changes in dimensions d1, d2 and d3 as shown in Figure 4.2.1 (left).

Figure 4.2.1 Illustration of HPHT test set-up for electrical connectors (left) and Temperature/Pressure profiles
used during the 8-hour tests (right).
Figure 4.2.2 shows a visual comparison of electrical connectors from PEK and Arlon 3000 XT after testing at
several temperature/pressure combinations. At 20,000 psi / 350°F(177°C), neither any significant
deformation nor any major visible differences were observed on the connectors of two materials. In reality,
this condition is known to be a limit for the use of PEK material on this particular application. This becomes
clearer after observing the extreme deformation and failure of the PEK connectors at 25,000 psi
/389°F(198°C) and 30,000 psi /428°F(220°C) . However, structural stability and functionality of Arlon 3000
XT connectors remain valid even at these conditions, thanks to its inherent high temperature properties.
Figure 4.2.2 Comparison of deformation observed on PEK and Arlon 3000 XT connectors, tested at various
temperature/pressure combinations.
As clearly demonstrated, the use of Arlon 3000 XT in electrical connectors shows great promise in extending
the boundaries of connector applications to higher temperatures and pressures. In addition, at the conditions
where satisfactory performance may be available from traditional polyketones, Arlon 3000 XT can also be
employed for improved reliability and better structural stability. For a numerical comparison, Figure 4.2.3
shows the changes in the d1 dimension of Arlon 3000 XT connectors before and after the test at various
temperature and pressure conditions. As a baseline, the same dimensional change observed in PEK at
20,000 psi / 350°F(177°C)(which is the highest condition PEK connectors did not fail) is also plotted next to
the values from Arlon 3000 XT connectors. For example, for the same conditions where PEK was deformed
by about 0.028”, the deformation for the Arlon 3000 XT remains five to six times lower. Even though the
deformation seen on Arlon 3000 XT gradually increases with increasing temperature and/or pressure, it
remains below the baseline level even at 30,000 psi and 350°F(177°C).

0.035
PEK Arlon 3000 XT
0.03
Change in dimension d1 (inch)

0.025
0.02
0.015
0.01
0.005
0
PEK Arlon 3000 XT Arlon 3000 XT Arlon 3000 XT Arlon 3000 XT
(20K, 350°F) (20K, 350°F) (25K, 389°F) (20K, 428°F) (30K, 350°F)
Figure 4.2.3 Comparison of the change in d1 dimension before and after pressure tests at several
conditions.
Testing of Arlon 3000 XT in product form showed significant advantages over incumbent materials.
Customer feedback and testing conducted thus far on this new material have verified the product
performance enhancements which were discussed in the preceding sections.
5. Summary
Arlon 3000 XT is a new material designed specifically for aggressive high temperature applications.
Principles of materials engineering and HPHT testing and modeling were utilized in the development of this
material, and product testing was used to validate its performance and advantages over other polymers.
With Arlon 3000 XT, Greene, Tweed has developed a new enabling polymer technology that should provide
performance advantages in applications where high temperature performance coupled with excellent
chemical resistance are key requirements. Arlon 3000 XT can be fabricated into finished shapes and
products through a variety of processes, similar to current PAEK materials. It is an unfilled polymer that
exceeds creep resistance of filled polymer grades, without the sacrifice in impact strength or toughness seen
when fillers are used. Product testing has shown its properties translate well into existing product forms of
back-up rings, electrical connectors and other sealing components.
Figure 5.1 details comparative properties of this new material vs. industry standard polymers PEEK and
PEKEKK (PEEK=best in class chemical resistance, PEKEKK =best in class high temperature performance
before the introduction of Arlon 3000XT).

Figure 5.1. Comparative properties of Arlon 3000 XT vs. PEEK (best in class chemical resistance) and
PEKEKK (former best in class high temperature mechanical properties)
Data presented shows that Arlon 3000 XT provides an excellent combination of properties that when taken
together provide significant performance advantages over all other commercially available polyketones. Its
chemical resistance is similar to best in class materials; its creep resistance is superior to all other
polyketones, and the ability to fabricate it into a variety of shapes gives great flexibility for its potential use in
new products. Its enhanced features should provide an increased safety factor in demanding applications,
and thus enhanced reliability over incumbent materials in oil and gas service.
6. REFERENCES
1. www.eia.gov, data tables of global hydrocarbon consumption.
2. Hyne, N. J. In Nontechnical Guide to Petroleum Geology, Exploration, Drilling, and Production, Penwell
Books: 2001.
3. Baird, T., Drummond, R., Langseth, B., Silipigno,L. “High pressure, high temperature well logging,
perforating, and testing”, Oilfield Review, Summer 1998, 51-67.
4. Total, Inc “The Know How Series.Exploration and Production: Deeply Buried Reservoirs, New
Conquests”, 2007.
5. Beims, T.”Davy Jones Discovery Opening New Shelf Frontier in Ultradeep Geology Below Salt”,
American Oil & Gas Reporter, April 2010.
6. Shadravan,A., Amani, M. HPHT 101-what pertroleum engineers and geoscientists should know about
high pressure high temperature wells environment,Energy Science and Tech,(2012),4,36-60.
7. Scheirs, J. In Compositional and failure analysis of polymers: a practical approach; John Wiley and
Sons: 2000.
8. Menard,K. In Dynamic Mechanical Analysis: A Practical Introduction; 2nd edition, CRC press: 2008.
9. Duda, J., RomdhaneI, I., Danner, R., “Diffusion in glassy polymers- relaxation and antiplasticization”,
Journal of Non-Crystalline Solids 172-174 (1994) 715-720.
10. Campbell, F. “Temperature Dependence, of Hydrolysis of Polyimide Wire Insulation”, Naval Research
Laboratory Memorandum Report 5158 ,1983.

11. Cogswell, F. N. In Thermoplastic Aromatic Polymer Composites, Butterworth-Heinemann: 1992.
12. NORSOK M-710 Standard “ Qualification of non-metallic sealing material and manufacturers”, Revision
2, 2001.
13. Ren,J. et al, “Testing and Challenge of High-Performance Thermoplastics for Oil and Gas High
Pressure/High-Temperature Sealing Applications”, ANTEC,2012.
14. McGrail, P, “Polyaromatics”, Polymer International, (1996),41,103-121.
15. Drake K., Bekisli B., “Arlon 3000 XT: A New High-Temperature Material for Oil and Gas Applications”,
Energy Rubber Group 2013 Fall Technical Meeting, Galveston, TX, Sept 2013.
16. Bekisli B., Aripirala A., Thoman R., “Numerical and Experimental Analysis of Polymer Behavior at
HPHT; A Case Study: Back-up Ring Extrusion”, MERL Oilfield Engineering with Polymers Conf., 2012.
17. "Victrex Material Properties Guide", available at www.victrex.com, 2007.

FATIGUE OF RUBBER AND PLASTIC MATERIALS

Dr Andrew Hulme & Jenny Cooper
Smithers Rapra & Smithers Pira Limited
Shawbury, Shrewsbury, Shropshire, UK SY4 4NR
Tel: +44 (0)1939 252306 Fax: +44 (0)1939 250383 email: ahulme@smithers.com
BIOGRAPHICAL NOTES
Dr Andrew Hulme is a Principal Consultant in plastics at Smithers Rapra, where he has

worked since 2001, providing independent advice on plastics design and manufacturing
to all industries. He specialises in providing durability and lifetime predictions for plastic
components in their operating environments. This involves providing material selection,
design advice on suitability of materials & manufacturing, injection moulding & FEA
simulations, dimensional management and the generation & use of long term design
data to improve confidence in designs. Prior to Smithers Rapra, Dr Hulme worked in the
automotive industry and in composites manufacturing. He is a materials science
graduate from Imperial College & has a PhD from the University of Birmingham.
Jenny Cooper is the Commercial Manager for the provision of technical services to
support the Industrial Sector at Smithers Rapra. She has specific responsibility for clients
such as oil and gas companies, raw material suppliers, compounders & distributors,
utility companies, electrical & electronic device manufacturers and companies in the
construction industry. Prior to working at Smithers Rapra, Jenny worked for GKN, initially
working on continuous fibre-reinforced resin systems for suspension and propshaft
applications. She was then the Polymer Test Laboratory Manager and later the Global
Technology Manager for the development of rubber & TPE boots used on automotive
constant velocity joints. She has a materials engineering degree from Loughborough
University.
ABSTRACT
There is a need to generate fatigue data to demonstrate the long term durability of polymers for use in Oil &
Gas applications. Fatigue testing is specified in standards such as ISO 13628-16 / API 17L1 (Specification
for flexible pipe ancillary equipment), but no prescribed test method is identified. The generation of long-term
data is particularly important for polymeric materials as their properties are not only time and temperature
dependent but can be significantly affected by the fluids they come into contact with.
This paper discusses the different models used for the prediction of fatigue life in rubber and plastic
materials and includes an overview of both crack nucleation and crack growth test approaches. The
commonly used International standard methods for rubber and plastic materials are also identified for
reference. The standard methods are generally limited to tests under laboratory conditions and therefore
custom material or product tests are often required to meet business needs. Since the early 1980’s, Smithers
Rapra has been generating fatigue data for other industries and offers solutions for providing engineering
data for the prediction of lifetime in the operating environment.
A test methodology for fatigue tests under fixed load or displacement is described which can be used in
either tension or flexure. The design of the test fixture enables it to be immersed into the test environment
under controlled temperature conditions. With the expected demand in the future for higher operating
temperatures and pressures, test equipment and methods continually need to be developed to enable
accelerated life predictions to be carried out.
FATIGUE METHODOLOGY
Introduction
When designing with rubber or plastics for a particular application, it is relatively straight forward to calculate
the expected stresses and strains associated with operational load conditions. However, as material
properties provided on data sheets, such as tensile strength, tend to be based on the short term or
Page 1 of 12 pages Paper 13 - Hulme
instantaneous behaviour of the material, the big question is, how will the component perform over a 10 to 20
year lifespan and beyond? Therefore it is important to understand the long term behaviour of the material.
Typically this involves either accelerated fatigue or creep testing, depending on the loading conditions.
Models for predicting fatigue life in rubber follow two overall approaches:
 Predicting crack nucleation based on the applied stress or strain levels

 Predicting growth of a particular crack given the initial geometry and energy release rate history of
the crack based on ideas from fracture mechanics.
Finite element stress analysis on CAD models can predict the maximum stress or strain values but fatigue
performance is affected by many variables other than just the stress level. These variables include
frequency, waveform and environmental effects (temperature, ozone levels, fluids, pressure). It is therefore
difficult to isolate specific test parameters for material testing which simulate the range of conditions the
product may experience.
In addition, fatigue life is very much dependent on the compound formulation, cure state (if appropriate) and
process history. For example, rubbers can be formulated for improved fatigue performance through selection
of the best polymer grade, filler type & loading, protective system and cure type. In addition, mixing quality
will influence the number and size of defects (voids, carbon black agglomerations etc.) from which cracks will
propagate.
For plastics, the material behaviour tends to be more uniform but is affected by processing conditions.
Residual moulding stresses and crystallinity can have a significant effect on the fatigue performance of a
component.
Many academic studies are often carried out on idealised compounds such as unfilled polymers. The tyre
industry has also undertaken a significant amount of research on fatigue; it should be noted that tyre
compound formulations generally incorporate strain crystallisable polymers. Strain crystallisation has the
effect of blunting the crack tip reducing the rate of crack propagation through the rubber. This effect occurs
for natural rubber, high cis-isoprene, chloroprene and non-phenyl silicone.
The following discussion has been supplemented with information from “A literature survey on fatigue
analysis approaches for rubber” by W V Mars and A Fatemi which includes 150 references and is published
in the International Journal of Fatigue 24 (2002) 949-961.
Crack Nucleation
The crack nucleation approach for predicting fatigue considers that a material has a life determined by the
history of stress or strain at a specific value. The fatigue crack nucleation life may be defined as the number
of cycles required to cause the appearance of a crack of a certain size. The two widely used fatigue life
parameters are ‘maximum principal strain’ and ‘strain-energy density’. The strain-energy density of a material
is defined as the strain energy per unit volume and is equal to the area under the stress-strain curve for the
material.
In fatigue, cracks only form when the material is pulled apart either under a tensile or shear stress. Fatigue
testing of polymers is complicated by extension of the sample due to creep during the test. This can result in
the test piece buckling if the test piece is forced to return to its original starting position. This can be
overcome by maintaining a minimum strain on the test piece; this is expressed as the R-ratio (minimum
strain : maximum strain, as shown in Figure 1). For rubbers which strain crystallise, it has been reported that
increasing the minimum strain (i.e. increasing the R-ratio) can significantly lengthen the fatigue life.
Paper 13 - Hulme Page 2 of 12 pages

Figure 1. Fatigue loading showing an applied R-ratio (min/max)
Crack Growth
The crack growth approach for predicting fatigue explicitly considers pre-existing cracks or flaws. Crack
growth is due to the conservation of a structure’s stored potential energy to surface energy associated with
new crack surfaces. In polymers, the potential energy released from the surrounding material is spent on
both reversible and irreversible changes to create the new surfaces. The energy release rate (or tearing
energy T) is simply the change in the stored mechanical energy (∆U), per unit change in crack surface area
(∆A).
The two most commonly used specimens for fatigue crack growth studies are;
 The single edge notched/cut planar tension specimen (or pure shear specimen)
 The single edge notched/cut simple tension
Fracture mechanic studies using test pieces of different thickness have found that the influence of thickness
on crack growth rates is greatest in thin specimens. To measure the number of cycles to rupture, standard or
modified tensile specimens are normally used, see Figure 2. Modified specimens with a minimal gauge
length have the advantage of concentrating rupture within a well-defined region and elongation of the
specimen is minimised.

Figure 2. Single edge notched tensile (SENT) fatigue sample, where fatigue crack growth is measured
up to a defined limit. The number of cycles to reach this limit is determined at different stress
amplitudes. More often than not the limit is defined as the point when the specimen ruptures.
Fatigue crack growth in rubber materials depends on the polymer type. Non-crystallising rubbers, such as
SBR, exhibit time dependent growth of cracks provided the energy release rate is above a threshold value;
this has been found to be independent of the type of test specimen (centre cracked sheet, edge cracked
sheet and a trouser tear test). Strain crystallising rubbers do not exhibit time dependent crack growth unless
the energy release rate is very high, approaching that at which catastrophic (single cycle) failure occurs.
General Comments
The published fatigue papers demonstrate that fatigue life is affected by many test variables (for example,
thickness, R-ratio, stress state, loading history, frequency). If the influences of the environment e.g. sea
water or hot oil, along with the possibility of variable material quality are also taken into account, it makes it
extremely difficult to predict the life of a component. Researchers have spent a lifetime studying relatively
simple geometries and idealised materials and still don’t have all the answers.
The nucleation approach has received little attention in the literature, although popular with engineers for its
simplicity and familiarity. The crack growth approach has been used more widely for modelling but one of the
limitations is that it requires upfront knowledge of the initial location and state of the crack that causes the
final failure.
The difficulty predicting fatigue life of components from material fatigue tests is that the applied stresses and
strains are usually multi-axial in nature. The literature review suggests that the use of uniaxial tensile fatigue
and energy release rate has limited value for these applications.
Depending on the objective of the work, it may be simpler to assess the relative fatigue performance of
different materials as part of an overall test programme which includes the influence of long-term ageing. For
life prediction of components, tests should be carried out on full size or scaled models where possible. More
importantly, testing should be carried out in the operating environment.
TEST METHODS
Smithers Rapra offer three main approaches to fatigue testing:
 Generation of data using standard tests

 Non-standard tests for the generation of fatigue curves of stress or strain versus number of cycles to
failure (crack nucleation).
 Fatigue testing on products

Historically, crack growth tests have been carried out but these are not favoured due to the practical
difficulties with generating a valid test result; the crack growth often forks or deviates towards the grips
making the test invalid. However, Smithers Rapra still has the capability to undertake these tests if required.
Standard Methods
Standard Methods for Rubber
The following tables summarise the common standard methods published by the BS, ISO and ASTM
committees.
Standard Title
ISO 132:2011 Rubber, vulcanized or thermoplastic -- Determination of flex cracking and
crack growth (De Mattia)
Rubber, vulcanized or thermoplastic -- Determination of dynamic properties
ISO 4664-1:2011 Part 1: General guidance
Rubber, vulcanized or thermoplastic -- Determination of dynamic properties

ISO 4666-1:2010 Part 1: General guidance
ISO 4666-2:2008 Part 2: Rotary flexometer
ISO 4666-3:2010 Part 3: Compression flexometer (constant-strain type)
ISO 4666-4:2007 Part 4: Constant-stress flexomer
ISO 6943:2011 Rubber, vulcanized -- Determination of tension fatigue
ISO 27727:2008 Rubber, vulcanized -- Measurement of fatigue crack growth rate
ISO 32100:2010 Rubber- or plastics-coated fabrics -- Physical and mechanical tests --
Determination of flex resistance by the flexometer method
ASTM D4482-11 Standard Test Method for Rubber Property—Extension Cycling Fatigue
ASTM D623-07 Standard Test Methods for Rubber Property—Heat Generation and Flexing
Fatigue In Compression
ASTM D430-06(2012) Standard Test Methods for Rubber Deterioration—Dynamic Fatigue
Method A: Scott Flexing Machine
Method B: DeMattia Flexing Machine
Method C: E. I. DuPont de Nemours and Co. Flexing Machine
ASTM D813-07 Standard Test Method for Rubber Deterioration—Crack Growth
Table 1. Standard Rubber Fatigue Methods
In terms of generating a fatigue life curve of strain versus number of cycles to failure, the most flexible
standard method is BS ISO 6943:2011 Rubber, vulcanized, Determination of tension fatigue. The aim of the
test method is to determine the resistance of vulcanized rubbers to fatigue under repeated tensile
deformation, the test piece size and frequency of cycling being such that there is little or no temperature rise.
Under these conditions, failure results from the growth of a crack that ultimately severs the test piece. The
method is restricted to repeated deformations in which the test piece is relaxed to zero strain for part of each
cycle. The method is believed to be suitable for rubbers that have reasonably stable stress-strain properties,
at least after a period of cycling, and do not show undue stress softening or set, or highly viscous behaviour.
The maximum permitted level of set defined in the standard is 10%. Two test piece geometries are
described; a dumbbell and a ring test piece. The test equipment for the ring test piece is shown in Figure 3.

Figure 3. Wallace/MRPRA test machine
Standard Test Methods for Plastics
The published standards most commonly used for fatigue testing of plastics is given below; the methods for
plastics are significantly fewer in number than those for rubber materials and this may be a reflection that
plastic materials have not been in existence for as long as rubber materials.
Standard Title
ASTM D7791-12 Standard Test Method for Uniaxial Fatigue Properties of Plastics
ISO 15850:2004 Plastics -- Determination of tension-tension fatigue crack propagation --
Linear elastic fracture mechanics (LEFM) approach
ISO 13003:2003 Fibre-reinforced plastics -- Determination of fatigue properties under
cyclic loading conditions
Table 2. Standard Plastics Fatigue Methods
For fatigue crack growth tests in plastics, the compact tension specimen is most commonly used (ISO
15850:2004). The specimens have a machined slot which is then notched with a razor blade (Figure 4). Most
academic studies with this kind of arrangement concentrate on transparent, brittle materials such as
polystyrene and poly-methyl methacrylate, where it is relatively easy to observe crack propagation and linear
elastic behaviour. In semi-crystalline engineering polymers, this is much more difficult, especially with fibre
reinforced grades.

Figure 4. Compact tension specimen for fatigue crack propagation in plastics.
Comments
For most standards, there is a comment in the scope that no exact correlation between the test results and
service life is given or implied. Another limitation is that all the above test methods are carried out under
standard laboratory conditions of 23°C/50% RH. If these conditions do not match the operating temperature
of the application, then the fatigue data cannot really be used for lifetime predictions and can only provide
comparative evaluations.
Non-Standard Methods
When a lifetime prediction is required for a material, bespoke tests need to be carried out which combine
fatigue cycling with temperature and the environment. Depending on the complexity and duration of the test,
it may be more cost effective the build a product test rig. This is particularly required for fatigue tests in fluid
environments. Smithers Rapra has a modular approach to rig building; existing load cells, linear actuators,
conditioning chambers etc. are used to minimise setup costs.
Where materials fatigue data is required, the equipment listed in the following sections is routinely used due
to the flexibility of test parameters.
Servo-Hydraulic
Smithers Rapra’s servo-hydraulic test machine allows specific conditions to be set and on-going
measurement of stress-strain properties is performed during testing. The main disadvantage is the high
running cost of the machine and it is therefore only suitable for short-term tests. A summary of the test
variables are:
Speed: 0 to 1 metre/second
Amplitude: ± 50 mm
Force: Up to 12 kN
Temperature: -50 to +160⁰C
Humidity: Approx 25% to 80%
Electro-Pneumatic
Most of Smithers Rapra’s fatigue tests on plastic materials are carried out using electro-pneumatic test
machines. This methodology has been also applied for some thermoplastic elastomers. The Smithers Rapra
standard tensile fatigue specimens are detailed in the Figure 5 below.

Figure 5. Typical tensile specimen dimensions
The test approach is flexible in that tests can be carried out under fixed stress or fixed displacement, tension
(
Figure 6) or flexure (3 and 4 point bending (
Figure 7)) and using test pieces either with or without a notch. The test fixture can be placed in an oven or
immersed in a test fluid to allow the effects of chemical degradation or environmental stress cracking (ESC)
to be assessed (
Figure 8). The effects of ESC can only really be determined by carrying out long-term, multi-point testing,
since the onset of crack growth in the environment is time, temperature and stress dependent (Figure 9). It is
also worth noting that with ESC, there is no chemical change to the polymer structure, so simple ageing tests
may not highlight the effect.
Figure 6. Electro-pneumatic test rigs showing specimen under test and a ruptured specimen where
the cycles to failure are automatically recorded.

Figure 7. Flexural fatigue using either four or three point loading. The rig is operated using the same
electro-pneumatic system.
Figure 8. Fatigue testing in warm fluid environment.
The typical method of testing is with load control and a square wave cycle (sine wave or other frequency
patterns are also possible). Test frequency is commonly 0.5 - 1.0 Hz, but other frequencies are possible.
With load controlled cycling, the test compensates for any change in specimen dimensions during the test. It
is possible to use displacement transducers to determine the degree of permanent stretch in the material.
With this test arrangement it is relatively straightforward to plot stress or strain versus number of cycles to
failure (S-N curves) with multiple data points at each test temperature. Elevated test temperatures are
commonly used as time temperature superposition is applied to extrapolate fatigue curves to much longer
time periods.
Tests can be carried out between -30°C and +160⁰C in air. However, there are practical limitations
associated with testing in fluids at higher temperatures, especially with the open bath arrangement.
Figure 9. Very different fatigue curves are seen for a number of polymers in a chemical environment
due to ESC effects. Some show dramatic change in behaviour after a moderate number of cycles.
Product Tests
An alternative to carrying out material tests on standard test specimens is to perform cyclic tests on the
products or components themselves. This may be impractical in some cases due to the size and
configuration of the part. The number of cycles to failure should be determined for a number of different load
levels, with temperature used to accelerate testing. This type of testing is advantageous when complex
loading is applied. For example, Figure 10, below shows a section of pressurised pipe being subjected to
cyclic flexing. The pipe assembly is maintained at constant pressure and a bending load applied using an
actuator to supply a cyclic lateral load or strain between two points.

Figure 10. Fatigue bending of a pressurised pipe. The number of cycles to failure are recorded for
different amplitudes of lateral displacement.
CONCLUDING REMARKS
The most useful fatigue test data, whether measured on standard material test or by product testing, is multi-
point S-N data as this can be used to construct long term predictions of behaviour, whereas fatigue testing,
where samples either pass or fail at specified number of cycles, can only be used for the most basic
screening or qualification.
With increasing demands for increasing performance of plastics and elastomers in the oil and gas industries,
where temperatures are increasing and chemical environments are becoming more aggressive, it is
desirable to be able to produce long term predictive fatigue data that is more representative of these
operating conditions. At present it is difficult to achieve this with the current arrangements of the test
equipment. It is envisaged that new test methods will need to be developed to allow fatigue performance to
be evaluated in a hot, pressurised chemical environment. This is an area where Smithers Rapra is currently
looking to develop test methods.

REFERENCES
1. “Life prediction of polymers for the oil and gas industry” A Hulme & J Cooper, Smithers Rapra,
Proceedings of High performance elastomers and polymers for oil and gas applications, April 17-18
(2012)
2. “Life prediction of polymers: model validation” A. Hulme S Speake J Cooper, Proceedings of High
performance polymers for oil and gas, April 10-12 (2013)
3. “Predicting the Life of Polymers for Industry”, J Andrasik, Smithers Rapra, Proceedings of Hose
Manufacturers Conference August 27-28, (2013)
4. “A literature survey on fatigue analysis approaches for rubber” by W V Mars and A Fatemi published
in the International Journal of Fatigue 24 (2002) 949-961

DRILLING FLUID INFLUENCE ON ELASTOMERS

Helmut Benning & Marcus Davidson
Baker Hughes
Email: Helmut.Benning@bakerhughes.com and Marcus.Davidson@bakerhughes.com
BIOGRAPHICAL NOTES
Dr. Helmut Benning studied Chemistry at the University of Hanover, earning a

diploma of chemistry, and continued to stay at Hanover University, Institute of Organic
Chemistry, and earned a Dr. rer. nat. He then completed his postdoctoral positions at
the University of Siegen working on polyurethanes and the German Institute of Rubber
Technology, working on elastic membranes for fuel cells and plasma treatment and
coating of metals, polymers and elastomers. He has now worked for Baker Hughes at
stator manufacturing and elastomer development for over 7 years.
Marcus Davidson has a PhD in Material Science from the University of Dundee. He
has 14 years post-doctoral experience in research and development, the last 7 years
being with Baker Hughes. Currently working as research and development group
leader on new product development for the drilling and completion fluids product line.
ABSTRACT
Elastomers may interact with drilling fluids both physically and chemically. Physical swelling is driven by the
solubility of the liquid in the elastomer. Following the observation of early alchemists: “like dissolves like”
swelling is increased if the molecular structures of the elastomer and solvent are similar.i Swelling decreases
with increased filler concentration and network density as well as by providing leachable ingredients like
softeners.ii The latter may lead to negative swelling which is not always welcome.
Once fluid molecules are inside the elastomer reactions may take place. Acids or alkalinity may cause
addition to double bonds, cleavage of ester bridges, elimination reactions and therefore degradation of the
network or embrittlement. If reactions take place while the elastomer is deformed, permanent set will be the
result which is detrimental to sealing.
Drilling fluids may be water or oil-based and often are emulsions containing two or more distinct liquid
phases. Hydrocarbons do not show significant reactivity towards most elastomers. Their solubility and
amount of swelling depends on the amount of C=C-double bonds or aromaticity and chain length. The
reactivity of aqueous fluids generally increases with temperature. Drilling fluid liquid phases may contain
dissolved components including: emulsifiers, hydroxides, salts, corrosion inhibitors, sour gas scavengers and
thixotropy enhancers. Often additives influence pH. Acid solutions may cause brittleness of hydrogenated
nitrile rubber (HNBR) while alkaline solutions deteriorate standard fluorinated rubber (FKM).
This paper will discuss the composition of drilling fluids and examine the ways in which they can impact the
integrity of elastomers.
Introduction
Drilling and completion fluids provide a number of vital functions while drilling and completing a well. The
multiple functions of a drilling fluid are described in some detail elsewhere.iii There are numerous different
types of fluids and they are often categorised by their base fluid. Each type of fluid contains a complex
mixture of soluble and insoluble components to provide a variety of vital functions. The following will
describe the main functions of additives and detail some information on the composition and look at the
effect that component has on elastomer stability.
Page 1 of 10 pages Paper 14 - Benning

Experimental
Standard elastomer samples (S2-dogbones according to DIN 53 504 for tensile tests, 2-mm-discs according
to DIN 53 479 for swelling measurements and 6-mm-discs according to DIN 53 505 for hardness
measurements) are immersed in the fluid of interest, separated by glass beads in a sealed autoclave and
heated as specified. After cooling down the samples are wiped with a tissue and examined instantly. Original
data are derived from samples from the same elastomer batch. The influence of the amount of air filling the
space above the fluid is considered as negligible.
Results and discussion
In water-based muds the base fluid can either be fresh water, seawater, brackish water or a brine. The
reason for choosing one water phase over another is driven by many factors including availability, cost,
inhibitive qualities, formation water compatibility and density. Error! Reference source not found. contains
a list of the most common water-based fluids. A primary driver is the brine chemistry, higher density brines
allow muds to be made with lower solids concentration. This can be beneficial as removing solids increases
the stability of the fluid and also reduces the likelihood of that fluid from plugging screens and other
production systems. Some brines can be incompatible with the reservoir fluids, for example brines with
divalent cations such as calcium or zinc can cause scaling in some reservoirs.
Standard FKM is not the best material to choose if water is present at 200° C. Error! Reference source not
found. shows what a bellow may look like after ageing in deionized water for 200 hours at 200° C. At
elevated temperatures water develops a significant extent of nucleophilicity that leads to an attack of the
standard FKM polymer backbone and disintegrates the elastomer network. This is observed at 175° C and
very pronounced at 200° C. At this temperature standard HNBR is attacked slowly by pure water, but the
reaction rate increases rapidly after an induction period as outlined in Error! Reference source not found.
and Error! Reference source not found.. After 72 hours volume swelling is below 5% and residual tensile
strength above 90% but after 168 hours swelling is close to 70% and residual strength below 20%. Standard
FKM exhibits a steadily increasing volume swelling from 25% after 72 hours to 60% after 240 hours, while
residual tensile strength is at only 25% after 72 hours and as low as 9% after 240 hours.
Paper 14 - Benning Page 2 of 10 pages

Base resistant FKM keeps swelling below 10% and residual strength above 90%. An improved HNBR
compound behaves even slightly better.
Error! Reference source not found. shows swelling and hardness changes of base-resistant FKM,
improved HNBR and FFKM for up to 200 hours in water at 200° C. Base resistant FKM is swelling by slightly
more than 10 percent while hardness is reduced simultaneously. Improved HNBR and FFKM take up around
5% by volume while hardness rises by around 3 percent.
Figure 4: Swelling at 200 °C in pure water
Nonaqueous fluids also have a number of different types of base fluid. In the early days of drilling the natural
crude oil would be used. This progressed to using diesel oils, mineral oils, lower toxicity mineral oils and
more recently synthetic oils. Synthetic oils include linear paraffin, internal olefins, linear alpha olefins, poly
alpha olefin, esters and gas to liquid oils. The properties of some typical non-aqueous fluids are given in
Error! Reference source not found. below.
Hydrocarbons tend to interact with elastomers mainly physically, as chemical reactions are unlikely due to
the nearly inert nature of hydrocarbons. Parts of the oil enter the elastomer network and lead to swelling.
This happens against the elastic recovery of the elastomer. Therefore swelling reaches equilibrium. The
extent of swelling depends on the compatibility of the solvent and the polymer. “Like dissolves like” was
already known by the alchemists. Error! Reference source not found. shows the effect of the aromatic
content of the hydrocarbon mixture on swelling of two nitrile types. As nitrile elastomers are partly polar, high
aromatic content increases the compatibility of elastomer and solvent. At the same time, compounds which
are not fixed to the network like softeners and fragments of crosslinking agents may leach out. Depending on
the elastomer recipe and the swelling, hydrocarbon leaching out may even be more pronounced than
swelling, which leads to a negative volume swelling.

To get an idea of the extent of swelling oil may cause in a polar elastomer, the aniline point may be taken
into account.iv Aniline is used as a molecular model which combines polarity and aromaticity. The lower the
temperature at which miscibility is observed the higher the capability of the oil to swell polar elastomers.
Error! Reference source not found. shows predictions of swelling can be made for different elastomers
based on the aniline point of oil.
Drilling fluids sold with the same brand name may still be different. This is illustrated by Error! Reference
source not found., where swelling of one nitrile elastomer is shown with oil-based mud from different
locations.
The hardness is reduced while the elastomer is swelling.
Error! Reference source not found. shows the devolution of volume swelling, hardness, tensile strength
and ultimate elongation if the aromatic content of oil is increased. Volume swelling increases nearly linearly
while hardness decreases progressively. Tensile strength and ultimate elongation are reduced in the first of
the illustrated examples. The second example shows similar devolution of volume swelling and hardness
with aromatic content but tensile strength and elongation pass through a maximum. The evolution of a
maximum may be explained by starting at negative volume swelling, which means that the elastomer chain
molecules are compressed related to their position during crosslinking. By increasing the aromatic content of
the oil the extraction of mobile compounds is surmounted by oil entering. Tensile strength and ultimate
elongation do not need to show their maxima at the same extent of swelling versus extraction.
Two nitrile elastomers were immersed in a second oil-based mud brand from three different origins. Both
show a wide variety of swelling while the hierarchy of elastomers as well as the different origin oils is
maintained regarding swelling. Hardness does not follow as simple rules in this case as shown in Error!
Reference source not found..
One brand of synthetic-based mud from different locations shows very moderate influence on nitrile II as
listed in Error! Reference source not found.. Extraction and swelling are nearly balanced. Here the
differences from batches of different origin are small. This also applies to further sbm brands.
Error! Reference source not found. shows a comparison of the influence of a single sbm on two nitrile and
one FKM elastomer. The two nitrile types reveal 8.8% swelling and 1.3% shrinkage, while the FKM reaches
3.3% swelling. Hardness change follows swelling.
In comparison to crude oil swelling in Error! Reference source not found. exhibits the swelling of FKM
staying at only 5.3% while nitrile II has passed to 6.6% and nitrile I even 16.3% volume swelling. This shows
that fluorinated elastomers experience only small interactions with non-fluorinated hydrocarbons, while nitrile
elastomers interact with aromatic hydrocarbons. Fluorinated elastomers are compounded without softeners,
so negative volume swelling caused by extraction of softeners is prohibited.

Viscosifier
One of the earliest types of viscosifier used in drilling fluids was natural clays found in the formation being
drilled. These imparted a viscosity to the fluid and it was found that wells could be drilled faster owing to the
improved capacity for carrying rock cuttings. Rather than relying on formation clays to give fluid viscosity an
number of products can be used to impart the desired rheological properties. Natural and processed clays
are still used. There include swelling clays such as montmorilonite of which Wyoming bentonite is probably
the most common. Other clays that can be used are attapulgite, sepiolite and hectorite. Attapulgite and
sepiolite clays are useful in drilling fluids as they are not sensitive to electrolyte levels that would flocculate
bentonite and hectorite suspensions.
Natural polymers including guar gum, xanthan gum, welan gum, diutan gum and scleroglucan. Xanthan gum
is a long-chain anionic polysaccharide with molecular weight in excess of 2 million Daltons. Whelan gum is
similar to xanthan but is more tolerant to higher pH and calcium ion environments. Guar gum is sourced
from the ground endosperm of the guar bean. It is also a polysaccharide made up of the sugars galactose
and mannose. It finds less use in drilling fluids than xanthan and other gums owing to the lower solid
suspending properties.
Synthetic and modified polymers form another family of viscosifiers for water-based muds. These include
hydroxyethyl cellulose (HEC) and carboxymethyl cellulose (CMC). These modified celluloses often used in
their sodium salt form.
In oil-based fluids the viscosifier of choice is organophilic clay. This is a clay treated with a quaternary amine
to render it hydrophobic and dispersible in oil. The amine with a fatty chain is covalently bonded to the clay
platelet. Bentonite or hectorite are common clays to use but attapulgite and sepiolite are also used for
applications where higher suspension characteristics are required. Polymer viscosifiers are also used, most
often in combination with organophilic clay. These can include synthetic copolymers or modified tall oil fatty
acids.
Dispersants
One inevitable fact about drilling fluids is that over time they will become contaminated with solids. The mud
circulating system will include equipment that can control the concentration of unwanted solids a
concentration of fine low-gravity solids will build up over time. Fine, low-density solids contribute to the
viscosity of the mud and if the concentrations are high enough the mud can be excessively thick or prone to
forming strong gels. A consequence of high low density solids is shown in Error! Reference source not
found..

Dispersants can be used in water-based fluids or oil wetting agents are common in oil-based muds. Both
types of chemicals have the same effect of reducing the association between the particles and mud
components to lower the viscosity. Lignosulphonates, a byproduct of the paper industry, tannins or modified
polyacrylates can be used in water-based muds. In oil muds, wetting agents are more commonly surface
active molecules such as modified fatty acids. Phosphate esters and other surfactants are also used.
Emulsifier
Oil muds are in most cases a water-in-oil emulsion. The internal, aqueous phase serves a number of
purposes. It provides additional viscosity, acts like a solid in terms of helping to lower fluid loss and provides
a place where water entering from the fluid can go. The internal phase is most often a brine, most commonly
calcium chloride. The salinity is typically about 25 wt.% and is usually higher than the salinity of the
formation brine so that it provides osmotic control to prevent water from the mud going into the formation
where it may cause swelling of mixed layer clays, etc. To stabilize the emulsion emulsifiers are necessary.
The emulsifier is a surfactant with a hydrophilic lipophilic balance (HLB) of typically between 8 and 10. Being
surface active with a more polar head and a non-polar tail, they will orient themselves at the interface
between the aqueous phase and the oil or at the surface of hydrophilic solids. Modern emulsifiers use two
types of chemistry to emulsify water or brine in oil. Tall oil fatty acids (TOFA) and chemical reaction products
of TOFA to form imidazolines or a complex polyamide mixture.
Tall oil fatty acids are a by-product of the Kraft Paper Process and are a cornerstone of oil-based fluids. To
prepare the TOFA for use as an emulsifier it is first oxidized, which produces more complex acids. The
functional components of TOFA are resin acids and fatty acids (oleic, linoleic and linolenic). When these
acids are added to a fluid containing lime Ca(OH)2 they are converted to calcium salts (soap).
Imidazolines and complex polyamide mixtures were developed. These emulsifiers are referred to as
secondary because they historically played a supporting role in stabilizing oil-based drilling fluids. They are
manufactured from TOFA, polyamine, and a polycarboxylic acid or anhydride. It is possible to make two
different types of emulsifiers by modification of the manufacturing process. One type consists of a complex
mixture is classified as imidazolines.
Fluid Loss Additive
When drilling through a permeable rock formation with a pressure overbalance, it is inevitable that fluid loss
will occur. Large solids in the drilling fluid help bridge across the pore openings but smaller particles, resins
and polymers fill the gaps and ensure that the fluid loss is low. Excessive fluid loss is undesirable as it can
destabilize the formation and lead to the collapse of the wellbore and associated problems such as stuck
pipe or loss of bottomhole assembly.

In oil-based drilling fluids the fluid loss additive can be selected from amine treated lignite, resins, synthetic
copolymers such as those based on acrylate/styrene, gilsonite, modified tall oil. In water-based systems it is
common to use polyanionic cellulose, CMC, starch, lignite, asphalt and sulphonated asphalt, gilsonite,
modified lignites and synthetic copolymers.
Inhibitors
Owing to the high cost of oil, water-based muds tend to be cheaper and easier to dispose of. There are
some applications where oil-based muds are preferred either due to their lower coefficient of friction or
inhibitive qualities. Shale is a major problem when drilling and especially so with water-based drilling fluids.
In the past couple of decades a water-based mud has been developed with similar inhibitive qualities to oil-
based mud while retaining the benefits of a low-cost base fluid and relative ease of disposal. These mud
systems, commonly referred to as high-performance water-based muds contain a raft of inhibitive products
designed to work together to prevent shale from swelling. The base fluid is typically a potassium chloride
brine as potassium is well known for its inhibitive qualities. Further shale inhibition is achieved by the
inclusion of some form of amine salt. The nitrogen groups on the amine are very effective at pairing with the
negative charges found on the clay platelets. Aluminium complexes and silicates are particularly effective at
reducing fluid loss into shale formations. At high pH (over 10.5) they are essentially soluble. When filtrate
containing the aluminium complex enters the formation and mixes with the connate water the pH drops and
the aluminium complex is no longer soluble. Precipitation occurs and the pore is effectively sealed.
Partially hydrated polyacrylamide is a cationic polymer that effectively bonds to the negative sites on the clay
surface encapsulating it and preventing water penetrating into the clay and causing swelling.
pH Control
It is desirable to control the pH of a water-based drilling fluid to typically between 9 and 10 but some fluid
systems such as silicate or aluminium complex fluids to over 11. The high pH in silicates and aluminium
complex fluids maintains the solubility of the metal complex and allows it to be inhibitive. A pH higher than 8
or 9 is also useful for minimizing the corrosive effects on down hole tubulars especially as temperature may
be elevated. The common products for modifying and maintaining drilling fluid pH are shown in Error!
Reference source not found.. In water-based mud the most commonly used is caustic soda and in oil-
based muds, lime is the pH modifier of choice. Some pH control additives are used to treat contaminations
of the mud, soda ash can be used to treat calcium contamination found when drilling cement for example.

Alkaline solutions attack standard FKM. Amines, sometimes used for corrosion protection, as well as oxides
and hydroxides mentioned in Error! Reference source not found. contribute to alkalinity. Basicity of
amines added to oil has an influence on the reactivity towards FKM.v This leads to elimination of hydrofluoric
acid, formation of a double bond in the polymer backbone and subsequently degradation of the elastomer
network. At pH= 10.5 peroxide-cured standard FKM is still applicable but above it will start being
deteriorated. Temperature and required service life will move this to higher or lower values. Therefore
applications in drilling services with some hundred hours of service life will be treated different to completions
where several years are required.
Alternatives to standard FKM are base resistant FKM, FFKM and HNBR.
Very high acidity, pH values as low as 2, are detrimental to nitrile rubber. Error! Reference source not
found. shows that even after 25 hours at 135 °C a lot of damage is done. While volume swelling and change
of tensile strength are still below 10 % hardness and elongation have changed +25 and –27 % respectively.
After 72 hours hardness has increased by more than 40% and tensile data cannot be determined because
samples break before the tensile test machine has reached initial stress. At higher temperatures the
outcome stays the same.
Weighting Material
One critical function of drilling fluids is to maintain downhole pressures. To do this the density of the drilling
fluid must equal or exceed the pore pressure of the formation. Brine can provide some density but it is often
more convenient or lower cost to use a high-density solid mineral. The type of mineral is usually chosen for
its insolubility in the base fluid, the hardness, abrasiveness, availability, cost and so on. Suitable materials
are listed in Error! Reference source not found..
By far the most common mineral used to add weight to drilling fluids is barite. Barite for drilling fluids is
typically a 200-mesh product with specific gravity of at least 4.2 (i.e., >90% BaSO4). Water-soluble alkaline
earth metals are controlled so as not to interfere with drilling fluid rheology [ref Industrial minerals]. Calcium
carbonate is used as either calcite or aragonite from natural limestone, marble and chalk sources. Aragonite
is a modified calcite. Calcium carbonates are used in fluids for drilling the reservoir where a high degree of
acid solubility can ensure no reduction in permeability that can impede oil or gas production or water
injection.

Addition of glycol derivatives to water-based drilling fluids results in increased swelling, which is illustrated in
Error! Reference source not found.. As expected, hardness is reduced simultaneously. The extent of
swelling depends on the percentage added and on the elastomer compound.
Furthermore, the rubber-to-metal bond may be weakened if glycol derivatives or other polar fluids like esters
migrate to the rubber to metal bond. vi
References
i
Bruno Vollmert: Grundriss der makromolekularen Chemie, Band IV, 115 ff.
ii
R. Hornig, K. Athanasopulu: GAK 3/2011, p 165 – 175 and GAK 4/2011, p 234 - 240
iii
Composition and Properties of Drilling and Completion Fluids (Sixth Edition), 2011, Pages 1-37,
Ryen Caenn, H.C.H. Darley, George R. Gray
iv
ASTM D611
v
C. Bergmann, J. Trimbach, W. Schmid: RFP - Rubber Fibres Plastics International 2013 1, 23ff
vi
J.R. Halladay, P.A. Warren: Rubber to Metal Bonding, in Handbook of Rubber Bonding, Rapra
Technology Limitied, 2001, p 66.

MODELING AND DESIGN OF REINFORCED ELASTOMERIC

PRODUCTS FOR OIL AND GAS APPLICATIONS
Stuart B. Brown, Jorgen Bergstrom, and Nagi Elabbasi
Veryst Engineering, LLC
47A Kearney Road, Needham, MA 02494 U.S.A.
Tel: +1 781 433 0433 Fax: +1 781 433 0933 email: sbrown@veryst.com
BIOGRAPHICAL NOTE
Dr. Stuart Brown, Managing Principal, Veryst Engineering, LLC. Dr. Brown is a
principal in Veryst Engineering, a Boston-based engineering firm specializing in
the analysis and testing of highly nonlinear materials and products using those
materials. Prior to Veryst Engineering, Dr. Brown was director of the Boston
office of Exponent, Inc., and before that he was on the faculty of the Department
of Materials Science and Engineering at the Massachusetts Institute of
Technology.
ABSTRACT
Reinforced elastomeric products are used commonly in oil and gas applications, including flexible hoses and
seals. Design of these products using engineering methods is difficult given the combination of rate-and-
temperature-dependent elastomers, highly nonlinear reinforcing materials such as Kevlar and other woven or
oriented materials, the potential for large deformations, and the high degree of anisotropy in the combined
product. Fundamental analyses are also complicated by the need to measure nonlinear material properties
and then calibrate the resulting test data to advanced material models. This presentation describes rigorous
methods to design these products using a combination of specialized material testing, implementation of
sophisticated models for elastomer behavior and reinforcements, and nonlinear finite element modelling. We
provide two examples of design analyses implementing these methods: a Kevlar and nylon cord reinforced
elastomeric hose (an example pictured below), and a gasket subjected to high temperatures and pressures.
INTRODUCTION
This project presents two case studies of modelling used for the design of reinforced polymer products used
in the oil and gas industry. The first case study simulates the deformation of a Kevlar and Nylon cord
reinforced, elastomeric mooring hose. The second case study models the deformation of a glass reinforced,
Teflon gasket used to seal a drill pipe connection.
CASE 1: CORD REINFORCED ELASTOMER HOSE
The hose is used as a stretchable portion of a mooring line connecting a surface-following buoy to the ocean
floor. An image of one such mooring hose is provided in Figure 1. The mooring lines are designed to stretch
several hundred percent and are typically deployed with an initial tension in the hose to provide a more
stable buoy position. (Irish, 2005; Paul, 2004, Paul, 2005)
Page 1 of 8 pages Paper 15 - Brown

Figure 1: Image of reinforced, elastomeric mooring hose
The hoses have a complicated structure. Described from the center, moving radially outward, they have an
inner, liquid-filled center, surrounded by an elastomer inner layer. Next are multiple layers of Nylon tire cord
oriented at an angle designed to be initially in compression as the hose elongates, transitioning to tension as
the hose lengthens. This transition from compression to tension develops from the rotation of the orientation
of the cords from a circumferential to a more longitudinal direction as the hose lengthens. Continuing radially
outward are more elastomers layers, which frequently contain embedded wires providing both power and
signal transmission to sensors located lower in the mooring system. The next layers are Kevlar cord for
protection from impact and shark bites. The final layer is a woven nylon fabric embedded in elastomer for
abrasion resistance. Figure 2 provides an overall view of the hose geometry.
The combination of the multiple layers, all with different material nonlinearities and anisotropy, made design
of these hoses difficult. Veryst Engineering was hired to produce a finite element model of the hose to
include explicitly more material nonlinearities than was possible using the design tools previously used to
construct these hoses.
Elastomer
Conductor
Kevlar Cord
Nylon Cord
Elastomer
Figure 2: Overview of reinforced hose cross section
The ends of the hose were not modelled, as the end constructions are complicated by metal end flanges.
The elastomer layers were modelled with an isotropic Bergstrom-Boyce material model with Mullins effect to
account for the change in response after the first few loading cycles (Bergstrom, 2001). Veryst measured
the stress/strain behavior of the corded materials both in multiple in-plane directions, at different strain rates,
as well as in compression and tension. As expected, the materials were highly anisotropic and, given the
elastomer, also rate-dependent. Figure 3 and Figure 4 provide example behavior for the Nylon reinforced
Paper 15 - Brown Page 2 of 8 pages
layers. Figure 3 provides the stress/strain response in the cord direction at two strain rates. Figure 4
provides similar data perpendicular to the cords. Notice the very large difference in mechanical response
given the fiber directions.
Figure 3: Nylon cord layer response in cord direction Figure 4: Nylon cord layer response perpendicular
to cord
The actual hose geometry was modelled both as a 3D structure and as a 2D axisymmetric structure
employing the Abaqus feature of “axisymmetric with twist” which permits analysis of axisymmetric structures
that can rotate about the axis of symmetry (Abaqus Analysis Users Manual). Figure 5 provides a strain
history applied to an actual hose.
Figure 5: Strain history
Figure 6 demonstrates the ability of the model to simulate the hose response under the loading history
shown in Figure 5.

Figure 6: Comparison of experimental and predicted load history resulting from strain history of Figure 5.
It is possible to capture the highly anisotropic and nonlinear behavior reinforced hoses using appropriate
material models. The axisymmetric with twist option within Abaqus provides a computationally efficient
means of capturing certain material anisotropies without requiring the complexity of a full three dimensional
model. The two dimensional idealization however does not permit important deformation modes, such as
buckling, ovalization of the cross section, or dynamic analysis. Ultimately a full three dimensional model is
needed to accommodate all potential deformation modes within the hose. Despite these limitations, we were
able to model the deformation behavior of the reinforced hose with an accuracy considered not possible by
our client.
CASE 2: GLASS FILLED TEFLON GASKET
This second case study involves the analysis of a glass filled Teflon gasket used as a seal in a drill pipe
application, shown in Figure 7. The gasket provides backup sealing for the threaded connection. The
gasket is reinforced with short glass fibers and although reinforced has much less anisotropy compared to
the reinforced hose described above.
Teflon Gasket
Figure 7: Schematic of glass-filled teflon gasket position within fitting
Figure 8 provides stress-strain data for the Teflon in both tension (at multiple temperatures) and
compression. This data, along with data from relaxation and constrained compression testing, was used to
calibrate a material model for fluoropolymer materials (Bergstrom, 2005). Figure 9 provides the fit of the
material model to the test data shown in Figure 8.
If possible, a material model should be validated under conditions that differ from those used to determine
the associated material model parameters. In this case, Veryst used punch testing to validate the material
model for this gasket material. A finite element model of the punch test is shown on the right hand side of
Figure 10. The punch test is convenient, as it can be modelled as an axisymmetric geometry. The punch
test experiment used the same glass-filled Teflon and simulated the test using the material model described
in Bergstrom 2005, and illustrated in Figure 9.
Figure 8: Tension and Compression Behavior of Gasket at Temperature
Figure 9: Material Model Comparison to Experimental Data
A representative test specimen is shown in Figure 10, and the experimental punch force/displacement
response is compared to the predicted response on the left side of the figure.

Figure 10: Validation of Model Using Punch Test
Figure 11 provides the compressed geometry of the gasket with associated von Mises stress field for short
time compression. The stresses will relax over time, and the stresses will also change with temperature.
Figure 11: Finite Element Stresses Predicted for Model
CONCLUSIONS
It is possible to simulate and design reinforced polymeric products using advanced finite element methods.
The primary constraint in performing these analyses is appropriate modelling of the highly nonlinear
materials comprising the different components. Nonlinearities include the natural rate-and-temperature
dependence of the polymers as well as a high degree of anisotropy introduced by directional reinforcement
such an embedded woven cords.
This capability provides enormous design benefits. Reinforced hoses can be very difficult to manufacture,
and design tools offer significant cost savings by reducing the number of design iterations required to
achieve the desired mechanical response. In addition, the addition of rate-and-temperature dependence
allows prediction of the performance of hoses and gaskets under a variety of operating conditions and over
prolonged periods of use.

References
1. Abaqus Analysis User’s Manual, “Generalized axisymmetric stress/displacement elements with

twist”.
2. Bergstrom, J., and Boyce, M, “Constitutive Modeling of the Time-dependent and Cyclic Loading of
Elastomers and Application to Soft Biological Tissues,” Mechanics of Materials, Vol. 33, pp. 523-530,
2001.
3. Irish, J, Paul, W, and Wyman, D, “The Determination of the Elastic Modulus of Rubber Mooring
Tethers and their use in Coastal Moorings”, Woods Hole Oceanographic Institution, WHOI-2005-10,
2005
4. Paul, W, Chaffey, M, Hamilton, A, and Boduch, S, “The Use of Snubbers as Strain Limiters in Ocean
Moorings”, Oceans 2005, Proceedings of MTS/IEEE.
5. Paul, W., “Hose Elements for Buoy Moorings: Design, Fabrication and Mechanical Properties”
Woods Hole Oceanographic Institution, WHOI-2004-06, 2004
6. PolyUMod Material Subroutine Library, Veryst Engineering, LLC.
7. Bergstrom, J., and Hilbert, L, “A Constitutive Model for Predicting the Large Deformation Behavior of
Fluoropolymers,” Mechanics of Materials, Vol 37, pp. 899-913, 2005.


DEVELOPMENT OF NEXT GENERATION OF MATERIALS FOR

SEALING SOLUTIONS IN HPHT CONDITIONS
Mathilde Leboeuf, Chien Nguyen, Filip Rousseau, Christophe Valdenaire, Rojendra Singh
Saint Gobain Seals Group
Kontich, Belgium
Email: Mathilde.Leboeuf@saint-gobain.com
BIOGRAPHICAL NOTE
Mathilde Leboeuf has a PhD on Material Science from the Ecole Des Mines de Paris,
France. She has been working for St Gobain since 2008 with a main focus on the
development of next generation materials and sealing solutions for the Oil and gas industry
ABSTRACT
UNAVAILABLE AT TIME OF PRINT
The oil and gas industry today is facing applications in high pressure and high temperature (HPHT)
conditions (Figure 1). Sealing solutions have been rated as the technology with the largest gap to overcome
for such applications (Figure 2). Therefore, there is an emphasis on the research and development of new
materials that can withstand demanding conditions. As a response to such needs, Saint-Gobain Seals Group
has been developing next generation of materials that can be used to fabricate next generation of sealing
solutions. The performance of the materials are confirmed by experimental work as well as proprietary FEA
material model.
Figure 1: High temperature high pressure reservoirs

(courtesy from Schlumberger)
Elastomers have shown evidence of reaching their limit in terms of HPHT resistance. Thermoplastic sealing
solutions offer a strong alternative because of their thermal and chemical resistant properties as well as their
rapid gas decompression resistance.
Saint-Gobain has developed specific thermoplastic formulations suitable for spring energised seal and back
up ring to withstand the high pressure and high temperature conditions and are commercially available as
Fluoroloy materials.
Page 1 of 4 pages Paper 16 - Leboeuf

Figure 2: HPHT technology gap (courtesy from Shradavan 2012 [1])
Materials for seal jacket have been exposed to isostatic pressure of 30,000 psi at room temperature and
30,000 psi at 250°C. Comparing the material properties before and after the exposure reveal no significant
based on the tensile properties. (Figure 3)
Figure 3: Tensile strength of materials before exposure, after exposure at 30 kspi and room
temperature, after exposure at 30 kspi and 250°C
A test rig to expose back up ring materials at HPHT conditions in configuration similar to those of application
was developed. It consists on an extrusion test rig that reproduces HPHT conditions (Figure 4).
Figure 4: Extrusion test rig that replicates application configuration
Paper 16 - Leboeuf Page 2 of 4 pages

Tests have been conducted on 16 materials. For each material, the height of extrusion is measured after
exposure to 25 kpsi at 200°C for 4 hours. The results are presented in Figure 5. In this test conditions, the
reference material (material 1) shows an extrusion of 0.038 inches. Among the fifteen other materials
evaluated, thirteen showed lower extrusion than the reference material and two had a spontaneous failure.
Among the materials tested, samples #4 and #7 show the lowest rate of extrusion.
Figure 5: Extrusion performance of 16 materials that include reference materials and 15 new
candidates for Fluoroloy materials for back up ring in HPHT application
The selected materials for jacket and back up rings are also being exposed to H2S concentration according
to NORSOK M710 rev2 specification. The experiments conducted in an external lab show that the materials
are compatible with H2S and passed the specification as prescribed by NORSOK M710 rev2 .
Representative results from the ageing test are presented in Figure 6 and Figure 7.
Figure 6: Representative examples of Volume swell performance of materials evaluated for HPHT
applications according to NOROSK M710 rev2 test specification
Page 3 of 4 pages Paper 16 - Leboeuf

Figure 7: Representative examples of change in elongation at break of materials evaluated for HPHT
applications according to NOROSK M710 rev2 test specification
Finally, as an outcome of this project, the thermoplastics materials that show promise at HPHT conditions
and chemical compatibility are being selected for sealing solutions in demanding applications. To validate
the performance of such sealing solution, a material model has been developed that captures thermo-
mechanical behaviour of thermoplastics.
REFERENCE
1- Shradavan 2012 HPHT 101 – What every engineer or Geoscientist should know about high pressure
high temperature wells, SPE 163376
Paper 16 - Leboeuf Page 4 of 4 pages

HNBR IN CO2
D L Hertz III
Seals Eastern Inc.
Red Bank, NJ, 07701, USA
Email: dhertziii@sealseastern.com
BIOGRAPHICAL NOTE
Mr Daniel L Hertz III has been involved with the elastomer seal business for over thirty years.
He is a principal of Seals Eastern Inc and has experience in all aspects of the seal business
from design and manufacture to his present position in executive management. He currently
serves as President of Seals Eastern, an American manufacturer committed to serving the
sealing needs of Fortune 500 companies across the globe for more than fifty years.
Mr. Hertz earned his bachelor’s degree from the University of Colorado, a Master of Science from Stevens Institute of
Technology, and a Juris Doctorate from Brooklyn Law School.
Mr. Hertz has lectured and presented numerous technical papers on the topic of high performance elastomers and
seals, with topics including fluorelastomers, polymer basics, elastomer behavior in brines, steam, extreme heat, cold
temperature, alternative methods of evaluating elastomers and other topics involving the seal industry and specialty
elastomers.
ABSTRACT
Nitrile rubber (“NBR”) remains one of the most popular oilfield elastomers. More recently, hydrogenated
nitrile rubber (“HNBR”) is being accepted in its place on account of its similar “toughness” and improved
stability in the presence of heat and reactive chemical species. However, carbon dioxide (a naturally
occurring gas that is frequently encountered in hydrocarbon environments) presents challenges for the NBR
class of polymers. Relatively small concentrations of CO2 in hydrocarbon mixtures can cause significant seal
swelling if consideration is not given to the specific choice of polymer, cure, and elastomer reinforcement.
More significantly, the effect of absorbed CO2 upon rapid gas decompression can be catastrophic if the same
consideration is not applied.
This study explores the interaction of CO2 and HNBR polymers. The relationships between CO2 and
acrylonitrile level is examined. This study is a continuation of work conducted and presented at RAPRA’s
2012 High Performance and Specialty Elastomer symposia but with an exclusive focus on HNBR. The
study’s objective is to provide reference data for both the application engineer and compounder when
designing for applications where CO2 will be encountered.
INTRODUCTION
Carbon Dioxide (CO2) is a naturally occurring colorless, odorless gas. It is frequently found in hydrocarbon
reserves. CO2, in the gaseous state, is denser than air with a specific gravity of 1.98 kg/m3.
CO2 is a linear molecule of two oxygen atoms bonded to one carbon atom through double bonds (C=O=C).
The molecule is symmetrical around the carbon atom and thus has no dipole moment. However, CO2 being
a linear triatomic molecule possesses four bending modes. The molecule presents symmetrical and
unsymmetrical stretch modes. The third and fourth bending modes include bending in the plane of page or
perpendicular to it (“doubly degenerate”). Given the CO2’s transient dipole moments, the molecule appears
benti (e.g. like an H2O molecule). Thus, the simple rule of thumb of “likes dissolves likes” is misleading if you
consider CO2 as a linear molecule.
Carbon dioxide becomes a supercritical fluid and hence a solvent at relatively modest pressures and
temperatures. The requisite parameters frequently exist in the reservoir and production conditions. Carbon
dioxide is only able to exist in the liquid state at pressures above 0.517 MPa (74.9 PSI). The triple pointii of
CO2 is about .518 MPa (75.1 PSI) at -56.6˚C. The critical pointiii is 7.375 MPa (1070.4 PSI) at 31.1˚C (88˚F).iv
In the course of this study, super-critical conditions were not present.
Page 1 of 12 pages Paper 17 - Hertz

The solvating powers of CO2 are well documented and applications utilizing supercritical CO2 have been
established for some time now. Unfortunately for the oil & gas field operator, these very same principles are
at work sabotaging elastomeric seals and the equipment they are designed to serve when CO2 is present in
the hydrocarbon stream. Modest amounts of CO2 present in the hydrocarbon reservoir can induce failure in
elastomeric seals that otherwise perform admirably in high pressure gases. Usually, the damage occurs
during rapid gas depressurization (“RGD”).
This study was conducted using 5 MPa (750 PSI) of pure CO2 which could be considered moderate pressure
in terms of most field conditions. However, the implications of Dalton’s “Law of Partial Pressures” should be
considered when viewing this data. Specifically, Dalton postulated that the total pressure of a mixture of
gases is just the sum of the pressures that each gas would exert if it were present alone and occupied the
same volume as the mixture of gases. Under most conditions, the molar fraction of CO2 in a hydrocarbon
gas mixture is substantially smaller than the molar fraction of the other gases present (e.g. N2, He, O2, CH4,
C2H6, C3H8, etc.). Thus, in the context of partial pressure, the CO2 condition in this study would exist in well
pressures of several thousand PSI where the CO2 molar fraction is only a few percentage points. On the
other hand, in a situation such as CO2 reinjection, field results might differ substantially from those observed
herein. . For a more critical discussion of the theoretical dynamics and associated references, the author
directs you to the published article “Elastomers in the Hot Sour Gas Environment” by Hertz, Jr.v
This study was undertaken to document HNBR’s interaction with CO2. HNBR is a copolymer of acrylonitrile
(“ACN”) and Butadiene. Unlike NBR, the copolymer is subsequently hydrogenated to increase saturation of
the butadiene component. NBR and HNBR are primarily graded by their acrylonitrile content. By varying the
ratio of ACN and butadiene, different properties are obtained. How this ratio affects HNBR behavior in CO2
was the question addressed by this study.
OBJECTIVES
The first objective of this study was to offer a comparative analysis of various HNBR grade’s swelling in
pressurized CO2 and swelling subsequent to rapid gas decompression (“RGD”). The enclosed data might
then serve as a quick reference for determining possible swelling of HNBR compounds in reservoirs known
to contain CO2.
The second objective was to offer details that could mitigate/exacerbate the swelling of HNBR compounds
subject to CO2 either while under pressure or subsequent to RGD. Specifically, this study examined
differences attributable to the amount of acrylonitrile, the amount of curative, grades of fine particle
reinforcement, and the amount of fine particle black.
SCOPE
CONTROLLED FACTORS:
Elastic modulus is a primary consideration of seal design. It is also one attribute affecting an elastomer’s
behavior under pressure and during RGD. However, there are several factors that will ultimately define
elastic modulus as well as other material attributes. An experimental array would be unwieldy if all these
factors and their possible levels were all examined. For purposes of this experiment, the author chose only
the most fundamental factors used to develop elastic modulus and solubility behavior of an HNBR oilfield
compound. A Taguchi L9 Orthogonal Arrays was used to study the factors and their associated levels.
Specifically, the controlled factors were:
1) The Acrylonitrile content;

2) The degree of cross-linking as controlled by part-per-hundred (“phr”) of curative;
3) The particle size/structure of carbon black, controlled by grade of carbon black, specifically N990,
N762, and N330;
4) The loading of carbon black reinforcement, controlled by phr of carbon black.
ENVIRONMENT:
Gas composition and testing temperature, while constant, were treated as uncontrolled factors in the
experiment. A pressure vessel, with a built in observation window, per Figure 1B, was flushed and charged
with a connected canister of 99.9% pure CO2 at room temperature 22.7˚C (73˚F) to evaluate the specimens
placed within it. The configuration is schematically detailed in Figure 1A.
Paper 17 - Hertz Page 2 of 12 pages

Figure 1A – Test Fixture Configuration Figure 1B – Observation Vessel and test vials
ELASTOMERS:
For the rubber compounder and application engineer, HNBR polymer is normally graded on the following
attributes:
1) Percentage content of acrylonitrile,

2) Degree of saturated butadiene in the backbone,
3) Mooney viscosity.
This is not an exhaustive list of attributes, but those most indicative of the materials behavior. This study
focused primarily on the acrylonitrile content while attempting to hold the other attributes constant. Because
the acrylonitrile content primarily defines the molecular composition of HNBR and its resistance to non-polar
(e.g. hydrocarbons) and polar (e.g. water) substances, it was the primary focus of this study.
All of the HNBR compounds herein were mixed on an open 12-inch roll mill.
Median
Percent Median
Acrylonitrile Median Percent Mooney
HNBR code (%) Saturation (%) Viscosity
1010 44% 96% 85
2010 36% 96% 85
3310 25% 95% 80
4310 17% 95% 72

Table 1 – Elastomer Test Groups and Specimens
TEST SPECIMENS:
Specimens conforming to those defined by ASTM D1460-86 (2010) Section 7.1 were utilized. The
specimens were die cut from ASTM slabs and measured 100 mm (4.0 in.) in length by ~1.6 mm (0.063 in.)
wide by ~ 2.0 mm (0.075 in.) thick. By so doing, the author could make reliance upon Table 1 of ASTM
D1460-86 (2010) for approximating the percentage change in volumevi.

EXPERIMENTAL
METHODOLOGY:
In Experiment 1, the acrylonitrile content was isolated while holding all other variables constant to the extent
this was possible. Four experimental compounds were mixed, tested, and measured as described infra.
In Experiment 2, specimen formulas were designed using Orthogonal Arrays, per Taguchi, and are detailed
infra. Orthogonal arrays are tables of numbers that allow for effective combinations of factors and levels for
an experiment. This approach allowed the study of a small fraction of the possible combinations of factors
(elastomer ingredients) and levels (ingredient loadings) to yield unbiased and meaningful results. Table 2
illustrates the L9 matrix used to test four (4) factors at three (3) levels.
LEVELS
FACTORS 1 2 3
A POLYMER A1 A2 A3
B CURE PHR B1 B2 B3
FILLER
C TYPE C1 C2 C3
D FILLER PHR D1 D2 D3
Table 2 - Taguchi L9 design of experiment
The conditions (i.e. compound formulas), numbered 1 through 9, contain no unfair biasing when Orthogonal
Arrays are utilized. Table 3 illustrates the resulting conditions utilizing a Taguchi L9 Orthogonal Array. The
measured result is the percent volume change during exposure to pressurized CO2 and subsequent to RGD.
Result A1 A2 A3 B1 B2 B3 C1 C2 C3 D1 D2 D3
#1 a a a a a
#2 b b b b b
#3 c c c c c
#4 d d d d d
#5 e e e e e
#6 f f f f f
#7 g g g g g
Condition
#8 h h h h h
#9 i i i i i
Total #1:9 #1:9 #1:9 #1:9 #1:9 #1:9 #1:9 #1:9 #1:9 #1:9 #1:9 #1:9
Avg #1:9 #1:9 #1:9 #1:9 #1:9 #1:9 #1:9 #1:9 #1:9 #1:9 #1:9 #1:9
Table 3 – Taguchi L9 Orthogonal Array
An L9 Orthogonal Array provides all combinations of any four factors, so that each level of each factor is
combined with each level of every other factor. The L9 array contains an equal number of conditions for each
factor, so each factor level is tested an equal number of times.vii.
Taguchi pleads “dig wide, not deep”. Orthogonal arrays are designed to offer an efficient approach to
discover effects and indicate where more comprehensive examination may be warranted.
MEASUREMENTS:
The 100 mm long high aspect ratio (50:1) test specimens were inserted into glass tubes printed with 1 mm
increments beginning at 100 mm (see Figure 2A). The glass tubes were then stood upright and sealed within
the pressure vessel such that the specimens could be observed and measured against the 1 mm increments
(see Figure 2B). The vessel was flushed once with CO2 and then charged and held at 750 PSI for 24-hours
(“24 Hr soak” / “soaking period”). During the soaking period, visual observation was made of the change in
linear length and the values recorded. The value after a 24-hour soaking period was used in this study.

Likewise, subsequent to RGD, visual observation was made of the change in linear length and the values
recorded. The value two (2) minutes after the RGD event was used in this study.
Figure 2A – Test vials Figure 2B – Vial increments
Evaluation of the HNBR compounds was based upon the specimen’s change in volume (“∆ Vol %”) as set
forth in Equation 3. The original specimen volume is calculated per Equation 1. Since observations are
measured as “length” in millimeters, the percent change in length is calculated per Equation 2 as an
intermediate step in calculating the percent change in volume. .
Eq. 1 : Volume initial = Vol i = Length Initial x Width Initial x Depth initial
Eq. 2 : ∆ Length % = ∆Len% = ( Length final - Length initial ) / Length initial
Eq. 3 : ∆ Vol % = { [Length I x (1 + ∆Len%)] x [Width I x (1 + ∆Len%)]

x [Depth I x (1 + ∆Len%)] – Vol i } / Vol i
PROCEDURE:
Test specimens were cut and placed in the measuring tubes. Three measuring tubes at a time were placed
inside a pressure vessel that was subsequently flushed with 99% CO2. After a single flushing with CO2, the
pressure vessel was pressurized with fresh CO2 to 750 PSI. This pressure was held for four (4) hours at
room temperature. After four hours, this pressure was released through a regulator over a two minute period
(350 PSI/minute). Upon reaching ambient pressure, the time was marked and measurements were made
after 2 minutes, 10 minutes, 30 minutes, 1 hour, and 2 hours.
EXPERIMENT 1 - EFFECT OF ACRYLONITRILE IN A MARGINALLY REINFORCED HNBR WHEN

IMMERSED IN CO2.
The first objective was to isolate the acrylonitrile content of HNBR and study its behavior when subject to
CO2. Four grades of HNBR polymer of similar saturation (95% – 96%) and viscosity (72 – 85 mooney) were
examined. Several criteria were used to determine the test formulas: 1) minimize the number of ingredients,
2) achieve a state-of-cure that would merely facilitate preparing the samples, and 3) minimize the interaction
of carbon-black and polymer.

The effect that different reinforcing particle sizes have upon swelling in high pressure gas relative to other
particle sizes has been previously reported by Hertzviii. N990 was settled upon since it is the largest particle
size with the least amount of elastomer reinforcement. It was determined that this particle size and a
moderate loading would minimize polymer-to-filler interaction and allow better observation of the polymer’s
behavior. The change in volume of the first trial specimens, subsequent to RGD, exceeded the measuring
apparatus. Thus, in this experiment, an additional trial used specimen lengths of 80 mm so that their
substantial change in volume could be accurately measured.
The polymers, filler, and cure system were mixed on an open roll 12 inch mill. The formulas used to study
the effect of acrylonitrile content in CO2 are listed in Table 4. Stress-strain data for these compounds was
calculated using ASTM D412 Test Method ‘A’ and compiled in Table 4a. Normalizing the formulas’ modulus
was not only impractical given the differences in acrylonitrile content, but unnecessary since there is no
correlation between volume change of the specimens and their measured modulus.
TEST FORMULAS:
44% acn 36% acn 25% acn 17% acn

Ingredient Phr Ingredient Phr Ingredient Phr Ingredient phr
Zetpol 2010 100 Zetpol 2010 100 Zetpol 3310 100 Zetpol 4310 100
Zinc Oxide 5 Zinc Oxide 5 Zinc Oxide 5 Zinc Oxide 5
Peroxide 5 Peroxide 5 Peroxide 5 Peroxide 5
N990 black 30 N990 black 30 N990 black 30 N990 black 30
Table 4 – Peroxide cured, 95-96% saturated HNBR polymers
ASTM D412 Test Method A - RESULTS:

HNBR Code ACN% M25 M50 M100 M300
HNBR-1010 44% 146 psi 175 psi 213 psi 688 psi
Table 4a – ASTM D412 Test Method A data
RESULTS AND DISCUSSION:
The swelling of elastomers under pressure in CO2 are merely a prelude to future behavior. A significantly
different story emerges subsequent to rapid gas decompression (“RGD”). Release of the hydrostatic load on
the materials’ surface allows the absorbed gas to expand causing significant swelling. Over a brief amount of
time, however, the gas diffuses from the elastomers allowing them to return to their initial geometry. Figure 4
illustrates swelling under pressurized CO2 and subsequent to RGD. Assuming elastomers to be isotropic
materials, the % linear change in the test specimens reflects approximately a 3X change in volume. The
changes in volume attributable to CO2 absorption presumably precede seal failure modes.
Previous work by Hertz III found that EPDM swelled slightly less than HNBR under pressurized CO2.
However, upon RGD, the EPDM swelled slightly more than HNBR but degassed more quickly and returned
to normal size more quicklyix. This past observation is relevant to the current study. In a limited sense, an
HNBR with 19% acrylonitrile content is more similar to an EPDM than an HNBR grade with higher ACN
content.
In this study, the amount of ACN content had minimal effect on swelling in CO2 under pressure during the
first 4 hours (240 minutes). The measured differences attributable to ACN content are mixed. Likewise,
during the first 10 minutes of an RGD event, the relationship of ACN to swelling was mixed. However, shortly
thereafter, the less ACN content there was in the HNBR the more quickly the compound released CO2.
The implications of the post RGD observation may be complicated for the seal engineer. Upon an RGD
event, the question becomes at which point in time does the elastomer compound suffer mechanical
damage?

The results of this experiment are recorded in Table 5 and depicted graphically in Figure 4.
10 30 60 120
4 hours 2 minutes minutes minutes minutes minutes
Elastomers 750 PSI post RGD post RGD Post RGD Post RGD Post RGD
HNBR 44% ACN 16% 139% 255% 264% 255% 229%
HNBR 36% ACN 13% 133% 221% 205% 153% 73%
HNBR 25% ACN 13% 272% 300% 167% 33% 8%
HNBR 19% ACN 16% 205% 264% 120% 12% 0%
Table 5 – HNBR % change in volume in CO2 under pressure and subsequent to RGD.
Figure 4 – Chronologic plot of HNBR swelling under pressure and subsequent to RGD
EXPERIMENT 2 - EFFECT OF ACRYLONITRILE, CURE, AND REINFORCEMENT ON HNBR WHEN

IMMERSED IN CO2.
The objective of Experiment 2 was to study the effects of CO2 on compositions of varying acrylonitrile
content, different grades of particle black (“carbon black”), different loadings of carbon black, and different
levels of cure. Evaluation was conducted using a Taguchi L9 Orthogonal Array. The test matrix is
documented in Table 6. Zinc oxide loading (5 phr) was constant and uncontrolled in this study. The percent
saturation of these polymers ranged from 95 to 96%.

LEVELS
FACTORS 1 2 3
25% 19%
A POLYMER 36% ACN ACN ACN
B CURE PHR 4 5 6
FILLER
C TYPE N330 N762 N990
D FILLER PHR 30 50 70
uncontrolled ZnO 5 5 5
Table 6 – Factors and associated levels for HNBR study
The test matrix in Table 6 results in the Taguchi L9 Orthogonal Array depicted in Table 7. The results of this
testing are found in Tables 8 and 9.
Cure Filler 4hr soak Post RGD

Condition Polymer PHR Filler type PHR % Vol ∆ % Vol ∆
36% ACN
#1 (A1) 4 (B1) N330 (C1) 30 (D1) 9% 186%
36% ACN
#2 (A1) 5 (B2) N762 (C2) 50 (D2) 12% 110%
36% ACN
#3 (A1) 6 (B3) N990 (C3) 70 (D3) 9% 60%
25% ACN
#4 (A2) 4 (B1) N762 (C2) 70 (D3) 19% 238%
25% ACN
#5 (A2) 5 (B2) N990 (C3) 30 (D1) 12% 186%
25% ACN
#6 (A2) 6 (B3) N330 (C1) 50 (D2) 12% 52%
19% ACN
#7 (A3) 4 (B1) N990 (C3) 50 (D2) 16% 186%
19% ACN
#8 (A3) 5 (B2) N330 (C1) 70 (D3) 12% 40%
19% ACN
#9 (A3) 6 (B3) N762 (C2) 30 (D1) 9% 82%
Table 7 – Taguchi L9 Orthogonal Array for HNBR study

Volume Change after 4 hour soak in 750 PSI CO2 at room temperature
4phr 5phr 6phr 30phr 50phr 70phr

36% 25% 19% cure cure cure N330 N762 N990 black black black
A1 A2 A3 B1 B2 B3 C1 C2 C3 D1 D2 D3
#1 9% 9% 9% 9%
#2 12% 12% 12% 12%
#3 9% 9% 9% 9%
#4 19% 19% 19% 19%
#5 12% 12% 12% 12%
#6 12% 12% 12% 12%
#7 16% 16% 16% 16%
#8 12% 12% 12% 12%
#9 9% 9% 9% 9%
Tota
l 0.31 0.44 0.38 0.44 0.37 0.31 0.34 0.41 0.38 0.31 0.41 0.41
Avg 0.10 0.15 0.13 0.15 0.12 0.10 0.11 0.14 0.13 0.10 0.14 0.14
Table 8 - % Volume Change after 4 hour soak in 750 PSI CO2
Volume Change subsequent to RGD
4phr 5phr 6phr 30phr 50phr 70 phr

36% 25% 19% cure cure cure N330 N762 N990 black black black
A1 A2 A3 B1 B2 B3 C1 C2 C3 D1 D2 D3
#1 186% 186% 186% 186%
#2 110% 110% 110% 110%
#3 60% 60% 60% 60%
#4 287% 287% 287% 287%
#5 186% 186% 186% 186%
#6 52% 52% 52% 52%
#7 261% 261% 261% 261%
#8 40% 40% 40% 40%
#9 82% 82% 82% 82%
Total 3.56 5.25 3.83 7.34 3.37 1.94 2.79 4.78 5.07 4.54 4.22 3.88
Avg 1.19 1.75 1.28 2.45 1.12 0.65 0.93 1.59 1.69 1.51 1.41 1.29
Table 9 - % Maximum Volume Change subsequent to RGD
RESULTS AND DISCUSSION:
The swelling of HNBR compounds in CO2 under pressure is appreciable. A range of volume increases from
9% to 19% was observed. However, upon RGD, the amount of swelling in HNBR is substantial. Ranges of
40% to 238% were observed. Nevertheless, the reader should not simply conclude that HNBR is unsuitable
for CO2. These test compounds were designed to provide guidance rather than optimal solutions.
Clearly, CO2 contra-indicates the use of HNBR in applications where the gas is present in appreciable
quantities. However, the seal engineer frequently finds the other merits of HNBR to require its use in spite of
this particular shortcoming.
The data clearly indicated that crosslink density, which is a function of the amount of curative (and the HNBR
grade’s degree of saturation) had the greatest impact on swelling subsequent to RGD. For this very reason,
an attempt was made to test materials possessing a similar degree of saturation. A reduction in saturation
would allow for a higher crosslink density.
It was further apparent that smaller particle size carbon black also mitigated swelling. N330 carbon black is
comprised of particles measuring 28 to 36 nm. N762 particles measure 60 to 100 nm while N990 particles
measure 250 to 350 nm. A smaller particle size presents greater surface area per unit of weight and hence
more reinforcement.
Higher loadings of carbon black demonstrate diminishing returns on mitigating swelling. The experienced
rubber compounder knows that these relationships will likely cause problems in achieving other material
attributes when designing an HNBR compound.
Figure 8- Factor/Level effects on HNBR compound swelling in CO2.
SOURCES OF ERROR
Changes in specimen length were recorded by visual examination. As such, a significant source of error
could be introduced. A one (1) millimeter error in reading the specimen length translates to roughly a 3mm3
error in volume. In evaluating data, the reader may want to consider volume change within a range rather
than as a single point.
All mixing of test batches was conducted on open roll mills, subject to loss of ingredients during the mixing
process or marginal errors during ingredient weigh up. Test batch weigh-up was conducted on industrial
scales with ±0.1 gram accuracy. Test compounds were mixed using 500 grams of polymer. With curatives
weighed as low as 4 phr (25 grams per batch), a 0.5 gram error would amount to a 2% deviation from the
test formula.
CO2 Pressure was regulated for the 750 PSI soak. Depressurization was also regulated to 375 PSI/minute.
Volume changes were rapid within the first 10 minutes of RGD and a simultaneous read of all three samples
was not practical. Thus, it is reasonable to assume a specimen length tolerance of ±1mm for the post RGD
data.

SUMMARY
This study utilized “Design of Experiments” to reveal significant relationships with an HNBR compound that
affect its interaction with CO2. No attempt was made to optimize these compounds.
Percent saturation of the HNBR, while mostly similar amongst the test compounds, was uncontrolled.
1) The predominant factor in reduction of CO2 induced swelling subsequent to RGD is the amount of
curative utilized.
2) The acrylonitrile content in HNBR is inconsequential to swelling in CO2 while the material is under
pressure. On the other hand, the acrylonitrile content of HNBR is a significant determinant of the
propensity to swell subsequent to an RGD event. While other factors may predominate in
determining the maximum swell of HNBR immediately subsequent to RGD, the rate of degassing
clearly increases as the amount of acrylonitrile decreases.
3) Smaller carbon black particle sizes mitigate swelling subsequent to an RGD event. Likewise,
increasing the loading of carbon black appears to mitigate swelling but with diminishing returns as
loading increases.
ACKNOWLEDGEMENTS
Foremost, the author would like to thank Zeon Chemicals for their contribution of HNBR samples to this
study. The author would also like to thank Taylor Boyle for his assistance in this work. Additional gratitude is
extended to Dan Hertz, Jr for his encouragement and prior research.
TRADEMARKS
Zetpol® is the registered trademark of Zeon Chemicals

REFERENCES
i
Knox, J.H., “Molecular Thermodynamics”, p129-130, John Wiley & Sons (Rev.Ed. 1978)
ii
The temperature and pressure at which the vapor, liquid, and solid phases of a substance are in
equilibrium.
iii
The state of fluid in which the fluid and gas both have the same density.
iv
Lide, David R.,CRC, “Handbook of Chemistry and Physics”, p.6-54 (77th Ed.1996).
v
Hertz, Jr., D.L., “Elastomers in the Hot Sour gas Environment”, Elastomerics (Sept 1986).
vi
The ASTM table simply calculates percent volume change as the difference between initial
calculated volume and final calculated volume divided by the initial volume. The final volume
assumes an isotropic material response such that percent change in length will be the same across
all three dimensions.
vii
“Taguchi Approach to Quality Optimization Series”, Technicomp,Inc., Cleveland, OH, p 2-2, (4th
Printing, 1988).
viii
Hertz, Jr., D.L., “Sealing Technology”, Rubber Products Manufacturing Technology, p.786, Marcel
Dekker, Inc. (1994).
ix
Hertz III, D.L., “Elastomers in CO2”, High Performance Elastomers & Polymers for Oil & Gas 2012,
Int’l Conference, Aberdeen, SCO, UK (April 2012)

High Performance POLYMERS for Oil & Gas 2014
Organised by:
Sponsors & Exhibitors:
INNOVATIONS WITH IMPACT
Supporting Associations & Media Partners:
World ils

High Performance Polymers For Oil - Rapra

Uploaded by

Copyright:

Available Formats

High Performance Polymers For Oil - Rapra

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

High Performance Polymers For Oil - Rapra

Uploaded by

Copyright:

Available Formats

High Performance POLYMERS for Oil & Gas 2014

Sponsors & Exhibitors:

INNOVATIONS WITH IMPACT

Supporting Associations & Media Partners:

15-16 April 2014

© Smithers Information Ltd, 2014

Paper 1 Opening Keynote

Paper 2 Offshore Drilling: Use of elastomeric materials and future challenges

CORROSION AND FAILURE: DEVELOPING QUALIFICATIONS AND SPECIFICATIONS FOR FUTURE

Paper 4 Nonmetallic Program for the Oil and Gas industry

Paper 5 Rapid Gas Decompression Resistance of Elastomeric O-Rings to Supercritical CO2

Paper 6 Industry Standards - A Blessing or a Curse?

Paper 7 CNT Technology and Dispersion

Paper 9 Polymer Graphene Nanocomposites for Oil and Gas Processes

Paper 10 Challenges of Temperature Extremes for Elastomer Materials

Paper 11 Cold Temperature Effects on Polymers - Cryogenic Spill Protection

Paper 12 Arlon 3000XT

MECHANISMS AND MANIFESTATIONS OF POLYMER AGEING AND FATIGUE, TECHNOLOGICAL

Paper 14 Drilling Fluid Influence on Elastomers

DEVELOPING AND MODELLING MATERIALS FOR THE FUTURE OF THE INDUSTRY

Paper 15 Modeling and Design of Reinforced Elastomeric Products

Paper 16 Development of Materials for Sealing Solutions in HPHT conditions

Paper 17 HNBR Rubber in CO2

PAPER UNAVAILABLE AT TIME OF PRINT

Page 1 of 2 pages Paper 1 - Lawson

Paper 1 - Lawson Page 2 of 2 pages

OFFSHORE DRILLING: USE OF ELASTOMERIC MATERIALS

Page 1 of 8 pages Paper 2 - Arteaga

Figure 1 – Schematic of BOP submitted by Cameron Iron Works for patent

Figure 2 – Generic layout of subsea drilling equipment

Paper 1 - Arteaga Page 2 of 8 pages

Figure 3 – Gimbal with six shock mounts

Figure 4 – Diverter cross section

Figure 5 – Double Seal Assembly

Paper 1 - Arteaga Page 4 of 8 pages

Figure 6 – Cameron DL Annular BOP

Paper 1 - Arteaga Page 6 of 8 pages

Figure 7 – Cameron BOP Rams

Page 7 of 8 pages Paper 2 - Arteaga

Paper 1 - Arteaga Page 8 of 8 pages

IFL™ - A NOVEL APPROACH TO THE REHABILITATION OF

Robert Walters is the founder and Chairman of Anticorrosion Protective Systems, a

THE DEVELOPMENT OF A VIABLE REHABILITATION SYSTEM FOR DEPLOYMENT IN NEW AND

PETRONAS Carigali (PCSB) is the owner and

Page 1 of 6 pages Paper 3 - Walters

THE MISSION STATEMENT:

- High tensile and good physical properties.

LINER QUALIFICATION AND DEVELOPMENT PROCESS

Paper 3 - Walters Page 2 of 6 pages

- Alteration of the fiber reinforcement, so as to provide improved tensile capabilities.

The final enhanced IFL™ Liner matrix comprises of a

Figure 2. IFL™ Liner matrix

Predictive Installation software has also been developed that will

The IFL™ Liner system is initially available in diameters suitable for

INDUSTRY MOTIVATION TO UTILIZE IFL™ LINERS

The majority of sub-sea pipelines are constructed from carbon steel,

Page 3 of 6 pages Paper 3 - Walters

IFL™ can be utilized for the rehabilitation of an existing sub-sea pipeline

- Unforeseen operational parameters such as CO2 or SRB corrosion