High Performance POLYMERSHigh Performance 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: [email protected];
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 High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
OFFSHORE DRILLING: USE OF ELASTOMERIC MATERIALS AND FUTURE CHALLENGES
Nicolas Arteaga Cameron 4601 Westway Park, Houston, TX 77040 Tel: +1 (281) 606-6190 email: [email protected]
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 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014
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 High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
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 Page 3 of 8 pages Paper 2 - Arteaga 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014
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
Paper 1 - Arteaga Page 4 of 8 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland 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. Page 5 of 8 pages Paper 2 - Arteaga 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014
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.
Paper 1 - Arteaga Page 6 of 8 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
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.
Page 7 of 8 pages Paper 2 - Arteaga 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014
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.
Paper 1 - Arteaga Page 8 of 8 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
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: [email protected]
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 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014
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 High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
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.
Page 3 of 6 pages Paper 3 - Walters 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014
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 anticipated corrosion is taking place and that unless the corrosion Figure 4. IFL™ Liner in 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.
Paper 3 - Walters Page 4 of 6 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
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 6. IFL™ Liner in Folded Form Ready for Offshore Deployment. Figure 5. Liner Shipping Reel Loaded with Liner. 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
Page 5 of 6 pages Paper 3 - Walters 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014
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.
Paper 3 - Walters Page 6 of 6 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
NONMETALLIC PROGRAM FOR THE OIL AND GAS INDUSTRY
Abdullah Al-Dossary Nonmetallic Engineer, Saudi Aramco Email: [email protected]
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
15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014
Paper 4 - Al-Dossary Page 2 of 2 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
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 [email protected]
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 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014
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 3 their findings for CO2 immersion . 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 High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
The 2012 presentation by Bjørn Melve is in relation to work carried out by Statoil to evaluate swelling and the 4 consequences of such . 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.
6 The final paper which gives significant information in relation to swell in CO2 is that of Zoltan Major et al . 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. 7 The paper by Morgan, Sully and Davies (1997) evaluated cycle tests in CO2 . 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 Page 3 of 10 pages Paper 5 - Warren 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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.
Paper 5 - Warren Page 4 of 10 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
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.
Page 5 of 10 pages Paper 5 - Warren 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014
.
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 Paper 5 - Warren Page 6 of 10 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland 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
Page 7 of 10 pages Paper 5 - Warren 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014
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 8 that HNBR is more suitable for RGD resistance than FKM with high levels of CO2 . 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.
Paper 5 - Warren Page 8 of 10 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
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.
Page 9 of 10 pages Paper 5 - Warren 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014
Paper 5 - Warren Page 10 of 10 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
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 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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 High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
NEXT GENERATION CNT/RUBBER SOLUTIONS FOR OIL AND GAS INDUSTRIES
Michael CLAES, Alicia RUL Nanocyl SA Rue de l’Essor, 4 - 5060 SAMBREVILLE - Belgium [email protected]
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 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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.
7 TEM image of Pure NC7000 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 High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland 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 Page 3 of 6 pages Paper 7 - Claes 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014
Electrical conductivity of NC7000™, the commercial CNT product of Nanocyl, in comparison with a conventional Carbon black N550.
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, in comparison with a conventional Carbon black N550.
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.
Paper 7 - Claes Page 4 of 6 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
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 …
Page 5 of 6 pages Paper 7 - Claes 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014
Paper 7 - Claes Page 6 of 6 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
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: [email protected]
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 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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 High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
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 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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 High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
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: [email protected]
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 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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 High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
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.
Page 3 of 6 pages Paper 9 Advincula 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014
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. Paper 9 Advincula Page 4 of 6 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
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.
Page 5 of 6 pages Paper 9 Advincula 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014
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.
Paper 9 Advincula Page 6 of 6 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
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: [email protected]
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 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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 High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
0
-5
-10 y = 0.0198x - 33.372 -15 R² = 0.9973 C) °
Tg ( -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 permeation
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?
Page 3 of 16 pages Paper 10 - Morgan 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014
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.
Paper 10 - Morgan Page 4 of 16 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
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
6
Stress (MPa) 4
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.
Page 5 of 16 pages Paper 10 - Morgan 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014
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
Paper 10 - Morgan Page 6 of 16 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
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 Page 7 of 16 pages Paper 10 - Morgan 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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. Paper 10 - Morgan Page 8 of 16 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
36
32
28
24
20 160°C 16 170°C 180°C 12 50% Modulus (MPa)
8
4
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
100 160°C 80 170°C 180°C 60
Elongation at Break (%) Break at Elongation 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).
Page 9 of 16 pages Paper 10 - Morgan 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014
-2.6 -2.8 -3 -3.2
-3.4 y = -7325.4x + 13.379 R² = 0.9991 -3.6 -3.8 -4
ln(1/time to 50% change in modulus) 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
2.5 200°C 215°C 2.0 230°C Modulus (MPa)Modulus
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
Paper 10 - Morgan Page 10 of 16 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
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
10000 80
8000 60
6000 40 pressure (psi) temperature ('C) 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.
Page 11 of 16 pages Paper 10 - Morgan 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014
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
Paper 10 - Morgan Page 12 of 16 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
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 0-20C 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.
Page 13 of 16 pages Paper 10 - Morgan 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014
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. Paper 10 - Morgan Page 14 of 16 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
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.
Page 15 of 16 pages Paper 10 - Morgan 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014
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
Paper 10 - Morgan Page 16 of 16 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
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 [email protected]
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 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014
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 High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
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
Page 3 of 8 pages Paper 11 - Viale 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014
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
Paper 11 - Viale Page 4 of 8 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
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).
Page 5 of 8 pages Paper 11 - Viale 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014
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.
Paper 11 - Viale Page 6 of 8 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
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. Page 7 of 8 pages Paper 11 - Viale 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014
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
Paper 11 - Viale Page 8 of 8 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
ARLON 3000XT: A NEW HIGH PERFORMANCE MATERIAL DESIGNED FOR EXTREME ENVIRONMENTS
Kerry Drake, Senior Scientist, and Burak Bekisli, Scientist Greene, Tweed & Co Email: [email protected] and [email protected]
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 150F(66C) to 300F(149C).
Page 1 of 22 pages Paper 12 - Drake 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014
At temperatures above 300F(149C) 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 440F(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 High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
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- Page 3 of 22 pages Paper 12 - Drake 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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 Paper 12 - Drake Page 4 of 22 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
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). Page 5 of 22 pages Paper 12 - Drake 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014
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(20C) higher than temperatures typically used for PEEK ISO/Norsok certification.
Paper 12 - Drake Page 6 of 22 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
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 752F(400C) that are not readily processable [McGrail].
(PEEK)
(PEK) (PEKEKK)
O O
aryl CO=ket one O=et her
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 325F(163C), which was equivalent to that of PEKEKK.
Page 7 of 22 pages Paper 12 - Drake 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014
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 above752F(400C), Arlon 3000 XT still maintained its structural and mechanical integrity (Figure 3.2.2).
Paper 12 - Drake Page 8 of 22 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
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 at752F(400C). (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.
Page 9 of 22 pages Paper 12 - Drake 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014
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.
Paper 12 - Drake Page 10 of 22 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
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 Page 11 of 22 pages Paper 12 - Drake 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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).
Paper 12 - Drake Page 12 of 22 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
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”).
Page 13 of 22 pages Paper 12 - Drake 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014
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.
Paper 12 - Drake Page 14 of 22 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland
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)
CTE (T CTE (T>Tg) ASTM E831 (mm/m per °F x10-5) 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). Page 15 of 22 pages Paper 12 - Drake 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 Arlon® 3000 Property Standard Units PEK XT *Dielectric Breakdown ASTM D618-A kV/mm 15.5 15.2 Voltage *Surface Resistivity ASTM D257 Ohm 6.62x1016 1.95x1017 *Volume Resistivity ASTM D257 Ohm-cm 4.03 x1017 3.49x1017 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. Paper 12 - Drake Page 16 of 22 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland 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. Page 17 of 22 pages Paper 12 - Drake 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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). Paper 12 - Drake Page 18 of 22 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland 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). Page 19 of 22 pages Paper 12 - Drake 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 0.035 PEK Arlon 3000 XT 0.03 0.025 0.02 0.015 0.01 Change dimension in d1 (inch) 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). Paper 12 - Drake Page 20 of 22 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland 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. Page 21 of 22 pages Paper 12 - Drake 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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. Paper 12 - Drake Page 22 of 22 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland 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: [email protected] 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 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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 High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland 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. Page 3 of 12 pages Paper 13 - Hulme 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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 Paper 13 - Hulme Page 4 of 12 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland 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. Page 5 of 12 pages Paper 13 - Hulme 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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. Paper 13 - Hulme Page 6 of 12 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland 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. Page 7 of 12 pages Paper 13 - Hulme 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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. Paper 13 - Hulme Page 8 of 12 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland 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 Page 9 of 12 pages Paper 13 - Hulme 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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. Paper 13 - Hulme Page 10 of 12 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland . 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. Page 11 of 12 pages Paper 13 - Hulme 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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 Paper 13 - Hulme Page 12 of 12 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland DRILLING FLUID INFLUENCE ON ELASTOMERS Helmut Benning & Marcus Davidson Baker Hughes Email: [email protected] and [email protected] 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 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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 High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland 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. Page 3 of 10 pages Paper 14 - Benning 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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 Paper 14 - Benning Page 4 of 10 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland 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. Page 5 of 10 pages Paper 14 - Benning 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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.. Paper 14 - Benning Page 6 of 10 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland 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. Page 7 of 10 pages Paper 14 - Benning 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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. Paper 14 - Benning Page 8 of 10 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland 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. Page 9 of 10 pages Paper 14 - Benning 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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. Paper 14 - Benning Page 10 of 10 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland 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: [email protected] 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 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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 High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland 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. Page 3 of 8 pages Paper 15 - Brown 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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 Paper 15 - Brown Page 4 of 8 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland 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. Page 5 of 8 pages Paper 15 - Brown 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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. Paper 15 - Brown Page 6 of 8 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland 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. Page 7 of 8 pages Paper 15 - Brown 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 Paper 15 - Brown Page 8 of 8 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland 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: [email protected] 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 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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 High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland 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 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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 High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland HNBR IN CO2 D L Hertz III Seals Eastern Inc. Red Bank, NJ, 07701, USA Email: [email protected] 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 i bent (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 iii iv CO2 is about .518 MPa (75.1 PSI) at -56.6˚C. The critical point is 7.375 MPa (1070.4 PSI) at 31.1˚C (88˚F). In the course of this study, super-critical conditions were not present. Page 1 of 12 pages Paper 17 - Hertz 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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 High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland 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. Page 3 of 12 pages Paper 17 - Hertz 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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 #8 h h h h h #9 i i i i i Condition 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. Paper 17 - Hertz Page 4 of 12 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland 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. Page 5 of 12 pages Paper 17 - Hertz 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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 HNBR-2010 36% 136 psi 161 psi 190 psi 625 psi HNBR-3310 25% 106 psi 130 psi 171 psi 746 psi HNBR-4310 19% 100 psi 132 psi 211 psi 994 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? Paper 17 - Hertz Page 6 of 12 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland 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%. Page 7 of 12 pages Paper 17 - Hertz 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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 Paper 17 - Hertz Page 8 of 12 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland 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 Page 9 of 12 pages Paper 17 - Hertz 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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. Paper 17 - Hertz Page 10 of 12 pages High Performance Polymers for Oil & Gas 2014 15-16 April 2014 – Edinburgh, Scotland 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 Page 11 of 12 pages Paper 17 - Hertz 15-16 April 2014 – Edinburgh, Scotland High Performance Polymers for Oil & Gas 2014 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) Paper 17 - Hertz Page 12 of 12 pages High Performance POLYMERSHigh Performance for Oil & Gas 2014 Organised by: Sponsors & Exhibitors: INNOVATIONS WITH IMPACT Supporting Associations & Media Partners: World ils