Elastomers in CO2
Presented at the
High Performance Elastomers & Polymers for Oil & Gas 2012
International Conference
Aberdeen, SCO, UK
20-21st April, 2012
D. L. Hertz III
Seals Eastern Inc.
Red Bank, NJ, 07701
USA
Prepare for your exams
Study with the several resources on Docsity
Earn points to download
Earn points by helping other students or get them with a premium plan
Guidelines and tips
Prepare for your exams
Study with the several resources on Docsity
Earn points to download
Earn points by helping other students or get them with a premium plan
Carbon dioxide is a naturally occurring gas that is frequently encountered in hydrocarbon ... greater absorption of CO2 by HNBR, a non-polar elastomer.
Typology: Schemes and Mind Maps
2022/2023
1 / 20
This page cannot be seen from the preview
Don't miss anything!
Related documents
Partial preview of the text
Download Elastomers in CO2 and more Schemes and Mind Maps Technology in PDF only on Docsity!
Elastomers in CO 2
Presented at the
High Performance Elastomers & Polymers for Oil & Gas 2012
International Conference
Aberdeen, SCO, UK
20-21 st^ April, 2012
D. L. Hertz III
Seals Eastern Inc.
Red Bank, NJ, 07701
USA
ABSTRACT
Carbon dioxide is a naturally occurring gas that is frequently encountered in hydrocarbon environments. The CO 2 molecule has unique characteristics causing popular oilfield elastomers to swell when in contact with it. Relatively small concentrations of CO 2 in hydrocarbon mixtures can cause significant seal swelling. More significantly, the effect of absorbed CO 2 upon rapid gas decompression can be catastrophic if consideration is not given to the choice of polymer, cure, and particle reinforcement.
This study utilized Taguchi’s approach to experimental design that studied the effect of cure system and particle black reinforcement upon an elastomeric compound’s swelling in CO 2. The elastomeric compounds studied included HNBR, EPDM, FKM Type I, FKM Type II, FKM Type III, FEPM (TFE/P), and FEPM (TFE/E/PMVE). Amine curatives in the FKM class materials were studied. Peroxide loading in the FEPM, FKM Type III and hydrocarbon elastomers was studied. Finally, particle black grades (N990, N550, and N330) and their respective loadings were studied.
INTRODUCTION
Carbon Dioxide (CO 2 ) is a naturally occurring colorless, odorless gas. It is frequently found in hydrocarbon reserves. CO 2 , in the gaseous state, is denser than air with a specific gravity of 1. kg/m3. CO 2 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, CO 2 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 CO 2 ’s transient dipole moments, the molecule appears benti^ (e.g. like an H 2 O molecule). Thus, the simple rule of thumb of “likes dissolves likes” is misleading if you consider CO 2 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 CO 2 is about .518 MPa (75.1 PSI) at -56.6˚C. The critical pointiii^ is 7.375 MPa (1070.4 PSI) at 31.1˚C (88˚F).iv^ In the course of this study, super-critical conditions were not present.
The solvating powers of CO 2 are well documented and applications utilizing supercritical CO 2 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 CO 2 is present in the hydrocarbon stream. Modest amounts of CO 2 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 CO 2 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 CO 2 in a hydrocarbon gas mixture is substantially smaller than the molar fraction of the other gases present (e.g. N 2 , He, O 2 , CH 4 , C 2 H 6 , C 3 H 8 , etc.). Thus, in the context of partial pressure, the CO 2 condition in this study would exist in well pressures of several thousand PSI where the CO 2 molar fraction is only a few percentage points. On the other hand, in a situation such as CO 2 reinjection, field results might differ substantially from those observed herein.
Figure 1A – Test Fixture Configuration Figure 1B – Observation Vessel and test vials
ELASTOMERS:
The following popular oilfield elastomers included in this study are found in Table 1. Polymers were studied in three experimental groups as determined by their molecular structure and/or cure system:
- Peroxide cured hydrocarbon polymers, both non-polar and polar,
- Amine cured FKM class fluoroelastomers (Types 1 & 2),
- Peroxide cured fluoroelastomers utilizing a triallyl-isocyanurate (“TAIC”) coagent, also referred to as “peroxy-TAIC” cured polymers. All the elastomer compounds herein were mixed on an open 12-inch roll mill. Since the primary objective was a comparative analysis of elastomers, the experiment design used elastomers as a controlled factor. Obviously an effort was made to group similar elastomers. Any future study whose purpose is product improvement should treat a specific polymer structure as an uncontrolled factor except in instances where grades of a specific polymer type are being examined.
(Experiment #) Test Group
D
Designation Trade name Specific gravity
(#1) Hydrocarbon HNBR Zetpol® 2010 1.
(#1) Hydrocarbon EPDM Royalene® 575^ 0.
(#2) FKM Amine cure FKM Type 1 Viton® A 1.
(#2) FKM Amine cure FKM Type 2 Viton® B 1.
(#3) Peroxy-TAIC cure FKM Type 3 Technoflon® PL 855 1.
(#3) Peroxy-TAIC cure FEPM Aflas® 100H 1.
(#3) Peroxy-TAIC cure FEPM Viton® ETP 600S 1. 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 volume vi^.
EXPERIMENTAL
METHODOLOGY:
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 L 4 matrix used to test three (3) factors at two (2) levels.
FACTOR
LEVEL
A A1 A
B B1 B
C C1 C
Table 2 – Taguchi L 4 design of experiment
Figure 2A – Test vials Figure 2B – Vial increments
CURED GUM POLYMER, NON-REINFORCED
Commercially available fluoroelastomer and hydrocarbon class polymers were selected based upon their acceptance and perception as high performance polymers in the oilfield. HNBR was selected over NBR since it is perceived as a high performance polymer and differs only in backbone saturation. Reinforcing fillers were omitted during the examination of cured gum polymer swelling in CO 2. In subsequent experiments different grades of particle black (“carbon black”) were incorporated. Specific gravity of the polymers is noted in Table1 should the reader want to consider differences in weight to volume when comparing volume swell of the different reinforced polymers. The effect that different reinforcing particle sizes have upon swelling in high pressure gas relative to other particle sizes has been previously reported by Hertz viii^. First, an evaluation of non-reinforced polymers was undertaken. The polymers and their respective cure systems were mixed on an open roll 12 inch mill. The formulas for non-reinforced, cured gum polymers are tabulated below in Tables 4 through 6.
TEST FORMULAS:
GROUP 1 - Hydrocarbon Elastomers, peroxide cured
HNBR EPDM
Ingredient PHR Ingredient PHR Zetpol® 2010 100 Royalene® 575 100 Peroxide 6 Peroxide 5
Table 4 – Peroxide cured hydrocarbon gum polymers
GROUP 2 – FKM Class polymers, Amine-BPAF cured
FKM Type 1 FKM Type 2
Ingredient PHR Ingredient PHR Viton® A 100 Viton® B 100 Diak® 1 2 Diak® 3 3 MgO 15 MgO 15
Table 5 – Amine cured FKM class gum polymers
GROUP 3 – Fluoroelastomers, peroxide-TAIC cured
FEPM (TFE/P)
FEPM
(TFE/E/PMVE) FKM Type 3 Ingredient PHR Ingredient PHR Ingredient PHR
Aflas® 100H 100 Viton® ETP 600 100 Technoflon® PL855 100 TAIC 5 TAIC 5 TAIC 5 Peroxide 3 Peroxide 3 peroxide 3
Table 6 – Peroxide-TAIC cured no/low VDF gum polymers
RESULTS AND DISCUSSION:
One could reasonably assume that the amount of elastomer swelling in CO 2 might be a function of elastic modulus. Table 7 documents the measured elastic modulus (G’) of the specimens prior to the soaking period and the recorded change in length and calculated percent change in volume of each specimen after a 24-hour soak under pressure. After the soaking period, pressure was released over a ten minute period to return to atmospheric pressure (sea level).
The swelling of elastomers under pressure in CO 2 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 CO 2 and subsequent to RGD. Due to a 30mm limitation in the test vials, linear change beyond 30% was recorded as 30%. Assuming elastomers to be isotropic materials, the % linear change in specimens reflects approximately a 3X change in volume. The changes in volume attributable to CO 2 absorption precedes seal failure modes. Although not within the scope of this study, RGD damage (e.g. blisters, fissures) was typically observed in the materials.
Figure 4 – Chronologic plot of swelling under pressure and subsequent to RGD
EXPERIMENT 1 - HYDROCARBON ELASTOMERS, REINFORCED
Evaluation of carbon black reinforced hydrocarbon elastomers lent itself to a Taguchi L (^4) Orthogonal Array. The test matrix is documented in Table 8. There are significant differences between the two hydrocarbon elastomers tested. Specifically, the EPDM is non-polar while HNBR is polar. The HNBR analyzed has a 36% acrylonitrile content.
FACTORS
LEVELS
A POLYMER EPDM HNBR
B CURE PPH 5 6
C FILLER TYPE N990 N
uncontrolled FILLER PPH 30 30
Table 8 – Factors and associated levels for reinforced hydrocarbon elastomer study
Rather than full factorial testing, requiring 2 3 or eight (8) conditions, the following orthogonal array in Table 9 was evaluated.
Condition Polymer Cure pph Filler Type
#1 EPDM (A1) 5 (B1) N990 (C1)
#2 EPDM (A1) 6 (B2) N330 (C2)
#3 HNBR (A2) 5 (B1) N330 (C2)
#4 HNBR (A2) 6 (B2) N990 (C1)
Table 9 – Taguchi L 4 Orthogonal Array for reinforced hydrocarbon elastomer study
RESULTS AND DISCUSSION:
Swelling at the end of a 24-hour soak is recorded in Table 10. Swelling 2 minutes subsequent to the 10 minute RGD is recorded in Table 11. EPDM presented less swelling under pressure than HNBR. EPDM also exhibited less swelling 2 minutes subsequent to RGD. Filler particle size had no effect on swelling under pressure but had significant impact subsequent to RGD, with the smaller particle size mitigating swelling. PHR of cure had no effect on swelling under pressure. On the hand, an increase in curative corresponded with an increase in swelling subsequent to RGD in HNBR. A full factorial examination of this relationship within HNBR revealed an interaction of particle size and increased curative subsequent to RGD. Specifically, the study (Table 9A) revealed that increasing curative with an N increased swelling subsequent to RGD while increasing curative with an N330 decreased swelling subsequent to RGD.
HNBR w/ 30 phr black
5 phr Cure
6 phr Cure N330 56.1% 40.5% N990 64.3% 104.8% Table 9A – Post RGD % Volume Change study of particle size to cure phr
The swelling response of EPDM is somewhat counter intuitive given its non-polar composition and CO 2 ’s non-polar configuration. However, when consideration is given to the doubly degenerate bending of the CO 2 molecule, the polarity is in fact different. Thus, we also observed a
EXPERIMENT 2 - AMINE CURED FKM ELASTOMERS, REINFORCED
Evaluation of carbon black reinforced amine cured FKM elastomers lent itself to a Taguchi L 4 orthogonal array. The test matrix is documented in Table 12. The FKM I and FKM II studied do not utilize any coagent. Cure is accomplished via amine induced dehydrohalogenation. Due to the brand nature of the FKM Type 1 and FKM Type 2 studied, the amine initiators are specific to the polymer and are thus chemically different. An assumption has been made that the differences in chemical structure of the amine compounds will participate nominally, if at all, in the results whereas the differences in respective quantity of curative will be a more significant factor. Carbon black grade and magnesium oxide loadings were constant and uncontrolled in this study.
FACTORS
LEVELS
A POLYMER FKM‐I FKM‐II
B AMINE PHR 2 3
C FILLER PHR 30 70
Uncontrolled FILLER TYPE N990 N
Uncontrolled MgO PPH 15 15
Table 12 – Factors and associated levels for reinforced amine cure FKM study
The following Taguchi L 4 Orthogonal array per Table 13 was evaluated.
Condition Polymer Cure pph
Filler
Type
#1 FKM‐I (A1) 2 (B1) 30 (C1)
#2 FKM‐I (A1) 3 (B2) 70 (C2)
#3 FKM‐II (A2) 2 (B1) 70 (C2)
#4 FKM‐II (A2) 3 (B2) 30 (C1)
Table 13 – Taguchi L 4 Orthogonal Array for reinforced amine cure FKM elastomer study
RESULTS AND DISCUSSION:
FKM Type 1 exhibited less swelling than the FKM Type 2 both under pressure and after RGD. Increasing the particle black loading decreased swelling both under pressure and after RGD. The premise that carbon black does not absorb CO 2 renders this observation unremarkable. Incremental change in curative had no effect on swelling under pressure but an increase in curative appears to have increased swelling subsequent to RGD. As discussed in the results of Experiment 1, there is an interaction between curative phr and large particle blacks. The FKM data suggests this same interaction is present in the amine cured FKM I and FKM II. Results are depicted graphically in Figure 6.
A B C
% Vol Δ
FKM-I
A
FKM-II
A
2 phr cure B
3 phr cure B
30 phr N C
70 phr N C #1 1 1 1 29.5% 29.5% 29.5% 29.5% #2 1 2 2 22.5% 22.5% 22.5% 22.5% #3 2 1 2 26.0% 26.0% 26.0% 26.0% #4 2 2 1 33.1% 33.1% 33.1% 33.1% Total 0.520 0.591 0.555 0.556 0.626 0. Average 0.260 0.295 0.277 0.278 0.313 0. Table 14 - % Volume Change of amine cure FKM elastomers in CO (^2)
A B C
% Vol Δ
FKM-I
A
FKM-II
A
2 phr cure B
3 phr cure B
30 phr N C
70 phr N C #1 1 1 1 19.1% 19.1% 19.1% 19.1% #2 1 2 2 15.8% 15.8% 15.8% 15.8% #3 2 1 2 15.8% 15.8% 15.8% 15.8% #4 2 2 1 33.1% 33.1% 33.1% 33.1% Total 0.349 0.489 0.349 0.489 0.522 0. Average 0.174 0.244 0.174 0.244 0.261 0. Table 15 - % Volume Change 2 minutes subsequent to RGD
Figure 6- Factor/Level effects on swelling of amine cure FKM in CO2.
Aflas ETP
FKM
III
2phr cure
3phr cure
4phr cure N330 N550 N
10phr black
30phr black
50phr black A1 A2 A3 B1 B2 B3 C1 C2 C3 D1 D2 D
Total 0.75 1.07 0.96 0.99 0.79 0.99 0.99 0.79 1.00 1.14 0.82 0. Avg 0.25 0.36 0.32 0.33 0.26 0.33 0.33 0.26 0.33 0.38 0.27 0. Table 18 - % Volume Change after2 minutes subsequent to RGD
Aflas A
ETP
A
FKM
III
A
2phr cure B
3phr cure B
4phr cure B
N
C
N
C
N
C
10phr black D
30phr black D
phr black D
Total 2.05 0.42 0.41 1.45 1.08 0.34 1.32 0.95 0.61 1.65 0.98 0. Avg 0.68 0.14 0.14 0.48 0.36 0.11 0.44 0.32 0.20 0.55 0.33 0. Table 19 - % Volume Change 2 minutes subsequent to RGD
RESULTS AND DISCUSSION:
Figure 8 offers a graphic presentation of the data. The three peroxide-TAIC cured polymers exhibited mostly the same swelling under pressure. Increasing cure phr produced no appreciable trend in volume change. Particle size had no effect on swelling under pressure. Finally, there is a trend of decreasing swelling as the phr of particle black increased. Contrary to the prior two experiments, an increase in peroxide corresponds to a decrease in swelling subsequent to RGD. It should be noted that this group of polymers contains a coagent (i.e. TAIC) whereas the preceding two groups bond through unsaturated sites in the polymer backbone. The constitution of their respective crosslink networks is substantially different. Contrary to the prior two experiments, an increase in particle size corresponded with a reduction in swelling subsequent to RGD. This is a remarkable but suspect observation given the premise of increased surface area and associated increased strength with smaller particle size black.
It is probable that the effects of particle black loading and peroxide loading are overshadowing this factor. Additional study in this regard would be warranted. Similar to the results of the FKM study, increasing the phr loading of particle black decreased swelling subsequent to RGD.
Figure 8- Factor/Level effects on peroxide - TAIC cured fluoroelastomer 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. Figure 8 plots the percent change to be added or subtracted to calculated volume changes in the event of 1mm error in observing the length of a specimen. The error associated with ±1 mm change in length is calculated per Equations 4 and 5.
Eq.4 : +Tolerance = +tol = [(1 + ΔLen% + .01)^3 – 1] /1 -- [(1 + ΔLen%) 3 – 1] / Eq.5 : -Tolerance = -tol = [(1 + ΔLen%) 3 – 1] /1 – [(1 + ΔLen% - .01)^3 – 1] /
In evaluating data, the reader may want to consider volume change within a range rather than as a single point.
SUMMARY
- Choice of Elastomer is the primary determinant of swelling in CO 2. a) EPDM possesses the least amount of swelling under pressurized CO 2. Particle black reinforced EPDM exhibits the least amount of swelling subsequent to RGD. b) ETP and FKM Type 3 gum possess the greatest amount of swelling under pressurized CO 2. c) FKM Type 1 gum possesses the least amount of swell subsequent to RGD and exhibits the most rapid out-gassing subsequent to RGD.
- Carbon black is a significant if not overshadowing factor in swelling subsequent to RGD. a) Increasing PHR of carbon black decreases swelling under CO 2 pressure and post RGD. b) Carbon black particle size has no effect on swelling under pressure. c) Decreasing particle size reduces swelling subsequent to RGD (Note conflicting data in Experiment 3).
- Curative loading has an effect on post RGD swelling and can interact with carbon black reinforcement. a) Increasing PHR of curative has no effect on swelling under pressure for all the elastomers studied. b) For the peroxide cure and amine cure compounds not utilizing a coagent (i.e. HNBR, FKM I, FKM II), there is an interaction between carbon black particle size and curative loading on swelling subsequent to RGD. Specifically, as studied in HNBR, increasing curative while using N990 black coincides with an increase in swelling subsequent to RGD while increasing curative with N330 reduces swelling subsequent to RGD. c) Increasing PHR of curative in peroxide TAIC cured fluoroelastomers has no effect on swelling under pressure but coincides with a reduction in swelling subsequent to RGD.
ACKNOWLEDGEMENTS
The author would like to thank Kris Solaas for his diligent work in Seals Eastern’s laboratory: both being the sine qua non of this paper. Additional gratitude is extended to Harold Cook for his DMRT work and Dan Hertz, Jr for his prior works and “hands-off” guidance.
TRADEMARKS
Aflas® is the registered trademark of Asahi Glass Company Viton® and Diak® are the registered trademarks of DuPont Performance Elastomers Zetpol® is the registered trademark of Zeon Chemicals Technoflon® is the registered trademark of Solvay Solexis. Royalene® is the registered trademark of Lion Copolymer Geismar, LLC
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 (77 th (^) 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, (4 th
Printing, 1988). viii (^) Hertz, Jr., D.L., “Sealing Technology”, Rubber Products Manufacturing Technology, p.786, Marcel
Dekker, Inc. (1994).