mechanisms of high-ph and near-neutral-ph scc of ... mechanisms of high-ph and near-neutral-ph scc...

In ternational P ipeline Conference — V o lu m e 1A SM E 1996

MECHANISMS OF HIGH-pH AND NEAR-NEUTRAL-pH SCC OF UNDERGROUND PIPELINES

John A. Beavers and Brent A. HarleCC Technologies 6141 Avery Road

Dublin, Ohio 43016

A BS TR A C TThis paper provides an overview of mechanisms for high-pH and

near-neutral-pH stress corrosion cracking of underground pipelines. Characteristics and historical information on both forms of cracking are discussed. This information is then used to support proposed mechanisms for crack initiation and growth.

IN TR O D U C TIO NThe first incident of stress corrosion cracking (SCC) on natural

gas pipelines occurred in the mid 1960’s and there have been hundreds of failures since that time (Wenk, 1974). Most of the early failures were intergranular in nature, whereas, many of the recent failures that have occurred in Canada are transgranular (Justice and Mackenzie, 1988). It is now recognized that there are at least two forms of external SCC on underground pipelines. The intergranular form is referred to as high-pH or classical SCC while the transgranular form is referred to as low-pH, non-classical, or near-neutral-pH SCC. Further details of these two forms of cracking, and possible mechanisms for initiation and propagation are given below.

C HA R A C TE R ISTIC STable 1 summarizes the characteristics of the two forms of SCC.

These data show that both occur in patches with hundreds of cracks on the pipe surface that are generally longitudinal in direction, and localized to a small area of pipe on a joint. These cracks link up to form long shallow flaws, that can lead to ruptures. In both cases, the fracture faces are usually covered with black magnetite or iron carbonate films. However, there are a number of differences in the two forms of cracking. The cracks associated with near-neutral-pH SCC are generally wide and transgranular in nature with evidence of corrosion of the crack walls. High-pH stress corrosion cracks are generally intergranular and are usually sharp with little evidence of corrosion of the crack walls. In many cases, the near-neutral-pH cracking is associated with the long-seams, where tenting of the tape

coating has led to a build-up of iron carbonate paste on the pipe surface. On the other hand, high-pH stress corrosion cracks are usually found in the body of the pipe. Pitting corrosion and CO-, attack also are occasionally associated with near-neutral-pH SCC sites while high-pH stress corrosion cracks usually are not associated with significant corrosion.

CO NTRO LLING FACTORSFor SCC to occur on an engineering structure, three conditions

must be met simultaneously; a specific crack-promoting environment must be present, the metallurgy of the material must be susceptible to SCC, and tensile stresses above some threshold value must be present. Each of these factors is discussed below for high-pH and near-neutral-pH SCC of natural gas pipelines.

Hiqh-pH SCCEnvironm ental Factors. When SCC was first observed on

natural gas pipelines, several chemical species were considered as possible causative agents, including nitrates, phosphates, and caustic. All of these species are passivating agents and can cause SCC of carbon steels. However, pH values high enough to cause caustic cracking were never detected in electrolyte samples taken from beneath coatings at failure sites and there was no evidence for concentration of nitrates or phosphates above the low levels found in groundwaters. In the early 1970’s, it was discovered that a concentrated carbonate-bicarbonate solution could also cause SCC of pipeline steels (Sutcliffe et al., 1972) and evidence for concentration of carbonates at the pipe surface was found in a limited number of cases (Wenk, 1974). Table 2 is a summary of the environmental analyses.

It was concluded that a concentrated carbonate-bicarbonate solution was responsible for the SCC observed on pipelines and a plausible mechanism for the development of the environment was developed (Wenk, 1974). Essentially, the cathodic protection (CP) system causes a pH increase at the pipe surface. Carbon dioxide,

Copyright © 1996 by ASME

IPC1996-1860

Downloaded From: https://proceedings.asmedigitalcollection.asme.org/pdfaccess.ashx?url=/data/conferences/asmep/89953/ on 07/14/2018 Terms of Use: http://www.asme.org/about-asme/terms-of-use

which is present in the soil, is rapidly absorbed by the high-pH solution to generate the concentrated carbonate-bicarbonate (C 03 - HCOj) environment. This environment is simulated in the laboratory using a IN NaHC03 - IN Na2C 0 3 solution, which has a pH of about 9.3.

In spite of the fact that its selection was based on very limited field measurements, the concentrated carbonate-bicarbonate environment was readily accepted by the pipeline industry and the scientific community as the causative agent for SCC of natural gas pipelines. This was the case primarily because the characteristics of the SCC generated on line pipe samples in the laboratory in this environment were consistent with most of the field observations at that time. These include the intergranular mode of cracking, the temperature, potential, and stress dépendance of cracking, the initiation of multiple cracks on pipeline samples, and the association of the cracking with a black magnetite film.

Stress corrosion cracking of pipeline steels in the high-pH environment occurs over a very limited potential range that is about 100 mV wide and is centered around about -722 mV Cu/CuS04 at 75°C and moves in the positive (noble) direction with decreasing temperature (Koch et al., 1985). This potential range is associated with the active-passive transition in a potentiodynamic polarization curve, as shown in Figure 1 (Parkins, 1974). Stress corrosion cracking in the high-pH environment also occurs over a limited pH range, centered around about pH 9, as shown in Figure 2. At pH values above ten, the solution is primarily composed of carbonate and the passive film characteristics and electrochemistry are not conducive to SCC. At pH values below eight, the solution is dominated by the bicarbonate ion and cracking does not occur unless cations such as NH4+ are present. The dominant soluble cation found in high-pH electrolytes is Na+.

The limited potential and pH ranges over which SCC occurs in the high-pH environment provides an explanation for the infrequent occurrence high-pH SCC on most pipeline systems. These dependencies also suggest that seasonal fluctuations are important in the cracking process. In order for the concentrated high-pH environment to develop, significant current flow to the pipe is required to generate the elevated pH environment required to absorb the C 0 2. On the other hand, the potential range for SCC lies between the native potential of most pipelines, and adequate protection, -850 mV (Cu/CuS04). This paradox can be resolved if one assumes that the environment and potential vary on a seasonal basis, with SCC occurring only during periods of the year where adequate CP is not achieved.

M etallurg ical Factors. In the early 1970’s, an extensive field survey was performed on SCC of natural gas pipelines, which was presented in the Fifth Symposium on Line Pipe Research (Wenk, 1974). The survey attempted to correlate the occurrence of SCC with metallurgical, environmental, and operating parameters. SCC was found to occur on pipelines having a variety of diameters, wall thicknesses, pipe grades, compositions, manufacturers, and joining techniques. The range of compositions in which SCC was found covered the range of almost all line pipe in use at the time.

Subsequently, the Pipeline Research Committee (PRC) of A.G.A. funded a number of programs to investigate the relationship between SCC susceptibility and metallurgical variables in an attempt to

identify cracking resistant steels for future construction. The program met with little success and practicable steel compositions or heat treatments that resist SCC were not identified (Barlow, 1979), (Beavers and Parkins, 1986). It was found that major alloy additions, >1%, of some elements could increase SCC resistance, but such additions are prohibitively expensive. Differences in the SCC susceptibility of different line pipe steels, as measured by differences in the threshold stress for crack initiation, were found in the studies, but the root cause for these differences was not identified.

M echanical Factors. In order for stress corrosion cracks to propagate in a material, tensile stresses must be present at the crack tip. Tensile stresses also are required for SCC initiation, and cracking usually initiates at flaws on the surface where a stress concentration exists. Below some value of the tensile stress, referred to as the threshold stress (o,h), crack initiation does not occur.

The tensile stresses can be applied or residual in nature. In the case of SCC of pipelines, both have been shown to play a role in cracking. Many cases of longitudinal SCC have occurred in the body of the pipe where the primary stress is the hoop stress generated by the internal pressure. This contention is supported by the fact that SCC occurs much less frequently on heavy wall pipe, where hoop stresses are lower.

The threshold stress parameter, Gth, for the initiation of SCC on line pipe steels has been studied extensively in the laboratory in the high-pH cracking environment. It has been found that G(h for machined surfaces approaches the yield stress of most materials but Oqj is significantly reduced for an actual pipe surface by the presence of mill scale and pits on the surface (Beavers et al., 1987). For mill scaled surfaces, values measured in the laboratory are typically between 60% and 70% of specified minimum yield strength (SMYS) in the high-pH environment (Beavers et al„ 1985). There is a significant amount of scatter in the o lh data for a given steel, which may be the result of variation in surface residual stresses, surface condition (pitting or mill scale distribution), or in the thermomechanical treatment during fabrication.

Cyclic stressing also affects SCC. Laboratory studies in the high-pH environment have shown that the small pressure fluctuations associated with operation of a natural gas pipeline tend to exacerbate SCC (Beavers and Parkins, 1986). This effect has been attributed to the occurrence of cyclic creep at the crack tips, which facilitates rupture of the passive films.

Near-N eutral-pH SCCEnvironm ental Factors. The environmental aspects of SCC

appeared to be reasonably well understood until one pipeline company started experiencing SCC on their polyethylene tape and asphalt coated pipelines in the 1980’s. An extensive field investigation program showed that the occurrence of SCC correlated with near-neutral-pH (pH <8) dilute C 0 2 containing electrolytes and that cracking was not observed where high-pH electrolytes were detected (Justice and Mackenzie, 1988).

There is significantly less laboratory information on the near- neutral-pH form of cracking. Initial attempts to reproduce the near- neutral-pH SCC in the laboratory met with limited success. In these initial studies, the test techniques used were similar to those that had


been successfully used or developed for the evaluation of high-pH SCC. For example, negligible cracking has been observed in tapered tensile tests performed in near-neutral-pH environments under cyclic load conditions that simulate pipeline operating conditions. Cracking has been reported in slow strain rate (SSR) tests but the cracks generally are confined to the necked regions of the specimens where extreme mechanical loading conditions are encountered, see Figure 3 (Beavers and Thompson, 1988). Other related studies have been somewhat more successful, although they shed only limited light on the cracking mechanism. For example, electrochemical studies of line pipe steels clearly demonstrate that the near-neutral-pH environments do not promote passivation and do not exhibit active-passive transitions, which are commonly observed with the high-pH cracking environment, see Figure 4 (Beavers and Thompson, 1988).

More recently, significant advances have been made in reproducing near-neutral-pH SCC, although additional research is needed to fully understand the mechanism of cracking. Parkins (1994) reported that environmentally induced cracks can be initiated in the surfaces of line pipe steels in the near-neutral-pH environments under cyclic loading conditions. The cracks initiate most readily on natural, corroded, mill scaled surfaces. However, the cyclic-load conditions (stress ranges, frequencies and number of cycles) required to obtain crack initiation in a reasonable length of time are generally more severe than those encountered in the field. Harle et al., (1994) produced transgranular crack propagation in a near-neutral-pH environment using precracked compact type specimens of line pipe steels. The fractography of the specimens was very similar to that observed in the field and the crack propagation rates were consistent with field observations. A typical photomicrograph from a laboratory test is shown in Figure 5.

The results of these studies and field observations all indicate that the environment responsible for near-neutral-pH cracking is dilute groundwater containing dissolved C 0 2. As is the case for high-pH SCC, the source of the C 0 2 is decay of organic matter in the soil. However, in the case of near-neutral-pH SCC, cathodic protection current cannot flow to the pipe either because of shielding of the current by coatings such as tapes, high resistance soils, or a poor cathodic protection system design. As a result, a dilute carbonic acid solution develops.

M etallurg ical Factors. Near-neutral-pH SCC has been observed on a number of pipelines having different grades and yield strengths. They do all appear to be carbon-manganese steels with a ferrite/pearlite microstructure. Near-neutral-pH SCC failures with the newer micro-alloyed very low carbon steels have not been reported. This may simply reflect the fact that these are newer steels (have not been in service as long) and have better coatings (i.e., no tapes). In any case, there are no laboratory or field data indicating that one particular microstructure or grade of steel is significantly more resistant to near-neutral-pH SCC.

Results of constant displacement rate tests performed at CC Technologies (Harle et al., 1995) on precracked compact type specimens have clearly demonstrated that the coarse grain heat affected zone (CGHAZ) microstructure adjacent to the weld, in an X-65 steel, is significantly more susceptible to cracking than the base material in the near-neutral-pH environment. Average crack

velocities are about a factor of two higher in the CGHAZ than in the base material, as shown in Figure 6. Higher crack velocities also have been observed at the bond line of the electric resistance weld than in the base metal in an X-52 line pipe steel.

M echanical Factors. The mechanical conditions for near- neutral-pH SCC appear to be more restrictive than those for high-pH SCC. In SSR tests, cracking is generally only observed within the necked region of the specimen where plastic deformation and true stresses are very high, as previously described. In constant displacement rate testing (Harle et al., 1995), cracking is only observed where dynamic loading conditions are present, as shown in Figure 7. No evidence of crack growth has been found under constant load or constant displacement conditions.

A threshold stress for crack initiation in the near-neutral-pH environment has not been established. A strong relationship has been established between the occurrence of SCC on one pipeline system and class location. On that system, SCC has led to about twenty failures in Class I locations, operating at 62% to 77% of SMYS, while only minor cracking and no failures have occurred on Class II or in locations, which operate at 64% and 53% of SMYS, respectively (Anonymous, 1992). This relationship suggests that either the threshold stress for SCC initiation is above about 65% of SMYS or that crack growth rates at these stresses are too low to cause significant problems.

M EC HA NISTIC CO N SIDER ATIO NSA number of mechanisms has been proposed for SCC of metals

(Jones and Ricker, 1992). Three well accepted mechanisms, at the present time, are the anodic dissolution mechanism, the film induced cleavage (FIC) mechanism, and hydrogen related mechanisms. Normally, a distinction is not made between initiation and propagation mechanisms. In the anodic dissolution mechanism, the applied stress causes plastic deformation and rupture of films on the metal surface. The rupture of the film exposes bare metal to the environment that undergoes rapid anodic dissolution. The crack actually propagates by dissolution but the passive film prevents corrosion of the crack walls so a crack geometry is maintained.

Very specific conditions, with respect to film formation behavior, film properties, metal creep rate, and corrosion rate of the bare metal surface must be present for this mechanism to operate. Where film formation rates are slow or creep rates are high, one would expect crack tip blunting. At the other extreme (high passivation rates and/or low creep rates) one would expect the passive film to maintain coverage of the metal surface, preventing crack advance. Within the ranges of the properties where cracking occurs, these properties affect behavior. Where creep rates are high and repassivation rates are low, one would expect a continuous process. Where creep rates are low and repassivation rates are high, one would expect a more discontinuous process. Intergranular SCC failures are frequently attributed to this mechanism since impurities at the grain boundaries have pronounced effects on repassivation and creep behavior.

Alloy-environment systems where the anodic dissolution mechanism is thought to operate include carbon steel and low alloy steels in caustic, nitrates, phosphates, and in concentrated carbonate- bicarbonate. In these systems, cracking is intergranular and the steel


exhibits active-passive behavior. Anodic dissolution rates of bare steel are high in these environments and maximum cracking velocities can be accurately predicted based on these rates.

The information available on near-neutral-pH SCC propagation is not consistent with this mechanism. The cracking in the near- neutral-pH environment is transgranular whereas the anodic dissolution mechanism is most commonly associated with intergranular cracking. As described above, line pipe steels do not exhibit passivation in the near-neutral-pH environments. The cyclic potentiodynamic polarization behavior is characteristic of active corrosion. The maximum anodic current densities also are low for the observed cracking velocities. With the anodic dissolution mechanism, one can estimate the maximum crack velocity based on Faraday’s law:

da/dt = ia*M/z*F*d

where da/dt is the crack velocity, ia is the anodic current density, M is the atomic weight, z is the oxidation state of the solvated species, d is the density, and F is Faraday’s constant. In research performed at CC Technologies (Beavers and Thompson, 1988) typical anodic current densities are of the order of 100 pA/cm2 for X-65 line pipe steel in carbon dioxide containing near-neutral-pH environments. This current density corresponds to a crack velocity of about 4 x 10-8 mm/s. In recent research, Harle et al., (1994) observed that crack velocities were at least two orders of magnitude higher in constant displacement rate testing of precracked specimens of line pipe steels in near-neutral-pH environments.

The lack of specificity of the conditions for crack propagation in the near-neutral-pH environment also is not consistent with an anodic dissolution mechanism for crack propagation. For example, near-neutral-pH cracking appears to be relatively less sensitive to potential than high- pH SCC (Beavers and Thompson, 1988). The severity of near-neutral-pH SCC does appear to increase with increasing negative potential (Parkins, 1994) in bulk solutions. In more recent work (Harle et al., 1994), evidence of environmental cracking also has been observed in deionized water in which carbon dioxide was bubbled. In contrast, the potential range for cracking in the concentrated carbonate-bicarbonate environment (high-pH environment) is very narrow as previously described. In the high- pH environment, both carbonate and bicarbonate species must be present and cracking occurs over a limited range of pH and potential as shown in Figure 2.

Finally, the results of recent cyclic load tests and interrupted constant displacement rate tests, performed for TCPL, have shown that near-neutral-pH stress corrosion crack propagation does not occur under constant load or constant displacement conditions (Harle et al., 1995). In contrast, alloy-environment systems where the anodic dissolution mechanism is thought to operate all exhibit crack growth under constant load or constant displacement conditions.

The characteristics of near-neutral-pH SCC also do not appear to be consistent with a film-induced cleavage (FIC) mechanism. This mechanism requires the presence of a brittle coherent oxide film, de-alloyed layer, or some other brittle film or layer. There is no evidence that a film or layer with the required properties is present. One could argue that the iron carbonate/ magnetite corrosion

products that are present on the fracture faces are responsible for FIC of the metal but fundamental studies to investigate this hypothesis have not been performed.

The most plausible mechanism for crack propagation in the near- neutral-pH environment is a hydrogen related mechanism. It is well established that hydrogen can cause several types of damage to steels including H2S cracking, delayed hydrogen embrittlement of high strength steels, and hydrogen effects on the plasticity of low strength steels (NACE - 1977). The mechanisms for these different forms of hydrogen embrittlement all involve the entry of hydrogen into the lattice of the metal and its concentration at crack tips and other stress raisers because of the high triaxial stresses at those locations. A number of effects of hydrogen at the crack tip have been postulated, including reduction in the cohesive strength of the lattice, the formation of unstable brittle hydrides, the formation of strain induced martensite, the pinning of dislocations and hydrogen enhanced localized plasticity (HELP). Cracking in H2S is normally associated with sulfides since they poison the recombination of hydrogen atoms on the surface of the metal and, in so doing, enhance the entry of hydrogen into the metal.

The most likely source of hydrogen in near-neutral-pH SCC is carbonic acid, formed by the dissolution of carbon dioxide in the groundwater (Schmidt, 1983):

C 0 2 + H20 = H2C 0 3 (sol)

The carbonic acid can further react with water to produce a hydronium ion and a bicarbonate ion:

H2C 0 3 (sol) + h 2o = + h 3o + + h c o 3~

The hydronium ion is then reduced to generate hydrogen at the metal surface. Schmidt also identified a second mechanism for hydrogen generation involving adsorbed species on the metal surface. Corrosion potential measurements (Harle et al„ 1995) are consistent with this mechanism for hydrogen generation. Typical corrosion potentials for X-65 Steel in a dilute groundwater solution (NS4) containing 5% C 02 (pH of about 6.75) is -725 mV SCE (-485 mV SHE). See Table 3 for NS4 composition. On a Pourbaix diagram for C 02, this pH and potential lie slightly below the hydrogen reduction line.

Crack propagation due to hydrogen mechanisms can be continuous or discontinuous on. a short term time scale (seconds to minutes). Over longer periods, a constant cracking velocity would be expected if the mechanical and environmental parameters controlling cracking are constant. There is strong evidence from field examinations of near-neutral-pH SCC that colonies of cracks become dormant. In some dormant colonies, crack tips are blunted while the crack walls have experienced significant corrosion, widening the cracks. The most likely explanation for this behavior is that the controlling parameters for crack growth changed at some time. The blunted crack tips are probably the result of either hydrostatic testing or a pressure spike. It is also possible that environmental changes, such as seasonal reductions in the CO, levels, cause the cracks to become dormant.

The above discussion has not distinguished between SCC initiation and SCC propagation. For many of the systems where the


anodic dissolution mechanism is thought to operate, it is generally considered that both initiation and propagation occur by the same mechanism. There are some exceptions, such as situations where stress corrosion cracks initiate at pits. It is probably more common for hydrogen related growth mechanisms to work in concert with other initiation mechanisms.

In the case of near-neutral-pH SCC, other possible initiation mechanisms include fatigue, pitting, or even high-pH SCC. A number of the features of near-neutral-pH SCC suggests that these other mechanisms may be involved with the initiation process. Near-neutral-pH SCC is frequently associated with discontinuous features on pipelines surfaces such as pits, scabs, scrapes, dents, and welds. Where near-neutral-pH SCC occurs in the pipe body away from these defects, the appearance of the crack colonies is nearly identical with high-pH colonies, suggesting that they may have similar origins in some cases. In the laboratory, it is very difficult to initiate or propagate colonies of near-neutral-pH cracks in the absence of stress concentrators and fatigue loading, such as that used by Parkins (1994).

C O N CLUSIO NSThe results of this analysis indicate that at least two mechanisms

of cracking are responsible for the SCC observed on operating natural gas pipelines. The first type is the high-pH SCC which was first reported in the mid-1960’s. Identifying characteristics of this type of cracking include the intergranular crack path and the absence of significant corrosion associated with the failure. The underlying mechanism for crack propagation with high-pH SCC is most likely an anodic dissolution mechanism. The environment responsible for high-pH SCC is a concentrated carbonate- bicarbonate electrolyte that forms at the pipe surface as a result of C 0 2 in the soil and application of cathodic protection on the pipe.

. The second mechanism of cracking is the near-neutral-pH SCC first reported in the 1980’s. Identifying characteristics of this type of cracking include the transgranular crack path with significant corrosion of the crack faces. Only recently have key aspects of this type of SCC been reproduced in the laboratory. The results of these studies suggest that the mechanism of initiation and propagation may be different in some cases. Initiation may occur by a corrosion fatigue mechanism. It is possible, although less likely, that the cracks initiate by an anodic dissolution mechanism. Propagation of near-neutral-pH stress corrosion cracks probably occurs by a hydrogen embrittlement mechanism and cyclic loading is important in the crack propagation process. The environment responsible for near-neutral-pH SCC is a dilute carbonic acid solution that forms at the pipe surface when C 0 2 in the soil dissolves in groundwater in the absence of cathodic protection on the pipe.

R EFERENCESAnonymous, 1992, “TransCanada PipeLines Responses to

National Energy Board Information Requests Regarding TransCanada Pipelines Maintenance Program and Stress Corrosion Cracking,” National Energy Board Inquiry Concerning Stress Corrosion Cracking, Proceeding MHW-1-92, November-December 1992.

Barlo, T. J., 1979, “Stress-Corrosion Cracking SteelSusceptibility,” 6th Symposium on Line Pipe Research, American Gas Association, Inc., Catalog No. L30175, p. P-1.

Beavers, J. A., Parkins, R. N., Berry, W. E., 1985, “Test Method for Defining Susceptibility of Pipeline Steels to Stress Corrosion Cracking,” NG-18 Report Number 146 Catalog No. L-51484, June 1985.

Beavers, J. A. and Parkins, R. N., 1986, “Recent Advances in Understanding The Factors Affecting SCC of Pipe Steels.” 7th Symposium on Line Pipe Research. American Gas Association. Inc., Catalog No. L51595, p. 25-1.

Beavers, J. A., Christman, T. K., Parkins, R. N., 1987, “Some Effects of Surface Condition on the Stress Corrosion Cracking of Line Pipe Steel,” CORROSION/87, Paper No. 178.

Beavers, J. A. and Thompson, N. G., 1988, Unpublished Research Performed at CC Technologies for TransCanada Pipelines Limited.

Harle, B. A., Beavers, J. A., and Jaske, C. E., 1994, “Low-pH Stress Corrosion Cracking of Natural Gas Pipelines,” Corrosion/94, Paper Number 242.

Harle, B. A., Beavers, J. A. and Jaske, C. E., 1995, “Mechanical and Metallurgical Effects on Low-pH Stress Corrosion Cracking of Natural Gas Pipelines,” Corrosion/95, Paper 646.

Jones, R. H. and Ricker, R. E., 1992, “Mechanisms of Stress- Corrosion Cracking,“ Stress Corrosion Cracking, Ed. R. H. Jones, ASM International.

Justice, J. T. and Mackenzie, J. D., 1988, "Progress in the Control of Stress Con-osion Cracking in a 914 mm O.D. Gas Transmission Pipeline,” Proceedings of the NG-18/EPRG Seventh Biennial Joint Technical Meeting on Line Pipe Research, Pipeline Research Committee of the American Gas Association, Paper No. 28.

Koch, G. H., Beavers, J. A. and Berry, W. E., 1985, “Effect of Temperature on Stress-Corrosion Cracking of Precracked Pipeline Steels,” NG-18 Report No. 148, American Gas Association, Inc., Catalog No. L51491.

NACE, 1977, “Stress Corrosion Cracking and Hydrogen Embrittlement of Iron Base Alloys." Proceedings of a Conference Held in Firminy, France, June 1973, NACE-5, 1977.

Parkins, R. N., 1974, “The Controlling Parameters In Stress Corrosion Cracking,” 5th Symposium on Line Pipe Research, American Gas Association, Inc., Catalog No. L30174, p. U-l.


Parkins, R. N., 1994, “Transgranular Stress Corrosion Cracking of High-Pressure Pipelines in Contact with Solutions of Near Neutral pH,” Corrosion, Vol. 50, No. 5, p. 384.

Schmidt, G., 1983, “Fundamental Aspects of C 0 2 Corrosion,” Corrosion/83, Paper No. 43.

Sutcliffe, J. M., Fessier, R. R., Boyd, W. K. and Parkins, R. N., 1972, “Stress Corrosion Cracking of Carbon Steel in Carbonate Solutions,” Corrosion, Vol 28, p. 313.

Wenk, R. L., 1974, “Field Investigation of Stress Corrosion Cracking,” 5th Symposium on Line Pipe Research, American Gas Association, Inc., Catalog No. L30174, p. T-l.

Table 1. Com parison of H igh-pH and Near-Neutral-pH SCC.

CHARACTERISTIC NEAR-NEUTRAL-pH SCC HIGH-pH SCC

Electrolyte pH 6.5 to 7.5 9 to 10

Fracture Mode Transgranular Intergranular

Cracking Orientation Longitudinal Longitudinal

Numerous Surface Cracks Yes Yes

Linking of Crack Yes Yes

Patches of Cracks Yes Yes

Corrosion of Crack Faces Yes No

Corrosion of Pipe Sometimes Usually Not

Association with Iron Carbonate and Magnetite Films Yes Yes

P/S Potential, CCS Native Potential -722 mV

Width of Potential Range Probably > 100 mV Narrow (< 100 mV)

Temperature Dependence Not Established Arrhenius Behavior

T able 2. Com position of L iquids Found Under Coatings Near T he Locations O f Stress- Corrosion Cracks (W enk, 1974).

AMOUNT IN SOLUTION, PERCENT

STATE PH c o 3 h c o 3 OH Cl n o 3

Alabama 9.7 0.5 0.5 — — ---

Arizona 12.3 1.0 N<*> 0.1 0.01 0.007

Mississippi 10 1.4 0.5 N 0.12 0.004

Mississippi 10 0.9 0.8 N 0.12 <0.01

Mississippi 9.6 0.5 0.6 N N —

North Carolina 10.5 0.7 0.4 N — N


POTE

NTI

AL,

mV

(SH

E)

Table 3. Com positions of S im ulated Near-Neutral-pH Cracking Electrolytes.

COMPOUNDS COMPOSITION, grams/liter

NS3 NS4

KC1 0.037 0.122

NaHC03 0.559 0.483

CaCl2-2H20 0.008 0.181

MgS04»7H20 0.089 0.131

CURRENT DENSITY, u A /c m 2

Figure 1. P otentiodynam ic Polarization Curves S how ing the Potential Range O ver W hich SCC O ccurs In Concentrated Carbonate- Bicarbonate Solution at 194°F (90°C) (after Parkins, 1974).


Figure 2. Potential - pH Region W here SCO is O bserved for Line P ipe Steels In Carbonate-B icarbonate Solutions at 167°F (75°C) (after Parkins, 1974).

Figure 3. Photom icrograph of S low Strain Rate Specim en of X-65 Line Pipe Steel Tested at a Strain Rate of 2.0 x 10~6 s '1 in NS3 E lectrolyte C ontaining Sparged 100% C 0 2 at Room Tem perature and a Potential of -6 5 0 mV SCE (Beavers and Thom pson, 19BB). See Table 3 fo r Com position of NS3 Electrolyte.


POTE

NTIA

L, m

V (S

CE)

Figure 4. Potentiodynam ic Polarization Curve for X-6S Line Pipe Steel at Room Tem perature in a Sim ulated Near-Neutral-pH Cracking E lectrolyte (NS3) Containing Sparged C 0 2; Scan Rate 0.17 mV/s (Beavers and Thom pson, 1988).

Figure 5. Scanning Electron M icroscope Photograph of Fracture Surface of Precracked C om pact Type Specim en of X-65 Line P ipe Steel Tested Under Constant D isplacem ent Rate (2.54 x 10“6 m m /s) in NS4 Electrolyte Containing Sparged 5% C 0 2 - 95% N2 at 35°C Under Freely Corroding Conditions (1000X M agnification) (H arle et al., 1994). See Table 3 for Com position of NS4 Electrolyte.


Comparisons of X-65 Crack Velocity vs Total J Curves At 5.0 E-8 ln/sec

Figure 6. C rack Velocity as a Function of the J Integral P aram eter fo r Four Precracked Com pact Type Specim ens of X-65 Line Pipe Steel Tested at a C onstant D isplacem ent Rate of 1.27 x 10“® mm/s in NS4 E lectrolyte Containing Sparged 5% C 0 2 - 95% Nz a t 35°C Under Freely C orroding C onditions. Specim ens W -4 and W -7 W ere M achined from the C G HAZ and Specim en T-12 and T -14 w ere M achined from the Base Metal. Specim en T -14 was a Control, Tested in Dry N itrogen (H arle et al., 1995).

IWE OaeandC<WhO SO M IjGE«4 1JE*6 J jOW 2 S M 10E«6 ! * • » 40E«6

Figure 7. Load and Potential Drop as a Function of Test T im e For Precracked C om pact Type Specim en of X-65 Line Pipe Steel Tested at a C onstant D isplacem ent Rate of 1 .27 x 10"6 mm/s in NS4 Electrolyte Containing Sparged 5% C 0 2 - 95% N2 at 35°C Under Freely Corroding C onditions. Cross Head W as Stopped After 1 x 10® s (Harle et al., 1995).


mechanisms of high-ph and near-neutral-ph scc of ... mechanisms of high-ph and near-neutral-ph scc...

Documents