CC BY
29
УДК 620.194.22:622.691.4.053:544.6.018.45
Regularities of the near-neutral pH stress corrosion cracking of gas pipelines
I.V. Ryakhovskikh
Gazprom VNIIGAZ LLC, Bld. 1, Estate 15, Proyektiruemyy proezd no. 5537, Razvilka village, Leninskiy district, Moscow Region, 142717, Russian Federation E-mail: I_Ryakhovskikh@vniigaz.gazprom.ru
Abstract. The article considers stress corrosion cracking (SCC) prediction models for pipes steels describing main stages of the process, their rates and implementation conditions. A number of tests, namely: electrochemical, corrosion-mechanical, mechanical, operational life tests, X-ray tests of layer-by-layer texture and residual stresses, analysis of micro and dislocation structure, were carried out for X70-steel pipes manufactured in Germany and USSR, after these pipes had been operated within the gas pipelines for a long time. It was shown that the texture nonuniformity of steels and residual process stresses in pipes are the parameters characterizing resource for growth or slowdown of cracks at initial stages. The cyclic tests did not identify any indicators of fatigue growth and significant changes of dislocational structure in the areas close to the cracks -0,28 deep (where 8 is the pipe wall thickness). During static and low-amplitude cyclic loads in the test environment with pH = 5,5...7,0 the rate of crack growth accelerates with available component (sulfide, carbonate, or phosphate) stimulating the anodic dissolution.
It was found out that SCC-cracks not deeper than 0,28 were not dangerous for pipelines reliability; on exposure of corrosion environment the forecasted growth rates of such defects did not exceed 0,3.0,4 mm per year, apart from cracks located in the welded joints and along the weld-fusion line. Growing of the 0,28 deep SCC-cracks would stop without electrolyte.
Keywords:
trunk gas pipeline,
stress corrosion,
cyclic testing of steel
pipes,
defect,
crack,
texture nonuniformity of steel,
rate of crack growth, pH.
In countries with a long system of high-pressure underground gas pipelines (UGP) the problem of stress corrosion cracking (SCC) is a common issue [1-4]. SCC accidents have been detected in USA, Canada, Australia, Iran, China [3-6]. The SCC defects formation and propagation takes place on the surface of pipes that are in contact with the soil electrolyte due to band coating separation [1-3]. A lot of UGPs constructed in Russia in 1970-1980's from X70-grade steel pipes 04O"-56" mainly produced by Mannessmann (Germany) and Khartsyzk Pipe Plant (USSR, the territory of modern Ukraine) suffer from SCC [2]. According to the results of specialized microscopic studies of SCC-cracks [7] and composition of soil electrolyte [8] basing on the tests of Russian second-hand UGPs, the near-neutral-pH SCC is observed as a rule.
It is a common belief that SCC of pipeline steels is a multistage process involving the active or slow development of cracks under various scenarios depending on the least favorable combination of the determining groups of «metal-environment-stress» factors [2, 4, 8-10]. Fig. 1 illustrates a generalized form of a stage-by-stage kinetic diagram of SCC defects formation and propagation in pipe steels as time goes by, which contains five specific cross-sections.
In light of the model, SCC gets through a flow of stages that characterize the variation of propagation rate including slow-down and stagnation. At every step the SCC process is designated by the combined effect of various parameters. To predict SCC propagation, it is necessary to describe the parameters affecting the kinetics of the propagation or slowdown of cracks, as well as to establish the types of functional relation between RCG and these parameters.
Scenarios of SCC propagation in UGP
Fig. 2 shows three specific scenarios of SCC in UGP. They describe the rate growth variation and depth of the SCC defects on the back of statistical data of the Russian UGP. It was supposed that the corrosion conditions for the formation of all kinds of SCC were formed in the first 5-7 years after gas pipeline exploitation had begun, which was typical for band
О
О &
Time
Fig. 1. The rate of crack growth (RCG) time course [4, 11]: 0 - an incubation period (IP) preceding formation of cracks; 1 - an area of formation and instant growth of cracks (it decreases with increase of time); 2 - an area
where RCG is constant quite long; 3 - an area where RCG increases due to external or operational loads, as well as the interaction of cracks in the colonies; 4 - an area of rapid RCG increase and impending breakdown
field-applied coatings in the 70s - 80s of the last century.
Scenario of accelerated («anomalous»)
SCC (further - ABC). When investigating the causes of gas pipelines accidents after 7-13 years of their commissioning [2, 12], metallurgical defects were detected as a result of corrosion weakening - visible shape failure or pipe galling in location of SCC defects [12]. Above mentioned special features of pipes explain the degradation of bearing capacity of pipes and the accelerated dynamics of SCC according to the ABC scenario. This scenario assumes the implementation of 2 main stages: formation and accelerated growth of a colony of cracks or one extended crack along the longitudinal weld of pipe in the AB section (to a depth of about 0,35.. .0,45, where 5 is pipe wall thickness), and its irreversible destruction on BC section. Further crack growth (section BC) with a depth of 0,35 or more occurs with the participation of soil electrolyte under project loads, for example, seasonal soil movements or changes in operating pressure. The kinetics of formation and initial growth of these cracks hinges upon the parameters of the corrosive
Time
Fig. 2. Specific scenarios of SCC propagation taking into account statistical information about accidents
2
environment [9, 13], and if depths of 0,405 or more are reached is determined by alternating loads on the gas pipeline caused by seasonal soil movements and gas transport modes, which are adequately described by the Peris formula [14].
Scenario of «intensive» SCC development (further - AB'C'D'E'). This scenario explains the random failures of UGP. The most part of them occurred after 20-30 years of operation. The implementation of UGP SCC according to the AB C D E model implicates the huge colonies formation of random orientation cracks [15] in location of pitting corrosion, pits and other defects on the overloaded pipe surface. The formation of a crowd of cracks in these areas provides partial balancing of residual process-related stresses (section AB ), for example, caused by plastic deformation of steel at the crack tip. The kinetics of further propagation of SCC is determined by specificity of soil electrolyte, residual stresses level, the structural and texture features in the pipe material surface layer. In a less degree the kinetics depends on the operational piping loads. In addition propagation of cracks in a big time range (B C section) occurs at a constant speed to an estimated depth of 0,25.0,305, and a forming colony of cracks orients along the axis of a gas pipeline under the action of circular stress. Then, on the CD' section it is possible to significantly slow down the development of cracks in colonies (see fig. 2) depending on their mutual location, distribution of stress fields [9, 15], or etching of crack tips as a result of active corrosion processes. When a defect reaches a depth of 0,405, its further development proceeds as in the ABC scenario.
Scenario of «slow» SCC propagation (further -ABC'D"). The time-extendedAB'CD' E scenario explains the identification of a significant proportion (more than 90 % of the total number) of SCC defects with a depth of less than 0,105 by repair actions on UGP after 30 years of operation [1]. As follows from the growth of cracks formed similarly to the previous scenario, if they have reached depth of 0,055.0,075, according to simulation results, it is expected that residual stresses effect decreases at section AB [15]. Against the background of textural inhomogeneity of material it leads to partial or complete stop of defects [16, 17] (section B'C"). Part of the cracks under influence of soil electrolyte can marginally grow in length [11]. As a reaction a dense colony of crevices is formed on the pipe surface. In the
light of plastic deformation or etched by soil electrolyte the tips of these crevices are becoming blunt. This process promotes complete stop of their propagation [16, 17]. It can be assumed that crack growth at B C" section is determined by the rate of local anodic metal dissolution [18, 19]. In addition, as the crack grows, the level of stresses in the pipeline will have an increasing effect on rate of its propagation. The repeated growth on B C" section under the influence of soil electrolyte is noted for part of cracks in the least favorable cases [20, 21]. These cases explain the detection of SCC-cracks with a depth of more than 0,15.0,205 based on the results of repeated surveys of UGP. When a defect reaches this depth, its further propagation proceeds similarly to the AB C D E scenario. It is of interest to calculate and experimentally confirm the propagation of SCC defects in accordance with the proposed scenarios to study the UGP working capacity taking into consideration SCC and service life of more than 30 years by the AB 'C'D model. It's better to realize this model taking into account the average service life of Gazprom PJSC UGP. The AB C"D" takes into account the influence of electrochemical processes on the kinetics of SCC process, residual stresses in the pipe walls, and the influence of structural and texture inhomogeneity of steel. The listed parameters affect significantly formation, slowing or further propagation of the SCC, however they are not currently taken into account when predicting the resource and specifying the interdiagnostic period of the UGP, which are subject to SCC.
To build an empirical multiparameter model of SCC UGP for reliable prediction of the life-period of pipes with SCC defects, a study discussed below was aimed at establishing the influence of the quantitative characteristics of meso-and microstructures of pipe steels in solutions simulating a near-neutral ground electrolyte on the pipe resistance against corrosion-mechanical failure.
Experimental methods
Material and solutions. Complex experimental studies were carried out for second-hand 040"...56" pipes with 16,5.18,7 mm thick walls produced at the Mannesmann (further - 1W steel) and Khartsyzsk (further - 2W steel) pipe plants from the controlled rolled steel (X70 strength grade). Before these pipes had been applied for more than 25 years as parts of UGP.
The chemical composition of the steel (tab. 1) was determined by spectral analysis using a Spectro Lab S emission spectrometer. One can see that the chemical compositions of steels differ in micro-alloying additives, namely: vanadium, Nb and Mo are present in 1W steel. Another difference in composition is the high content of Ti in the 2W steel, and content of Si in the 1W steel. Actual mechanical properties of pipe steels complied with regulatory requirements.
Metallographic studies were performed by optical microscopy using a Carl Zeiss Axiovert 40 MAT, and by scanning electron microscopy (SEM) using a JEOL JSM-6610LV with an INCA Energy Feature XT energy dispersive X-ray (EDX) microanalysis system and INCAWave 500 spectrometer. They made
it possible to determine composition of non-metallic inclusions and the corrosion products in the cracks. Specimens for the structural studies were pickled in 4 % alcohol solution of HNO3.
All investigated pipes are characterized by ferritic-pearlite structure (fig. 3, see a, b). Ferrite of the 2W steel has more rounded grains, while 1W steel has a more fine-grained structure (especially pearlite grains), as well as pronounced axial segregation on the background of significant banding, which is typical for controlled rolled steels. These morphological features of the structure seem like a difference in the technological parameters of hot deformation of steels. There were differences revealed neither in defective, nor in defect-free steel areas.
Table 1
Chemical composition and mechanical properties of steels
Steel Composition, wt % Yield strength (M, MPa Tensile strength (ats), MPa
C Mn Si S P Cr Ni Cu V Nb Ti Mo
1W 0,093 1,57 0,41 0,007 0,021 0,12 0,04 0,06 0,07 0,038 0,022 0,031 490 590
2W 0,115 1,63 0,34 0,003 0,021 0,04 0,02 0,007 - - 0,07 - 480 620
Fig. 3. The structure of 2W (a, *500) and 1W (b, *500) steels; polyhedral ferrite (PF) in 2W (c, x15 000); cementite precipitates at the junctions of ferrite grains (2W) (d, X30000)
The quantitative characteristics of pearlite were evaluated, and the morphology of the pearlite component of the steel microstructure was analyzed. Samples were chosen both without SCC defects and with them 4 mm deep. The study was performed by transmission electron microscopy on a foil 0,2.0,5 mm thick. There was no appreciable differences in the structural state of samples, including the morphology of pearlite. The structure of the samples is based on PF (see fig. 3c), which is characterized by smooth boundaries and almost complete absence of blocks. Occasionally, cementite precipitates are observed at the junctions of ferrite grains (see fig. 3d).
The NS4 (pH 7,0), C2 (pH 6,3) and NOWATV (pH 7,1) solutions have been chosen as the corrosive media modeling the soil electrolytes [4]. To eliminate changes in the pH of the solutions during the experiments, investigators added borate buffer to them, which stabilized the pH and did not affect the rate of electrochemical reactions [19]. In a number of experiments citrate buffer was also used [19]. In addition, additives were added to the background solutions, which were components of the natural soil electrolyte (Na2S, NaHCO3, NaNO3), and Mg(H2PO4)2, KI and metal corrosion inhibitor were used as model additives.
Procedures for investigating kinetics of SCC:
• stage «0». During the experiment the samples have been holded in the electrolyte, the processes occurring on the surface of the metallographic section have been observed using the Carton SPZT50 optical microscope (magnification *200) [22]. Reserchers have recorded an enlarged image of the surface area, using an Amoyca AC-300 digital CMOS video camera connected to the eyepiece via an adapter during the entire experiment. Data from the camera was transferred to a personal computer and processed in the Corel PHOTO-PAINT X3 graphic editor. All samples were polished on a diamond paste with a grain size of 7 to 0,5 ^m, then washed in an ultrasonic bath in a mixture of C2H5OH : C7H8 = 1:1 for 25 min before testing;
• stage «1». Researchers have carried out an experimental assessment of the rate of propagation of microcracks from local corrosion defects (LCD) on surface of steel on samples of 120*25*5 mm with variable loading in the NS4 solution in order to decrease the time of an experiment compared with static tests [9]. On one part of the samples, pits with a diameter and a depth of about 300 ^m
(aspect ratio 1:1) were formed by anodic etching in an alkaline solution [22]. In other samples pitting simulators were created in the form of drilled holes with a diameter of 600 ^m and a depth of 1200 ^m (aspect ratio 1:2), a diameter of 900 ^m and a depth of 2250 ^m (aspect ratio 1:2,5). Corrosion-mechanical tests were carried out on the installation, creating a variable stress by 4-point bending of the sample in a corrosive environment. The imposition of a variable (0,15 Hz) load was obtained due to a rotating eccentric mounted on the shaft of an electric motor with a gearbox. The tangential residual macrostresses acting in different layers along the wall thickness of pipes were measured in a section perpendicular to the axis of the pipe. X-ray lines were recorded on a Bruker D8 DISCOVER diffractometer. The textural heterogeneity of the pipe samples was studied on a D8 DISCOVER with DAVINCI diffractometer (Bruker, Germany). The obtained results were analyzed using Bruker AXS DIFFRAC.EVA v.4.2 and DIFFRAC.TOPAS v.5.0 software, texture analysis was made using LaboTex software. For visual confirmation of stabilization conditions of cracks an electron back-scatter diffraction (EBSD) analysis was performed;
• stage «2». The RCGs in corrosive media of various compositions were determined on beam samples (200*17*5 mm) under static and cyclic loading. Preliminarily, the specimens were pre-cracked by fatigue in the air for the initiation of sharp crack [18, 19]. Specimens were fastened in three-electrode cells that were filled with the electrolyte under study. A constant tensile strain was applied to the specimens [19]. Experiments were carried out at room temperature (20 ± 2 °C). According to A.I. Marshakov, VE. Ignatenko, et al. [19], the RCG was determined and the stress intensity factor (K) at the crack tip, MPa-m0,5, was calculated. In the present work K1SCC value (threshold stress intensity factor for SCC) was not determined (K1SCC = 11 MPa-m0-5 for X70 steel in solution [24]). Researchers tried to choose the load values in the middle of the interval between K1SCC and K1C (threshold stress intensity factor for mechanical cracks); this range of mechanical stress corresponds to the plateau on the kinetic fatigue fracture diagram of X70 pipeline steel [19]. In this range, the RCG is almost independent of the load [10, 19];
stage «3-4». Before hydrocyclical testing, depth of SCC defects was determined using an ultrasonic flaw detector Phasor XS with
L99HK phased sensor array. Authors measured the linear dimensions of cracks using a ruler and photographic equipment after visualization with the help of a Magnaflux magnetic powder control kit. The stand, as well as the list ofSCC defects have been described in detail earlier [13, 25]. Cyclic tests on stand no. 1 involved simulation of the effects of starting and stopping the compressor unit at the linear section of the main trunk pipeline. N = 210 is the number of load cycles implemented with internal pressure amplitude change from 0,1 to 7,5 MPa (coefficient of skewness R = 0,01) at frequency f = 5 • 104 Hz. In cyclic tests on stand no. 2, simulation of the pressure changes at the linear section of UGP was carried out. N = 500 is the number of load cycles implemented with internal pressure amplitude change from 5,5 to 7,5 MPa (R = 0,7) at frequency f = 8-10~4 Hz. During tests the stands no. 3 and 4 imitated pressure fluctuations during the operation of the real section of UGP, the loading was carried out in non-uniform cycles. The loading mode of the stand no. 3 is a repeating sequence of 1 cycle with R = 0,01 and 32 minor loading cycles (R = 0,7). The loading mode of the stand no. 4 is a repeating sequence of 1 cycle with R = 0,01 and 40 minor loading cycles (R = 0,7). Based on the average regulatory period of service of the tape coating, the stands no. 3 and 4 simulated operating time of 20 years. During the hydraulic test of pipe test
stands, the state of SCC-cracks was monitored using the acoustic emissions complex A-Line 32D and a system of specialized fiber optic strain sensors.
Result and discussion
Kinetics of SCC on stage «0». Fig. 4 shows the surface of steel samples after solutions exposure in model electrolytes (tab. 2). It can be seen that for a specified period of time multiple sites of local corrosion have been formed on the samplings surface. Corrosion defects on 1W samplings were similar to defects on 2W samplings. After 24 days of testing, the corresponding dependences of LCD RCGs on the surface of samples in solutions of different compositions were constructed (fig. 5).
It is seen that with increasing time (t) the RCG of defects decreases. These dependences (see fig. 5) are described with an equation
da dt0
= к (t -10)я
(1)
where a means depth of the crack; k and n are empirical coefficients depending on the composition of a solution and the metallurgical properties of steel; t0 is the time of initiation of LCD after access of the corrosive medium to the pipe surface.
The result testifies that the conditions favorable in context of LCD formation develop
Fig. 4. Surface of a 2W sample after exposure in model electrolytes (a, x200) and corrosive sites around inclusions (b, x3000)
Table 2
The results of the composition energy dispersive analysis of non-metallic inclusions (see fig. 4b)
Solution Spectrum, mass %
C O Mg Al Si S Ca V Mn Fe
1а 17,54 23,73 0,00 2,27 0,18 17,15 20,76 0,09 0,27 17,56
2а 16,85 18,29 0,00 0,27 0,28 1,08 1,94 0,21 0,82 60,27
within a short time period after destruction of the pipeline coating and the receipt of a corrosive environment for steel.
Kinetics of SCC on stage «1». As can be seen in fig. 6, microcracks occur at the bottom of the pits, and then go to the main steel surface and quickly develop becoming slower with a large aspect ratio. Microcracks originate from surface concentrators after IP. It depends on the aspect ratio of a defect on the steel surface. The obtained results correlate with the morphology of the SCC operational defects, which are also derived from LCD (see fig. 6b). It was found that 2W pipes are more prone to microcracks formation than 1W pipes, but this effect is minor, for example, the time to microcracks initiation for these
steels in NS4 is 24 and 28 days respectively. It is fair to assume that decrease of RCG into depth is primarily associated with the weakening of tangential tensile stresses with a distance from the outer surface of pipe, as well as with the passing of various structural steels layers occurred due to hot rolling [17].
The data obtained by the EBSD method (fig. 7) illustrate that a transgranular crack stops when it reaches a crystallite with an orientation differ from the initial one. In the fig. 7a, the colors of the RGB palette show the crystallites with different crystallographic orientations, defined within the elementary stereographic triangle for cubic materials. Significant differences in the texture of 1W and 2W pipes expressed in different
Solutions: — NS4 — C2 — NOVATW
Fig. 5. The dependences of the LCD RCGs: a- 1W; b-2W
Fig. 6. Formation of microcracks at the bottom of pits in the laboratory (a) and SCC cracks from pit on the pipe surface under operating conditions (b)
— 1W — 2W
Fig. 7. EBSD maps of crystallites orientations in the pipes surface layers (a) and distributions of textural parameter and relative residual stresses over 6 used for determination of the crack depth limit (b)
volume fraction ratios of texture components and inhomogeneity of their distributions over the pipes wall thickness lead to increase of the resistance against SCC due to cracks branching at the initial stages of growth. To identify the tendency of pipes to SCC authors suggest a textural parameter 7gs/rr which describes the volume fractions ratio of rolling texture components typical: so-called Goss component (gs) and y-fiber (see fig. 7b, left). If this parameter is above 0,8, most cracks do not grow deeper than 0,105.
The textural components master curves 7gs/7Y and their comparison with the residual stresses in the pipe surface layer are presented in fig. 7b. At depth lower than indicated on fig. 7b as 0,085
(1W) and 0,135 (2W), the textural components ratio in the pipe walls will affect the microcracks initiation by intercrystalline corrosion, while with decrease of tangential tensile stresses with distance from the pipe outer surface the microcracks growth will slow down or stop.
The RCG in the first stage with a depth of 0,01.0,15 can be eyeball estimated using the empirical expression [21]:
da ( a
— = r exp I--
d/ I m
(2)
where 0,01 < a < 0,085 for 1W; 0,01 < a < 0,135
dc
for 2W; r = — (с - half-length of a crack on pipe d/
surface); m - empirical coefficient taking into account the ratio of depth to the length of a single crack.
Kinetics of SCC on stage «2». The effect of stand tests conditions on existing SCC defects was assessed by the view of the crack tip, the state of the corrosion products inside the crack and the metal matrix at the crack tip. As can be seen in fig. 8, the left photo at fig. 8a shows the view of a crack without testing, it can be seen that the corrosion products deposit near the pipe surface has separated but there is no stratification at the crack tip. After cyclic loading there are the visible changes of crack form, namely this is separation of the deposit across the depth of the crack (see fig. 8a, the central photo). The study of samples after static loading to a pressure of 12,0 MPa showed the increase of the distance between the sides of the crack with microplastic deformation of the metal at the crack tip (see fig. 8a, the right photo). The study of the microstructure at the crack tip did not show any visible differences or changes in the conditions without testing and after cyclic testing.
After recalculation of deformation in the test stand sections with SCC defects up to 0,205 deep, the calculated stresses did not exceed the value of 0,7c02 (fig. 9, see a). This fact excludes the possibility of mechanical crack propagation even
as a result of cyclic effects on crack tip. The fatigue tests of samples according to ASTM E466 confirm this conclusion. At maximum stresses close to the steel yield tensile ~ 490 MPa, failure occurs after 15000 cycles (see fig. 9b), which does not correspond to the actual loads in the gas pipeline. The microscopic results of the steel samplings near the SCC cracks with a depth of up to 0,155 before and after the completion of cyclic tests confirmed the absence of significant changes in the dislocation substructure, which excludes the possibility of their fatigue growth (see fig. 8c,d).
Thus, at stage «2» cracks can propagate only by the mechanism of local anodic dissolution of metal, which involves the accelerated dissolution of metal at crack tip. The values of RCG in the NS4 solution do not exceed 0,15 mm/year (fig. 10, see points 1). Adding sulfide and carbonate ions as the activators of anodic dissolution (AD) (see points 1', 1") accelerates RCG twofold. Soil contains an extensive list of chemical elements that can accelerate the propagation of cracks in pipe steels; tests were carried out in a more aggressive environment with various additives (see points 3) for a rapid assessment of the influence.
The use of an aggressive environment allowed to reduce the time of testing, as well as to determine
the load threshold value Ku
80 MPa m05.
Fig. 10 shows that injection of the AD activator,
After variable loading
After static loading
Fig. 8. The results of metallographic studies of SCC cracks with depth up to 0,26 (a) and near 0,36 (b) after hydraulic testing of pipes, and dislocation structure of 1W steel before (c)
and after (d) testing
a
b
£ 1,2
£ 1
ld iel
y1 o1
lat 0,9
re0,8
0,7
о 1W О 2W
о
/
à о
/
о / с/
.••о /О 8
о
о о
0,1
0,2
0,3
a550
^ 500 c
the
. S 450
0,4
tre
400
x
a350
300
250
о
о
о о X.
о X
•. о
Maximum depth of cracks under the sensor, %
50 100 150
Number of cycles to failure, 103
Fig. 9. Dependence of local stresses relative to the yield strength of steel on the maximum depth of SCC cracks in area of fiber optic sensor at pressure of 7,5 MPa (a) and fatigue tests of typical samples (b)
a e
1O1
C
О
Pá
1OO
1O-
1O
Solutions: O citrate buffer (CB) □ NS4 OCB + sulfide □ NS4 + sulfide O CB + carbonate □ NS4 + carbonate O CB + phosphate O soil solution O CB + ingibitor from accident O CB + iod section pipeline 6°0 -oo-oo-
с >5 o 0 o 0 0
O 4 0
4 ^07 0
8 o7
9 > -O-
О -л-
02 °
О 9-
1' í L 1 □
□ 1Q □ □
3O
5O
7O
9O
K, MPamO5
Fig. 10. The change in RCG in pipeline steel depending on stress intensity factors in various solutions
e.g. phosphate (see points 6), accelerates RGG, the corrosion inhibitor injection (see point 9) slows down the crack growth [23], the nitrate and iodide introductions do not change RGG. The effect of the studied additives on the value of RGG comports with their effect on AD rate of steel.
The impact of composition of medium on the value of RGG was also measured under low-amplitude cyclic loads in NS4 solution and with the addition of sulfide and carbonate ions. It has been established that in NS4 the crack growth accelerates half as much as compared with an air,
in a solution with additives it is 2,5 times faster than in the air. These values were used to calculate RCG in stage «3».
The dependencies of RCG in stage «2» are described by the equation
da
~át
xBT
(
M
l - n zF p
( ( ln
1 1-й í t л 'o 1-n
\SF
N+1 n
к - Ks
nr0a2r
(3)
0
0
n
n
where B =
2N
N -1
fp.^ \ Fr
- dissolution rate
da .
— =AiNb dt ' "
of fresh surface metal, A/m2; n, M, z, F, t0, eF r0, N, P, gy E, K1SCC - constants [24].
Kinetics of SCC on stage «3-4». Determination of RCG at stages «3» was carried out according to the results of the described above complex of laboratory testing of samples and hydraulic tests of pipe lashes with SCC cracks of various depths. When cracks reach a depth of 0,255 or more, the local stresses in the defective zones exceed 0,7c02. As a result, in addition to the corrosive environment, when calculating RCG, variable loads on pipeline should be taken into account. The dependences of RCG on variable load parameters are satisfactorily described if they are constructed in RCG - (AKaK Pmax// Y)n coordinates [21] (fig. 11). With an increase in the depth of SCC cracks (from 3 to 8 mm) the value of RCG increases from 0,1 to 0,6 mm per year. Taking into account the obtained data on the influence of the corrosive environment on the RCG and the actual number of variable loads experienced by the pipeline during operation, the RCG dependencies in the NS4 solution without additives and in the presence of aggressive additives Na2S or NaHCO3 were constructed. Consideration of influence of corrosive environment and the modes of operation of gas pipelines rises the calculated rates of crack propagation, RCG values above 1 mm per year were observed for defects of more than 5 mm deep. The dependencies of RCG in stage «3-4» are described by the equation
K ax AK "
f r N
. 32
(4)
where N - number of low-amplitude cycles in year; Nb - number of high-amplitude cycles in year; At - complex empirical coefficient taking into account the parameters of pipe and operating conditions; K max - is the maximum stress intensity factor calculated for a pipe with SCC; AKa - the range of variation of K when implementing variable loading; / - the average frequency of cyclic changes in the internal pressure in the pipeline with small amplitudes, Hz; n - coefficient depending on the pipe parameters; a, P, y - empirical constants.
Comprehensive universal model of the time-dependent formation and propagation of SCC. Based on the calculated experimental data, the total rate of the formation and propagation process for a single SCC crack before the moment of destruction will be described by eq. (5) (see below). At the same time, the minimum RCG value at each subsequent stage cannot be less than the RCG at the previous stage. The above-mentioned inequality determines transition of the SCC process to the next stage of development. The graphical realization of the composed system of equations is shown in fig. 12. It takes into account the possible range of values and the areas of definition of each individual function for 1W and 2W pipes with a wall thickness of 16,5 mm.
S3 3 u
о
О
Tests: O in air
O in NS4 solution without additives O in NS4 solution in presence of aggressive additives Na2S or NaHCO3
......СГ'Оц_
101
102
103
104
(AKK /py
v max-' '
Fig. 11. The RCG dependences on AKaK emJf
2
1
0
Calculated RCG values: O befor destruction ■ o at the initial moment of time Stage II Stage III
о —
о
)
о с
О I _
о I 1
_ i В
On air NS4 + NS4 inhibitor
NS4 + NS4 + sulfide carbonate
On air NS4 + NS4 inhibitor
NS4 + NS4 + sulfide carbonate
Fig. 12. RCG dependence for 1W pipe at stages II and III
0
da
dT
i = 0| к (t - t0)n
■il f ' i = 11 rexpI —
( ;*
i = 21
il m >-» ( t0 У» „А i, ( k2 - K1s ^N-i 1-»
i = 3| A,Nb
1- n zF p
кт ax Ak'
B1
ln
V I nr0°Y J J
da da da da da — <—<—< — < —, (5)
dt0 dt1 dt2 dt3 dt4
f Г Nl 32
i = 41 0 - rupture
where i is the stage number of SCC according to the model.
A graphical implementation of the composed system of equations (5) for stages II, III is shown in fig. 12 related to model electrolytes and calculated measurement errors for 1W pipes with 5 of16,5 mm.
Recommended operation of gas pipelines with SCC. Generalization of the obtained research results makes authors turn to practical recommendations for operation of gas pipelines having SCC cracks. The calculated dependences of crack depths as the functions of pipeline operating time are shown in fig. 13. The dependences are plotted for 1W pipes, and take into account the variations of the calculated RCG minimums and maximums at the stages II and III. The following practical recommendations for operating of gas pipelines are formed from the obtained RCG curves:
• in the A, B, C areas the pipes with SCC cracks should be cut out since such defects are of dangerous
• in the D and E areas pipes with SCC cracks can be left in operation until repairs are carried out, providing that these pipes are subject to periodic in-line inspections. The RCG should be predicted in accordance with the intensive growth curve in areas D and B, and in accordance with the slow growth curve in areas E and C;
• in F area pipes with SCC cracks having a < 0,15 are recommended for re-isolation using coatings containing inhibitor. These SCC cracks belong to the dormant ones and grow very slowly, in case of reisolation the growth of defect will stop for sure.
I
Growth:
— intensive
— slow
Operating time, years Fig. 13. Rapid and slow development of the SCC cracks
0
To estimate the ceiling sizes of cracks, which outreaching will make a gas pipeline break down, a two-parameter fracture criterion is used in the form of the engineering method R61.
Conclusions
The characteristic scenarios for SCC cracks development in steel pipelines describing the process stages, as well as the temporal changes of RCG and depth of cracks are considered in relation to the Gazprom's statistics of accidents. Few quantitative effects on the rates of formation and propagation of SCC were established, namely these are the effects of residual stresses, pipe-steel structural components and textures, chemical composition of soil electrolyte, magnitude and type of the mechanical load. The rate of SCC cracks was calculated and experimentally determined in 040"...56" pipes with 16,5.18,7 mm thick walls made of the X70-grade controlled rolling steels, after their long-term operation in gas pipelines. It was found that SCC cracks down to 0,25 deep did not present immediate hazard to pipelines reliability. On exposure to corrosion environment the predicted development rates of such damages did not exceed 0,3.0,4 mm per year excluding the cracks located in the welded joints and along a welding line. Growth of the SCC cracks down to 0,25 deep would stop without electrolyte.
1 See: API 579-1/ASME FFS-1. Fitness-For-Service. 2007. R Gazprom 9.4-030-2014. Method for assessing the stability of compressor station process pipelines with stress corrosion defects [Metodika otsenki prochnosti tekhnologicheskikh truboprovodov kompressornykh stantsiy so stress-korrozionnymi defektami].
If a pipe was left in operation, or a crack was not detected during the technical diagnostics of pipelines sites, the in-line diagnostics of RCG for cracks with a < 0,255 would not reach the accident values during the standard inter-diagnostic period under 5 years. The obtained result provides leaving defected pipes in operation until the scientifically validated repairs.
Authors suggest an algorithm for operating a damaged pipeline with the SCC cracks of the base pipe metal. Pipes having the 0,1.0,255 deep cracks could be left in operation until a scheduled inspection, but only if their growth rates are predicted, and periodic observations are being made by means of the in-line diagnostic monitoring. It is recommended to leave pipes with SCC cracks of less than 0,15 depth upon condition of re-insulation of these pipes using coatings containing inhibitor.
References
1. ALIMOV, S.V., A.B. ARABEY,
I.V. RYAKHOVSKIKH, et al. The concept of diagnosis and repair of gas mains in regions with high susceptibility to stress corrosion cracking [Kontseptsiya diagnostirovaniya i remonta magistralnykh gazoprovodov v regionakh s vysokoy predraspolozhennostyu k stress-korrozii]. Gazovaya Promyshlennost. 2015, no. S2 (724), pp. 10-15. ISSN 0016-5581. (Russ.).
2. ANTONOV, V.G., et al. Stress corrosion cracking of main gas pipelines [Korrozionnoye rastreskivaniye pod napryazheniyem
trub magistralnykh gazoprovodov]: atlas. ARABEY, A.B., Z. KNOSHINSKI (eds.). Moscow: Nauka, 2006. (Russ.).
3. CHENG, Y.F. Stress corrosion cracking
ofpipelines. Hoboken, New Jersey: John Wiley & Sons, 2013.
4. NATIONAL ENERGY BOARD. Report of public inquiry concerning stress corrosion cracking
on Canadian oil and gas pipelines: report of the inquiry. Calgary, Alberta, Canada: NEB, 1996, MH-2-95.
5. GAMBOA, E. Inclined stress corrosion cracks
in steel pipelines. Corrosion Engineering, Science and Technology. 2015, vol. 50, no. 3, pp. 191-195. ISSN 1478-422X.
6. SADEGHI MERESHT, E., T. SHAHRABI FARAHANI, J. NESHATI. Failure analysis of stress corrosion cracking occurred in a gas transmission steel pipeline. Eng. Fail. Anal. 2011, vol. 18, no. 3, pp. 963-970. ISSN 1350-6307. DOI: https://doi.org/10.1016/ j.engfailanal.2010.11.014
7. ZAITSEV, A.I., I.G. RODIONOVA,
O.N. BAKLANOVA, et al. Structural factors governing main gas pipeline steel stress corrosion cracking resistance. Metallurgist. 2013, vol. 57, no. 7-8, pp. 695-706. ISSN 0026-0894.
8. RYAKHOVSKIKH, I.V., R.I. BOGDANOV, V.E. IGNATENKO. Intergranular stress corrosion cracking of steel gas pipelines in weak alkaline soil electrolytes. Eng. Fail. Anal. 2018, vol. 94, pp. 87-95. ISSN 1350-6307.
9. RYAKHOVSKIKH, I. V. A complex technique of research of corrosion-mechanical properties of low-carbon low-alloyed pipe steels and
an assessment of theirfirmness against SCC [Kompleksnaya metodika issledovaniya korrozionno-mekhanicheskikh svoystv malouglerodistykh nizkolegirovannykh trubnykh staley i otsenka ikh stoykosti protiv korrozionnogo rastreskivaniya pod napryazheniyem]. Candidate thesis (engineering). National Research Nuclear University MEPhI. Moscow, 2013. (Russ.).
10. MALKIN, A.I., A.I. MARSHAKOV,
A.B. ARABEY. Processes of crack initiation and propagation on the steels of main pipelines [Protsessy zarozhdeniya i rosta korrozionnykh treshchin na stali magistralnykh truboprovodov]. Pt. I: Modern understanding of the mechanisms of stress corrosion cracking of pipeline steels in aqueous media [Sovremennyye predstavleniya o mekhanizmakh korrozionnogo rastreskivaniya staley v vodnykh sredakh]. Korroziya: materialy, zashchita. 2009, vol. 10, pp. 1-15. ISSN 1813-7016. (Russ.).
11. CHEN, W. Modeling and prediction of stress corrosion cracking of pipeline steels. In: EL-SHERIK, A.M. (ed.). Trends in Oil and Gas Corrosion Research and Technologies. Production and Transmission. 1st ed. Sawston, Cambridge, UK: Woodhead Publishing, 2017, ch. 30,
pp. 707-748. DOI: http://dx.doi.org/10.1016/ B978-0-08-101105-8.00030-9
12. KARPOV, S.V., D.I. SHIRYAPOV, A.S. ALIKHASHKIN. Complex research of stress corrosion cracking at trunk gas pipelines: practice and outlooks [Kompleksnyye issledovaniya korrozionnogo rastreskivaniya pod napryazheniyem na magistralnykh gazoprovodakh: opyt i perspektivy]. Vesti Gazovoy Nauki. Moscow: Gazprom VNIIGAZ LLC, 2016,
no. 3 (27): Improving reliability of gas mains subject to stress corrosion cracking, pp. 144-154. ISSN 2306-8949. (Russ.).
13. ARABEY, A.B., O.N. MELEKHIN, I.V. RYAKHOVSKIKH, et al. Studying
a possibility of continuous service of pipes with minor stress-corrosion cracks [Issledovaniye vozmozhnosti dlitelnoy ekspluatatsii trub s neznachitelnymi stress-korrozionnymi povrezhdeniyami]. Vesti Gazovoy Nauki. Moscow: Gazprom VNIIGAZ LLC, 2016, no. 3 (27): Improving reliability of gas mains subject to stress corrosion cracking, pp. 4-11. ISSN 2306-8949. (Russ.).
14. EADIE, R.L., K.E. SZKLARZ, R.L. SUTHERBY. Corrosion fatigue and near-neutral pH stress corrosion cracking of pipeline steel and the effect of hydrogen sulfide. Corrosion. 2005, vol. 61,
no. 2, pp. 167. ISSN 0010-9312.
15. ARABEY, A.B., T.S. ESIYEV,
I.V. RYAKHOVSKIKH et al. Influence of features of the pipe production technology on resistance to stress corrosion cracking during the operation of main gas pipelines [Vliyaniye osobennostey tekhnologii proizvodstva trub na ustoychivost k korrozionnomy rastreskivaniyu pod napryazheniyem pri ekspluatatsii magistralnykh gazoprovodov]. Gazovaya Promyshlennost. 2012, no. 2(673), pp. 52-54 (in Russia). ISSN 0016-5581.
16. LEIS, B.N. Initiation of SCC on gas transmission pipelines in related cracking environments. In: Corrosion 96, 24-29 March, Denver, Colorado. NACE International, 1996, paper no. 268.
17. PERLOVICH, Y., O. KRYMSKAYA,
M. ISAENKOVA, et al. Effect of layer-by-layer texture inhomogeneity on the stress corrosion of gas steel tubes. Materials Science Forum. 2017, vol. 879, pp. 1025-1030. ISSN 1662-9752. DOI: 10.4028/www. scientific.net/MSF. 879.1025.
18. RYAKHOVSKIKH I., R. BOGDANOV, T. ESIEV, et al. Stress corrosion cracking of pipeline steel
in near-neutral pH environment. Proc. of Materials Science & Technology 2014, October 12-16. Pittsburgh, PA, USA, 2014, vol. 1, pp. 807-814.
19. MARSHAKOV, A.I., V.E. IGNATENKO, et al. Effect of electrolyte composition on crack growth rate in pipeline steel. Corrosion Science. 2014, vol. 83, pp. 209-216. ISSN 0010-938X.
20. KANTOR, M.M., V.V. SUDIN,
V.A. BOZHENOV. Application of the slow electron diffraction method for studying stress corrosion cracking of trunk pipelines [Primeneniye metoda diffraktsii otrazhennykh elektronov dlya izucheniya korrozionnogo rastreskivaniya pod napryazheniyem magistralnykh truboprovodov]. Vesti Gazovoy Nauki. Moscow: Gazprom VNIIGAZ LLC, 2016, no. 3 (27): Improving reliability of gas mains subject to stress corrosion cracking, pp. 30-36. ISSN 2306-8949. (Russ.).
21. CHEN, W. An overview of near-neutral pH stress corrosion cracking in pipelines and mitigation strategies for its initiation and growth. Corrosion. 2016, vol. 72, no. 7, pp. 962-977. ISSN 0010-9312.
22. GLADKIKH, N.A., M.A. MALEYEVA, et al. Study of the initial stages of local dissolution of carbon steel in a chloride solution [Issledovaniya nachalnykh stadiy
lokalnogo rastvoreniya uglerodistoy stali v khloridnom rastvore]. Korroziya: materialy, zashchita. 2016, no. 6, pp. 17-22. ISSN 1813-7016. (Russ.).
23. MARSHAKOV, A.I., I.V. RYAKHOVSKIKH, V.E. IGNATENKO, et al. Development
of inhibiting compositions aimed at preventing stress corrosion cracking of gas mains [Razrabotka ingibiruyushchikh kompozitsiy dlya predotvrashcheniya korrozionnogo rastreskivaniya pod napryazheniyem magistralnykh gazoprovodov]. Vesti Gazovoy Nauki. Moscow: Gazprom VNIIGAZ LLC, 2016, no. 3 (27): Improving reliability of gas mains subject to stress corrosion cracking, pp. 48-63. ISSN 2306-8949. (Russ.).
24. LU, B.T. Further study on crack growth model of buried pipelines exposed to concentrated carbonate-bicarbonate solution. Eng. Fract. Mech. 2014, vol. 131, pp. 296-314. ISSN 0013-7944.
25. ARABEY, A.B., I.V. RYAKHOVSKIKH, A.V. MELNIKOVA, et al. Technology for repair of gas mains subject to stress corrosion cracking [Tekhnologiya remonta magistralnykh gazoprovodov, podverzhennykh korrozionnomy rastreskivaniyu pod napryazheniyem]. Nauka
i Tekhnika v Gazovoy Promyshlennosti. 2017, no. 3, pp. 3-16. ISSN 2070-6820. (Russ.).
Закономерности развития стресс-коррозии труб магистральных газопроводов в средах с околонейтральным водородным индексом
И. В. Ряховских
ООО «Газпром ВНИИГАЗ», Российская Федерация, 142717, Московская обл., Ленинский р-н, с.п. Развилковское, пос. Развилка, Проектируемый пр-д № 5537, вл. 15, стр. 1 E-mail: I_Ryakhovskikh@vniigaz.gazprom.ru
Тезисы. В статье рассмотрены характерные сценарии коррозионного растрескивания под напряжением (КРН) стальных труб в составе магистральных газопроводов (МГ), описывающие основные стадии процесса, скорости и условия их реализации. Для конструктивно различных труб диаметром 1420 мм, произведенных на заводах Германии и CCCP из малоуглеродистых сталей класса прочности Х70, после длительной эксплуатации в составе МГ реализован комплекс электрохимических, коррозионных, коррозионно-механических, механических, ресурсных испытаний, а также рентгеновских исследований послойной текстурной неоднородности и остаточных напряжений, металлофизических исследований микроструктуры и дислокационной субструктуры фрагментов труб. Кинетика роста вновь образованных трещин определяется уровнем остаточных технологических напряжений на поверхности стали и аспектным отношением трещин. Показано, что параметрами, характеризующими возможность роста или торможения трещин на начальных стадиях процесса, являются уровни текстурной неоднородности материала и остаточных технологических напряжений во внешних слоях стенки трубы. Фактором, способствующим релаксации напряжений в вершине трещины, является расположенная здесь зона пластической деформации. После циклических испытаний труб, моделирующих работу МГ, вблизи вершин коррозионно-механических трещин глубиной до 3 мм не установлено признаков усталостного прироста и существенных изменений дислокационной субструктуры. Однако по результатам фрактографических исследований изломов для большинства труб после циклических испытаний отмечается нарушение целостности оксидных пленок, что при доступе коррозионной среды может стимулировать рост трещин по механизму локального анодного растворения. Показано, что при статических и малоамплитудных циклических нагрузках скорость роста трещины в испытательных средах с рН, равными 5,5 и 7,0, ускоряется в присутствии компонентов грунта,
стимулирующих анодное растворения металла (сульфид-, карбонат- и фосфат-ионы). Установлено, что КРН-трещины глубиной до 0,25 (где 5 - толщина стенки трубы) не оказывают существенного влияния на надежность трубопроводов, при воздействии коррозионной среды прогнозные темпы роста таких дефектов не превышают 0,3.0,4 мм/год. Исключением являются коррозионно-механические трещины любых размеров, расположенные в сварных соединениях и по линии сплавления, которые всегда следует рассматривать как потенциально опасные и подлежащие устранению в возможно кроткие сроки после обнаружения.
Ключевые слова: магистральный газопровод, стресс-коррозия, циклические испытания стальных труб, дефект, трещина, текстурная неоднородность материала, остаточные технологические напряжения, скорость развития трещины, водородный индекс.
Список литературы
1. Алимов С.В. Концепция диагностирования и ремонта магистральных газопроводов в регионах с высокой предрасположенностью к стресс-коррозии / С.В. Алимов, А.Б. Арабей, И.В. Ряховских и др. // Газовая промышленность. - 2015. - № S2 (724). - С. 10-15.
2. Антонов В.Г. Коррозионное растрескивание под напряжением труб магистральных газопроводов: атлас / В.Г. Антонов и др.; под общ. ред. А.Б. Арабея, З. Коношински. - М.: Наука, 2006.
3. Cheng Y.F. Stress corrosion cracking of pipelines / Y.F. Cheng. - Hoboken, New Jersey: John Wiley & Sons, 2013.
4. MH-2-95. Report of public inquiry concerning stress corrosion cracking on Canadian oil and gas pipelines: report of the inquiry / National Energy Board. - Calgary, Alberta, Canada: NEB, 1996.
5. Gamboa E. Inclined stress corrosion cracks in steel pipelines / E. Gamboa // Corrosion Engineering, Science and Technology. - 2015. - Т. 50. - № 3. - С. 191-195.
6. Sadeghi Meresht E. Failure analysis of stress corrosion cracking occurred in a gas transmission steel pipeline / E. Sadeghi Meresht, T. Shahrabi Farahani, J. Neshati // Eng. Fail. Anal. - 2011. - Т. 18. - №. 3. - С. 963-970. -DOI: https://doi.org/10.1016/j.engfailanal.2010.11.014
7. Zaitsev A.I. Structural factors governing main gas pipeline steel stress corrosion cracking resistance / A.I. Zaitsev, I.G. Rodionova, O.N. Baklanova, et al. // Metallurgist. - 2013. - Т. 57. - № 7-8. - С. 695-706.
8. Ryakhovskikh I.V. Intergranular stress corrosion cracking of steel gas pipelines in weak alkaline soil electrolytes / I.V. Ryakhovskikh, R.I. Bogdanov, V.E. Ignatenko // Eng. Fail. Anal. - 2018. - Т. 94. - С. 87-95.
9. Ряховских И.В. Комплексная методика исследования коррозионно-механических свойств малоуглеродистых низколегированных трубных сталей и оценка их стойкости против коррозионного растрескивания под напряжением: дис. ... к.т.н.: 01.04.07 / И.В. Ряховских [Место защиты: Нац. исслед. ядерный ун-т]. - М., 2013. - 155 с.
10. Малкин А.И. Процессы зарождения и роста коррозионных трещин на стали магистральных газопроводов. Ч. 1: Современные представления о механизмах коррозионного растрескивания сталей в водных средах / А.И. Малкин, А.И. Маршаков, А. Б. Арабей // Коррозия: материалы, защита. - 2009. -№ 10. - С. 1-15.
11. Chen W. Modeling and prediction of stress corrosion cracking of pipeline steels / W. Chen // Trends in oil and gas corrosion research and technologies. Production and transmission / A.M. El-Sherik (ed.). - 1-е изд. - Sawston, Cambridge, UK: Woodhead Publishing, 2017. - Гл. 30. - С. 707-748. - DOI: http://dx.doi.org/10.1016/B978-0-08-101105-8.00030-9
12. Карпов С.В. Комплексные исследования коррозионного растрескивания под напряжением на магистральных газопроводах: опыт и перспективы / С.В. Карпов, Д.И. Ширяпов, А.С. Алихашкин // Вести газовой науки: Повышение надежности магистральных газопроводов, подверженных коррозионному растрескиванию под напряжением. - М.: Газпром ВНИИГАЗ, 2016. - № 3 (27). -С. 144-154.
13. Арабей А.Б. Исследование возможности длительной эксплуатации труб с незначительными стресс-коррозионными повреждениями / А.Б. Арабей, О.Н. Мелёхин, И.В. Ряховских и др. // Вести газовой науки: Повышение надежности магистральных газопроводов, подверженных коррозионному растрескиванию под напряжением. - М.: Газпром ВНИИГАЗ, 2016. - № 3 (27). - С. 4-11.
14. Eadie R.L. Corrosion fatigue and near-neutral pH stress corrosion cracking of pipeline steel and the effect of hydrogen sulfide / R.L. Eadie, K.E. Szklarz, R.L. Sutherby // Corrosion. - 2005. - Т. 61. - № 2. - С. 167.
15. Арабей А.Б. Влияние особенностей технологии производства труб на устойчивость к коррозионному растрескиванию под напряжением при эксплуатации магистральных газопроводов / А. Б. Арабей, Т.С. Есиев, И.В. Ряховских и др. // Газовая промышленность. - 2012. - № 2 (673). - С. 52-54.
16. Leis B.N. Initiation of SCC on gas transmission pipelines in related cracking environments / B.N. Leis // Corrosion 96 conf., 24-29 March, Denver, Colorado. - NACE International, 1996. - Paper no. 268.
17. Perlovich Y. Effect of layer-by-layer texture inhomogeneity on the stress corrosion of gas steel tubes / Y. Perlovich, O. Krymskaya, M. Isaenkova, et al. // Materials Science Forum. - 2017. - Т. 879. - С. 1025-1030. -DOI: 10.4028/www. scientific.net/MSF. 879.1025.
18. Ryakhovskikh I. Stress corrosion cracking of pipeline steel in near-neutral pH environment / I. Ryakhovskikh, R. Bogdanov, T. Esiev, et al. // Proc. of Materials Science & Technology 2014, October 12-16. - Pittsburgh, PA, USA, 2014. - Т. 1. - С. 807-814.
19. Marshakov A.I. Effect of electrolyte composition on crack growth rate in pipeline steel / A.I. Marshakov, V.E. Ignatenko, et al. // Corrosion Science. - 2014. - Т. 83. - С. 209-216.
20. Кантор М.М. Применение метода дифракции отраженных электронов для изучения коррозионного растрескивания под напряжением магистральных трубопроводов / М.М. Кантор, В.В. Судьин, В.А. Боженов // Вести газовой науки: Повышение надежности магистральных газопроводов, подверженных коррозионному растрескиванию под напряжением. - М.: Газпром ВНИИГАЗ, 2016. -№ 3 (27). - С. 30-36.
21. Chen W. An overview of near-neutral pH stress corrosion cracking in pipelines and mitigation strategies for its initiation and growth / W. Chen // Corrosion. - 2016. - Т. 72. - № 7. - С. 962-977.
22. Гладких Н.А. Исследование начальных стадий локального растворения углеродистой стали в хлоридном растворе / Н.А. Гладких, М.А. Малеева и др. // Коррозия: материалы, защита. - 2016. - № 6. - С. 17-22.
23. Маршаков А.И. Разработка ингибирующих композиций для предотвращения коррозионного растрескивания под напряжением магистральных газопроводов / А.И. Маршаков, И.В. Ряховских, В.Э. Игнатенко и др. // Вести газовой науки: Повышение надежности магистральных газопроводов, подверженных коррозионному растрескиванию под напряжением. - М.: Газпром ВНИИГАЗ, 2016. -№ 3 (27). - С. 48-63.
24. Lu B.T. Further study on crack growth model of buried pipelines exposed to concentrated carbonate-bicarbonate solution / B.T. Lu // Eng. Fract. Mech. - 2014. - Т. 131. - С. 296-314.
25. Арабей А.Б. Технология ремонта магистральных газопроводов, подверженных коррозионному растрескиванию под напряжением / А.Б. Арабей, Ряховских И.В., Мельникова А.В. и др. // Наука и техника в газовой промышленности. - 2017. - № 3. - С. 3-16.
