Investigation of pH 2 S influence on 3% Cr tubing steel corrosion behaviours in CO 2 –H 2 S

Abstract

A series of experiments were conducted in autoclave to investigate corrosion properties of low Cr oil country tubular goods in CO₂/H₂S saturated brine solution (3 wt-%NaCl) at 90°C. Test material is 3Cr-80, which is gradually accepted in oil and gas industry, especially for downhole tubing and casing. The CO₂ partial pressure was fixed at 4·0 bar in all experiments, and the H₂S partial pressure was varied from 0, 2, 4, 8, 20, 40 and 800 mbar to 4000 mbar (4 bar). The results show that when the pH₂S varied at a lower range (2-40 mbar), 3Cr-80 materials’ corrosion rates were restricted, and coupons’ surface morphologies became glossy compared to the pure CO₂ corrosion. Under the condition of pH₂S = 800 mbar, H₂S no longer have a retardation effect on CO₂ corrosion but acceleration. The average corrosion rate was higher than pure CO₂ corrosion. When the pH₂S rise to 4000 mbar (equal to the CO₂ partial pressure), 3Cr corroded much more severe and serious mesa attack was observed.

Keywords

3%Cr steel Mesa attack Water chemistry

This work was supported by the funding of “Introduction Talents Scientific Research Project (YJRC-2013-19)”, China University of Petroleum, Beijing.

Introduction

CO₂/H₂S corrosion is one of the most prevalent forms of degradation of transportation pipe line, downhole tubing and refine facilities. In past several years, researches gradually tend to accept that trace amounts of H₂S can retard CO₂ corrosion in carbon steel,^1–5 including general corrosion and pitting attack. The main reason for this retardation is the fast formed protective iron sulphide layer.^5,6 In an H₂S present environment, many types of iron sulphides may form such as amorphous ferrous sulphide, mackinawite, cubic ferrous sulphide, smythite, greigite and pyrrhotite, among which mackinawite is considered to form first on the steel surface by a direct surface reaction,⁷ and performs a good protection. Even very small amounts of H₂S still can form this FeS layer. Smith and Pacheco presented a detail work on predicting minimum H₂S concentration for reduction of CO₂ corrosion in mild steel in slightly sour environment. He found that basic thermodynamic principles can be used to calculate the minimum H₂S levels for retarding CO₂ corrosion, and 25 ppm H₂S was the critical value for his experimental condition.⁷ Based on their calculation, most of previous studies were in reduction zone; thus, lower corrosion rate was obtained.

This type of inner layer was observed also at high temperature and seemed much denser than the outer layer.⁴ It also was reported in an early time that the crystalline precipitation product of ferrous ions by H₂S was solely composed of mackinawite <100°C in the absence of oxidants.⁸

However, some researchers who hold an opposite opinion still argue that a certain amounts of H₂S will lead to severe localised corrosion in mild steel. Tang et al. reported that the corrosion rate of SAE-1020 carbon steel at 90°C increased with the increase of H₂S concentrations from 58·91 to 408·44 mg L⁻¹, and severe localised corrosion on the carbon steel surfaces was observed in H₂S containing solutions.⁹ For carbon steel, pitting attack usually related to the cementite structure in steel, which acts as a cathode with respect to ferrite in acidic solution.¹⁰ In addition, galvanic coupling between FeS covered and uncovered area is another possible reason for localised attack in sour environment.¹¹

Another popular topic in this area was the difference of sweet or sour environment, which was based on the partial pressure ratio of CO₂/H₂S.^12–15 The earliest mention of a value of 500 for the CO₂/H₂S ratio was posted by Dunlop et al. in 1983.¹⁶ Based on their research, Srinivasan and Tebbal presented a more detail diagram for clarifying the corrosion environment. If the ratio of CO₂/H₂S is less than 200, the environment was dominated by sour corrosion; if this ratio is between 200 and 500, it was mixed control; and if this ratio is less than 500, the environment was dominated by CO₂.¹³ A few years later, Bernardus et al. changed this lower border from 200 to 20.¹⁷ This rule of thumb has been cited in a lot of paper. However, Smith pointed out that this rule of thumb value (500) was obtained just by chance selection from a wrong iron sulphide free energy data in the early time (1983). In addition, this value was extremely sensitive to thermodynamic data inputs. This rule of thumb value actually makes no sense to the sweet or sour environment.¹² The only useful values for corrosion research are the CO₂ and H₂S absolute partial pressures in gas phase or the concentrations in aqueous phase.

In addition, most of the previous studies focus on carbon steel; few of them mentioned low Cr content steel. This paper presented a full range of experiments on 3%Cr materials in oil and gas industry and tries to find the critical point for H₂S retarding on CO₂ corrosion.

Experimental

Materials

Low Cr content steels were used and gradually accepted in recent 10 years because of the low cost and good performance on resisting CO₂ pitting attack. Chemical composition of 3Cr-80 steel is shown in Table 1.

Table 1

Chemical composition of 3Cr-80 steel/mass-%

Elements
Materials	C	Mn	Si	P	S	Cr	Mo	Ni	Fe
3Cr-80	0·19	0·32	0·32	<0·01	<0·006	2·93	0·39	0·17	Balance

Test matrix and solutions

In total, eight series test conditions were performed in this study and test matrix is shown in Table 2. Temperature, CO₂ partial pressure, liquid velocity and solution chemistry were not varied. The only varied parameter was H₂S partial pressure.

Table 2

Experimental matrix

Condition no.	1	2	3	4	5	6	7	8
Constant parameter	T = 90°C, pCO₂ = 4·0 bar, stir rate 550 rev min⁻¹, 7 days
pH₂S/mbar	0	2	4	8	20	40	800	4000

The solution was simulated oilfield water. The original solution pH is 6·8, and the ion concentrations are shown in Table 3.

Table 3

Oil field water ions concentration/mg L⁻¹

K⁺Na⁺	Ca²⁺	Mg²⁻	Cl⁻	SO₄ ²⁻	HCO₃ ⁻	CO₃ ²⁻
8350	3885	57	19634	850	337	0

Equipment and procedures

All of the experiments in this study were conducted in a 2.0 L autoclave (Fig. 1). This autoclave was described specifically in an earlier study.¹⁸ The specimens were polished with silicon carbide no. 320 and no. 600 papers, rinsed with distilled water, degreased in acetone and then weighed after drying. The specimens were mounted on a holder, which was made of poly(tetrafluoroethylene). The holder was placed in the autoclave, then the brine solution was added into the vessel and the autoclave was closed. In the beginning, the temperature was raised to 45°C to remove some of the dissolved oxygen from the solution, and then purged with N₂ gas for ∼2 h. After purging with N₂, CO₂ or CO₂/H₂S mixed gas (pre-prepared) was injected into the autoclave and kept flowing for about 5 min in order to remove the residual N₂, then the outlet valve was closed, and the temperature and pressure were raised to designed values. The gas saturated process usually takes ∼1 h to fully saturate, and the judgment is pressure drop <0·1 bar within 30 min after the inlet was closed. Flow velocity was ∼1·7 m s⁻¹ (rotation speed, 550 rev min⁻¹). The testing duration was 7 days (168 h).

Figure 1

a high temperature and high pressure dynamic autoclave; b coupon specimen

Two parallel samples of each material were tested in the same condition. One specimen was used to calculate the average corrosion rate; the other one was used for morphology analysis.

After the experiment, weight loss specimens were rinsed with distilled water and corrosion scale was removed by Clark's solution. Then, the specimens were dried and photographed to record the corrosion status. Finally, the residual masses of the specimens were determined using an electronic balance (accuracy, 0·1 mg), and the average corrosion rate was calculated by the weight loss method. The analysed specimens were carefully wetted with deoxidised isopropanol and DI water without creating physical damage to the corrosion product. All samples are stored in a vacuum desiccator before and throughout the analysis process.

Surface analysis

Scanning electron microscopy (SEM) and energy dispersive spectrometer (EDS) were used to analyse the corrosion product layer, and profilometer was used to analyse the surface condition after removing corrosion product.

Results and discussion

Corrosion rate

General corrosion rates under different conditions are shown in Fig. 2, and the corresponding coupon pictures are shown in Fig. 3. In pure CO₂ corrosion condition, the general corrosion rate is 1·15 mm/a. General corrosion was observed and no localised corrosion appeared in the coupon surface. Keeping the CO₂ partial pressure constant, with increasing H₂S partial pressure from 2 to 20 mbar, the general corrosion rates were decreased continuously. The surface condition became better than the pure CO₂ corrosion. Some polishing gloss still can be seen after removing the scale. The minimum corrosion rate value appeared at a certain H₂S partial pressure between 20 and 40 mbar. However, when the H₂S partial pressure reached 800 mbar, the corrosion rate no longer was restricted by the adding H₂S but accelerated 30% compared to the pure CO₂ environment. General corrosion was observed as well, but the coupon surface changed from gloss to rough. Under the condition of 4000 mbar H₂S, the same as CO₂ partial pressure, the corrosion rate was almost twice of the pure CO₂ condition. Severe localised corrosion (mesa attack) occurred.

Figure 2

Corrosion rate versus pH₂S (T = 90°C, pCO₂ = 4·0 bar, 1·7 m s⁻¹, 7 days)

Figure 3

3Cr specimen pictures after corrosion product cleaning (T = 90°C, pCO₂ = 4·0 bar, 1·7 m s⁻¹, 7 days)

Corrosion product analysis

Surface morphology images of corrosion products were shown in Fig. 4. EDS analysis results were shown in Fig. 5. For pure CO₂ corrosion (Fig. 4a), typical cubic FeCO3 grains were observed and the grains were distributed evenly. EDS data (Fig. 5a) show that a little CaCO₃ also coexisted with FeCO₃. Underneath these crystals, a Cr rich product layer could also be formed, which has a barrier effect on ion exchanging.¹⁸ After very small amount of H₂S added in the system, these cubic grains disappeared and some fluffy structure formed on the surface (Fig. 4b). Based on previous research,⁹ this type of corrosion products could be fluffy mackinawite (FeS). According to the EDS analysis (Fig. 5b), FeS and FeCO₃ should be coexisted. As the H2S partial pressure increases but <40 mbar, the fluffy mackinawite FeS was transferred into granular FeS gradually (Fig. 4c and d ). FeS and FeCO3 coexisted as well. However, when the H₂S partial pressure reached 800 mbar (Fig. 4e), a density FeS grain layer was formed and some fluffy structure was formed on top of this FeS grain layer. According to the EDS analysis, almost all of the corrosion products are purely FeS and no C and O elements were detected (Fig. 5e). H₂S corrosion dominated the corrosion process, and this type of sole FeS corrosion products cannot protect the surface well and a higher corrosion rate happened. A seriously localised corrosion happened in the final corrosion condition; a much higher H₂S content (4000 mbar) induced product layer, generating an “internal stress” and then leading to film cracking (Fig. 4f); and severe localised attack occurred (Fig. 3, no.6). EDS data show that FeS₂ formed under this condition (Fig. 5f). This type of corrosion products has the worst protectiveness.

Figure 4

a pCO₂ = 4·0 bar, no H₂S; b pCO₂ = 0·4 MPa, pH₂S = 2 mbar; c pCO₂ = 0·4 MPa, pH₂S = 8 mbar; d pCO₂ = 0·4 MPa, pH₂S = 40 mbar; e pCO₂ = 0·4 MPa, pH₂S = 800 mbar; f pCO₂ = 0·4 MPa, pH₂S = 4000 mbar

Figure 5

a no H₂S; b pH₂S = 2 mbar; c pH₂S = 8 mbar; d pH₂S = 40 mbar; e pH₂S = 800 mbar; f pH₂S = 4000 mbar

Sulphur element content (At%) in corrosion product layer that changed with H₂S partial pressure is shown in Fig. 6. X-ray diffraction (XRD) analysis results are shown in Fig. 7. Sulphur content increased with the H₂S partial pressure. When the H₂S partial pressure are <40 mbar, the maximum sulphur content are <25%(At), FeCO₃ and FeS coexisted (Fig. 7), and the trace amount of H₂S restricted the CO₂ corrosion (Fig. 2). Nevertheless, when the H2S partial pressure rises to 800 mbar, the products sulphur content exceeds 50%, FeS or FeS₂ are the dominated corrosion products, and this type of product layers is less protective. In some circumstances, it will induce film cracking and lead to localised attack (Fig. 4f). It needs to be stressed that under the condition of 4000 mbar H₂S, the product layer cracked seriously and parts of the layer fell off during the XRD sample preparation process, so the full surface XRD scanning pattern did not show a typical peak. It did not reflect the real crystal structure of the corrosion products. Fortunately, the EDS data clearly show that the atom ratio of S/Fe is quite closed to 2∶1 and FeS₂ speculation is reasonable. Figure 8

Figure 6

Comparison of sulphur element contents at different pH₂S environment

Figure 7

XRD patterns of 3Cr specimens at different pH₂S

Figure 8

Concentration ratio of [H₂S]/[CO₂] that varied with partial pressure ratio of pH₂S/pCO₂ (90°C, pCO₂ = 4·0 bar,1·7 m s⁻¹,7 days)

Water chemistry analysis

Nordsveen et al. have published a series of influential papers to demonstrate the process of CO₂ and H₂S corrosion.¹⁹ A brief introduction is summarised as follows.

When CO₂ is dissolved in water: It is hydrated to give carbonic acid (H₂CO₃): Then, H₂CO₃ is dissociated into two steps: Similar to CO₂, H₂S dissolution and dissociation equations are written as follows: The overall electroneutrality equation for this system can then be written as follows: The concentration of the species that are derived from CO₂ and H₂S solubility and dissociation can be calculated by solving these equations if both the CO₂ and H₂S partial pressure (pCO₂ and pH₂S) and the appropriate expressions for the other four constants are known. These equilibrium constants are referencing from Nordsveen's research.¹⁹ Based on dissociated process, [H₂CO₃](aq.) and [H₂S](aq.) can be calculated based on these constants.

It is assumed that the direct reductions of H₂CO₃(aq.) and H₂S(aq.) are dominantly reducing the reaction during the corrosion process. Many researchers hold this opinion. The concentration of this two main species is shown in Table 4, and the curve between concentration ratio of [H₂S]/[CO₂] and partial pressure ratio of pH₂S/pCO₂ is shown in Fig. 8. When the pH₂S/pCO₂ ratio is <1∶100, the concentration ratio of [H₂S]/[CO₂] is located in a low range and [H₂S] is lower than [H₂CO₃] or they are at the same order of magnitude. Under this condition (H₂S partial pressure are <40 mbar in Fig. 3), H₂S will retard the CO₂ corrosion and lower corrosion rates occurred compared to the pure CO₂ corrosion; meanwhile, a better surface condition was obtained (Fig. 3, nos.2-4). However, when pH₂S/pCO₂ ratio reached 1∶5 (the partial pressure of H₂S is 800∶mbar), [H2S]aq. is almost 400 times of [H₂CO₃]aq., and H₂S no longer have the retardation effect on CO₂ corrosion, but acceleration. Under this condition, H₂S dominated the corrosion process and the mechanism is totally different from the previous one. Less protective sulphide film leads to a higher corrosion rate (Fig. 3). This situation became more serious when pH₂S/pCO₂ is equal to 1·0, pH₂S = 4000∶mbar. Concentration of [H₂S]aq. is almost 2000 times of [H₂CO₃]. H₂S(aq.) reducing reaction dominated the corrosion system absolutely, and it can be considered as a pure H₂S corrosion. The corrosion products will dominantly controlled the outward diffusion of Fe²⁺ and inward diffusion of [H₂S]aq. under this condition.²⁰

Table 4

Calculating concentration of each species

No.	pCO₂,bar	pH₂S,bar	pCO₂/pH₂S	pH₂S/pCO₂	[H₂CO₃]	[H₂S]	[H₂S]/[H₂CO2]
1	4	0	∞	0·00000025	1·53×10⁻⁴	0	0
2	4	0·002	2000	0·0005	1·53×10⁻⁴	1·52×10⁻⁴	9·95×10⁻¹
3	4	0·004	1000	0·001	1·53×10⁻⁴	3·05×10⁻⁴	1·99×10⁰
4	4	0·008	500	0·002	1·53×10⁻⁴	6·10×10⁻⁴	3·98×10⁰
5	4	0·04	100	0·01	1·53×10⁻⁴	3·05×10⁻³	1·99×10¹
6	4	0·8	5	0·2	1·53×10⁻⁴	6·10×10⁻²	3·98×10²
7	4	4	1	1	1·53×10⁻⁴	3·05×10⁻¹	1·99×10³

Conclusions

In a CO₂ dominated corrosion environment, a trace amount of H₂S (<40 mbar, pH₂S/pCO₂<1∶100) will decrease the CO₂ corrosion rate in 3%Cr steel and the steel surface became better as well. However, when the H₂S partial pressure increases to a certain value (800 mbar in this paper), the system will change from CO₂ domination to H₂S domination and the corrosion rates will increase compared to the pure CO₂ corrosion.

A seriously localised corrosion happened when H₂S partial pressure is 4000 mbar; a possible reason is that FeS₂ product layer generates an “internal stress” and then leads to film cracking. This type of corrosion products has the worst protectiveness, and severe localised attack occurred consequently.

Because of the huge solubility difference of CO₂ and H₂S gas in brine water, 1∶500 of H₂S partial pressure in CO₂ can form the same order of magnitude of corrosion species H₂S (aq.) and CO₂ (aq.). When the partial pressure of H₂S is 1∶5 of CO₂, the concentration of H₂S (aq.) is almost 400 times of H₂CO₃(aq.). H₂S corrosion takes over the system domination, and corrosion mechanisms are totally different.

H₂S not always restricts or accelerates CO₂ in low alloy steel; it depends on the partial pressure of both corrosive gas and their ratios. Generally, it can be concluded that trace amount of H₂S will retard CO₂ corrosion and a much higher H₂S will induce a seriously localised attack on low alloy steel.

Acknowledgements

The authors would like to acknowledge the support from the Institute for Corrosion and Multiphase Technology, Ohio University, USA, and the China National Offshore Oil Corporation Research Institute.

References

Abelev

.: ‘Effect of H₂S on Fe corrosion in CO₂-saturated brine’ J. Mater. Sci. 2009 44 (22) 6167–6181. doi: 10.1007/s10853-009-3854-4

Valdes

Case

Ramirez

and Ruiz

: ‘The effect of small amounts of H₂S on CO₂ corrosion of a carbon steel’; 1998

Houston, TX

NACE International.

Choi

Y.-S.

Nesic

and Ling

: ‘Effect of H₂S on the CO₂ corrosion of carbon steel in acidic solutions’ Electrochim. Acta 2011 56 (4) 1752–1760. doi: 10.1016/j.electacta.2010.08.049

Jingen

: ‘Influence of H₂S content on CO₂ corrosion behaviors of N80 tubing steel’ Pet. Sci. Technol. 2011 29 (13) 1387–1396. doi: 10.1080/10916466.2010.545784

Lee

K.-LJ

and Nesic

: ‘The effect of trace amount of H₂S on CO₂ corrosion investigated by using the EIS technique’; 2005

Houston, TX

NACE International.

Yin

Z. F.

.: ‘Corrosion behavior of SM 80SS tube steel in stimulant solution containing H₂S and CO₂’ Electrochim. Acta 2008 53 (10) 3690–3700. doi: 10.1016/j.electacta.2007.12.039

Smith

S. N.

and Pacheco

J. L.

: ‘Prediction of corrosion in slightly sour environments’; 2002

Houston, TX

NACE International.

Berner

R. A.

: ‘Iron sulfides formed from aqueous solution at low temperatures and atmospheric pressure’ J. Geol. 1964 72 (3) 293–306. doi: 10.1086/626987

Tang

: ‘The effect of H₂S concentration on the corrosion behavior of carbon steel at 90°C’ Corros. Sci. 2010 52 (6) 2050–2058. doi: 10.1016/j.corsci.2010.02.004

10.

Berntsen

Seiersten

and Hemmingsen

: ‘Effect of FeCO₃ supersaturation and carbide exposure on the CO₂ corrosion rate of carbon steel’; 2011

Houston, TX

NACE International.

11.

Durnie

W. H.

Woollam

R. C.

Vera

J. R.

Huggins

and Mendez

: ‘Localized corrosion due to galvanic coupling between FeS-covered and uncovered areas: another oilfield myth’ Corrosion 2013.

12.

Smith

S. N.

: ‘Discussion of the history and relevance of the CO₂/H₂S ratio’; 2001

Houston, TX

NACE International.

13.

Srinivasan

and Tebbal

: ‘Critical factors in predicting CO₂/H₂S corrosion in multiphase systems’ Corrosion 1998.

14.

Yan

Deng

Dong

Zhang

and Li

: ‘Experimental study of 3% Cr tubing steel in CO₂ and CO₂/H₂S corrosion environment’ Oil Gas Facilities 2012 10 43–48. doi: 10.2118/157302-PA

15.

Zhang

.: ‘Corrosion rate of hydrogenation to C110 casing in high H₂S environment’ J. Wuhan Univ. Technol. Mater. Sci. Ed. 2012 27 (6) 1081–1083. doi: 10.1007/s11595-012-0605-1

16.

Dunlop

A. K.

Hassel

H. L.

and Rhodes

P. R.

: ‘Fundamental considerations in sweet gas well corrosion’ Corrosion 1983.

17.

Bernardus

Kapusta

Whitham

Girgis

Rippon

Thomas

M. J. J.

and John

: ‘Improvements on De Waard–Milliams corrosion prediction and applications to corrosion management’ Corrosion 2002.

18.

Wei

JinGen

XiaoRong

: ‘Effects of silty sand on CO₂ corrosion behavior of low-Cr tubing steel’ Chin. Sci. Bull. 2012 57 (2) 927–934.

19.

Nordsveen

Nesic

Nyborg

.: ‘A mechanistic model for carbon dioxide corrosion of mild steel in the presence of protective iron carbonate films—Part 1: Theory and verification’ Corrosion 2003 59 (5) 443–456. doi: 10.5006/1.3277576

20.

Liu

Wang

and Ke

: ‘Corrosion behaviors of X52 pipeline steel in high H₂S concentration solutions at temperatures ranging from 25°C to 140°C’ Corros. Eng. Sci. Technol. 2013 48 (5) 380–387. doi: 10.1179/1743278213Y.0000000095

Investigation of pH 2 S influence on 3% Cr tubing steel corrosion behaviours in CO 2 –H 2 S–Cl − environment

Abstract

Keywords

Introduction

Experimental

Materials

Test matrix and solutions

Equipment and procedures

Surface analysis

Results and discussion

Corrosion rate

Corrosion product analysis

Water chemistry analysis

Conclusions

Acknowledgements

References