Aqueous corrosion of plain carbon steel during sliding wear

Aqueous corrosion of plain carbon steel during sliding wear

167 Wear, 154 (1992) 167-176 Aqueous wear A. Hedayat corrosion of plain carbon steel during sliding and S. Yannacopoulos Deparhnent of Mechanical...

1MB Sizes 0 Downloads 16 Views

167

Wear, 154 (1992) 167-176

Aqueous wear A. Hedayat

corrosion of plain carbon steel during sliding

and S. Yannacopoulos

Deparhnent of Mechanical S7N OWO (Canada)

Engineering,

University of Saskatchewan,

Saskatoon,

Sask

J. Postlethwaite Department of Chemical S7N OWO (Canada)

Engineering,

University of Saskatchewan,

Saskatoon,

Sask.

S. Sangal Department of Mechanical S7N OWO (Canada)

Engineering,

University of Saskatchewan,

Saskatoon,

Sask.

(Received June 21, 1991; revised and accepted August 19, 1991)

Abstract

The effects of the cojoint action of corrosion and sliding wear are found in many industrial applications; for example in the area of sucker rods and tubing in slant and directionally drilled oil wells. A prototype apparatus has been designed and built to simulate the cojoint action of corrosion and wear on sliding steel components. Sliding wear tests were conducted on AISI 1045 steel samples in a 10% NaCl aqueous solution. The samples were subjected to various loads to study the effect of contact pressure on the wear-corrosion process. A set of tests were also run under the same mechanical conditions but in the presence of potassium chromate to inhibit corrosion. The effect of the corrosion inhibitor was very pronounced in the case of samples representing sucker rods. The experimental results show a marked drop in the material loss rates of these samples. Optical and scanning electron microscopy was used to study the characteristics of wear scars on the samples after testing. An explanation for the role of corrosion in the sliding wear of the steel samples is presented.

1. Introduction

The corrosion-wear of sucker rods and tubing has long been a problem associated with oil production [l, 21. The cojoint action of corrosion and wear on these steel components is much more destructive than the action of each by itself [3, 41. Mechanical wear occurs as a result of the reciprocating sliding contact of the sucker rod and tubing [3]. Corrosion, however, results from the combined presence of gases and liquids in the oil. Hydrogen sulphide, carbon dioxide and oxygen are the major corrodent gases encountered in oil fields [l, 5-71. The corrosivity of the gases is enhanced by the presence of water in the oil [8]. The water content is also responsible for wetting the metal surface. Thus, the lubricating effect provided by an oil film between the rods and tubing is diminished, and the components become more susceptible to mechanical wear [3]. Moreover, the presence of sand and other hard solid particles

0043-1648/92/$5.00

0 1992 - Elsevier Sequoia. All rights reserved

168

increases the wear of the sucker rods and tubing. Mechanical wear processes may remove protective films, resulting in an enhanced corrosion rate. Corrosion reactions may in turn produce hard abrasive particles, resulting in an enhanced rate of mechanical wear [9]. The injection of corrosion inhibitors down the oil casing is a common practice to control the corrosion of the sucker rods and tubing during oil production. Most of the inhibitors used in the oil fields are liquids of an organic nature, and are injected intermittently or continuously depending on well condition and requirement. The inhibitors form a chemisorbed film on the metal surface. The technique is most effective in the case of oil wells with high water contents. By being water repellent, the inhibitors decrease the corrosion rate of the sucker rods and tubing significantly [l, 9, lo]. The research conducted in the area of sliding wear of sucker rods and tubing in the oil industry so far has been limited to a qualitative description. While some studies focused on the serious damage caused by the cojoint action of corrosion and wear on these structures [l, 91, other studies summarized the practices used at the oil site for protecting these structures [7, lo]. However, none of the research work addressed the problem in a quantitative manner, nor explained in depth the role of corrosion in the process of sliding wear. In view of this, the objective of the research work presented here was to study the role of corrosion in the sliding wear of plain carbon steel of which rods and tubing are made. 2. Experimental

procedure

A variety of experimental machines have been developed to study the sliding wear of metals in aqueous corrosive media [ll-131. The design of these prototypes was either based on the concept of pin on disc [ll, 121, or on the concept of a reciprocating body in contact with a stationary body [13]. Since the corrosion-wear process in the oil field occurs as a result of the reciprocating motion of the sucker rod against the tubing, a prototype machine based on the concept of reciprocating motion was assembled. The test apparatus was designed to operate with test specimens having a variety of geometries. Although specimens with curved surfaces represent a more realistic simulation of the sliding of sucker rod couplings against the tubing wall in oil wells, the present tests were run using specimens with flat surfaces to enable the contact pressure between the two surfaces to be determined accurately. The geometries of the top (stationary) and bottom (reciprocating) specimens are shown in Fig. 1. The top specimens, representing the sucker rods, were 12.7 mm in diameter, while the bottom specimens, representing the tubing, were 110 mm in length

6

TOP

SPECIMEN

__

BOTTOM

SPECIMEN

Fig. 1. Geometries of the stationary (top) and oscillating (bottom) specimens.

169

and 19 mm in width. The test specimens were machined from hot rolled AISI 1045 steel. They were rinsed in methanol prior to coating the bottom and the sides with an insulating laquer to prevent corrosion. The surfaces in contact during sliding were left uncoated. The exposed surface areas of the top and bottom specimens were 127 mm2 and 2200 mm2 respectively. However, the wear areas on the bottom specimens depended on the stroke length and were therefore smaller than the actual exposed area. Throughout the tests, a stroke length of 85 mm was used, which resulted in a wear area on the bottom specimen of 1040 mm2. A sliding speed of 85 mm s-r was used which yielded approximately 54 strokes min-‘. The solutions used throughout the experiments were made with distilled water. In the first set of experiments, wear tests were carried out in an aerated 10% NaCl aqueous solution. In the second set, 500 ppm potassium chromate was added to the solution to inhibit corrosion. The applied normal pressures on the specimens were approximately in the range 60-800 kPa which correspond to a load range of 0.8-10.1 kg. In each chamber, 1.7 1 of the solution completely submerged the specimens. Weight loss measurements were made at periodic intervals for a duration of approximately 7 days for each specimen pair. Finally, optical and scanning electron microscopy was used to study the surface characteristics of the samples after completion of each test.

3. Experimental

results

The weight loss data were converted into thickness loss values and plotted as a function of the cumulative number of strokes. Figures 2(a) and 2(b) illustrate the change in thickness of the top and bottom specimens respectively, as a function of the number of strokes. These samples were tested in the absence of potassium chromate, and under various load conditions. Figure 2(a) shows that, as expected, the thickness loss in the upper specimen increased monotonically with contact pressure. However, as shown in Fig. 2(b), the thickness loss data of the bottom specimens did not exhibit such behaviour. The addition of potassium chromate to the corrosive aqueous solution in which the sliding wear tests of the steel samples took place, yielded interesting results. The change in thicknesses of the top and bottom specimens, as a function of the cumulative number of strokes, under various loading conditions is shown in Figs. 3(a) and 3(b) respectively. The total wear of both top and bottom specimens increased monotonically with pressure. The wear of the top specimens was very much decreased by the presence of the inhibitor. The wear of the bottom specimens, however, was of a similar magnitude to that obtained without the inhibitor. The effect of contact pressure on the average rate of thickness loss of specimens tested in the absence and presence of the inhibitor is shown in Fig. 4, where again the major influence of the corrosion inhibitor on the overall wear behaviour of the top specimens is illustrated.

4. Discussion

A comparison of the thickness loss rates as a function of contact pressure with and without inhibitor (Fig. 4) leads to several observations. In both cases, the loss rates of the small stationary top specimens were higher than those of the larger sliding

x ??S 0 I A

0

0

I

CUMMULATIVE

(a)

2 NUMBER

3

4 OF

STROKES/

kP0

431 kP0 562 kPa 211 kpS 4OkPa

5 IO’

N

*o \ b 2

X 0 . t, 0

IO

E

0 @I

;

CUMMULATIVE

779 451 362 2it 60

kPa kPa kPa CPa kPa

5 NUMBER

OF

STROKES

/ IO’

Fig. 2. Thickness loss vs. the cumulative number of strokes for samples tested in a 10% NaCl solution in the absence of potassium chromate under different applied normal pressures: (a) top specimens, and (b) bottom specimens. bottom

specimens at all times. This is attributed to the fact that in the process of sliding, the entire contact surface of the top specimen is continuously sliding against a fraction of the area of the bottom specimen at any given time. Thus, the top specimen experiences continuous sliding contact, while the bottom specimen experiences only intermittent sliding contact. This probably explains the fact that sucker rods fail more frequently in service than the tubing [2]. In the absence of the corrosion inhibitor, the ratio of corrosion-wear rates between the top and bottom specimens increased with the increase in contact pressure, as shown in Fig. 4. The coefficient of kinetic friction between the top and bottom specimens was determined experimentally as 0.2. The results in Fig. 4 show that a contact pressure of 60 kPa caused the top specimen, experiencing ~nt~uous sliding contact, to have a thickness loss rate about three times that of the bottom specimen, experiencing only

171 IO _ __ INTERMITTENT VALUES NOT AVAILABLE /X / / /+

/ / ’

-

/’

x t

704 666

0

539

. A 0

(a)

CUMMULATIVE

IO

NUMBER

OF

STROKES

km LPo

kPa 326 kPa 167 kF0 6OkPa

I IO’

INTERMITTENT VALUES NOT AVAILABLE

r

X 764 + 666 0 539 . 326 A 167 0 60

/ / / /

,+ /



kPa kPa kPa kPa kPo kPa

O 0 @I

I

CUMMULATIVE

2 NUMBER

3 OF

4 STROKES

5 /IO5

Fig. 3. Thickness loss ZIS.the cumulative number of strokes for samples tested in a 10% NaCl solution and 500 ppm potassium chromate under different applied normal pressures: (a) top specimens, and (b) bottom specimens. intermittent sliding contact. The application of a contact pressure of about 800 kPa, however, caused the top specimen to lose thickness at a rate about eleven times that of the bottom specimen. The above phenomenon could be explained as follows. The sliding of the top and bottom specimens against each other created scratches.on their surfaces, and the scratches provided fresh metal sites for corrosion action, specifically chloride pitting [14]. The continuous sliding of the specimens against each other did not allow for surface films to reform. Thus, bare metal sites remained exposed, and additional fresh metal sites were also furnished for corrosion action. As the contact load increased, the specimen surface was scratched to a greater extent, and more sites on the surface were vulnerable to corrosion attack in the form of pitting, Upon the addition of the corrosion inhibitor, the fresh metal sites become less vulnerable to pitting corrosion. This effect is illustrated in Fig. 4 where in the presence

172

-22 d

TOP

0

BOTTOM

)I

TOP

0

BOTTOM

I

I00

SPECIMEN

(WITHOUT

SPECIMEN

SPECIMEN

I

200

f WITHOUT

I WITH

SPECIMEN

1

400

CONTACT

Fig. 4.

1

INHIBITOR

INHIBITOR

(WITH

t

300

INHIBITOR

INHIBITOR

I

500 PRESSURE

)

1 I

E

I

600

1

700

I

600

!

0

900

/ kPa

Rate of thickness loss vs. applied normal pressure.

of the corrosion inhibitor and throughout the contact pressure range of 60-800 kPa, the top specimens, experiencing continuous sliding contact, lost thickness at a rate about 1.5 times that of the bottom specimens, experiencing intermittent sliding contact. It is also shown in Fig. 4 that the ratio of the difference in metal loss between the top and bottom specimens at 800 and 60 Wa in the absence of corrosion inhibitor as compared with that in its presence is 7 to 4 respectively. Moreover, the results in Fig. 4 show that the inhibition of corrosion led to a considerable drop in the rate of thickness loss of the top specimens. Thus, it is inferred that the corrosion component contributing to the total rate of thickness loss in the top specimens was substantial. The weight loss data for AISI 1045 steel immersed in a 10% NaCl solution without sliding wear showed that the steel corrodes at a rate of 0.4 mm per year. This rate is much lower than that caused by the cojoint action of corrosion and wear. Figure 4 shows that for the bottom specimen, the weight loss due to the cojoint action of corrosion and wear throughout the pressure range of 60-800 kPa was about 4.5 times that of corrosion only. In the case of the top specimen, at 60 and 780 kPa, the material loss due to corrosion-wear was about 14 and 50 times that of the corrosion action by itself respectively. Figures 5-10 are micrographs illustrating the role of corrosion in the sliding wear of the steel specimens. Figures 5(a) and S(b) are optical micrographs showing the main characteristics of the wear scars and pits on the samples subjected to a contact pressure of 60 kPa in the presence and absence of the corrosion inhibitor respectively. The samples tested without inhibitor, Fig. 5(a), show more intense wear tracks, pits, and scars than those tested in the presence of the corrosion inhibitor, Fig. 5(b). Figures 6(a) and 6(b) also show the difference between the characteristics of the samples

Fig. 5. Optical micrographs of a specimen tested under a 60 kPa contact pressure NaCl solution: (a) without inhibitor, (b) with 500 ppm potassium chromate.

(4

in a 10%

@I

Fig. 6. Optical micrographs of a specimen tested under a 362 kPa contact pressure in a 10% NaCl solution: (a) without inhibitor, (b) with 500 ppm potassium chromate.

Fig. 7. Scanning electron photomicrograph in a 10% NaCl solution.

of a specimen tested under a 60 kPa contact pressure

tested in the presence and absence of the inhibitor respectively, but subjected to a higher contact pressure of about 362 @a. It is clear that the wear sears and pits of the samples tested without the inhibitor, Fig. 6(a), are longer and more intense than those on the surface of the samples tested in the presence of potassium chromate, Fig. 6(b).

(4

(b)

Fig. 8. (a) Scanning electron photomicrograph of a specimen tested under a 60 kPa contact pressure in a 10% NaCl solution; (b) enlargement of the outlined area in (a).

Fig. 9. Scanning electron photomicrograph of a specimen tested under a 60 kPa contact pressure in a 10% NaCl solution with 500 ppm potassium chromate.

Fig. 10. Scanning electron photomicrograph of a specimen tested under pressure in a 10% NaCl solution and 500 ppm potassium chromate.

a 362 kPa contact

175

Scanning electron microscopy was used to study the characteristics of the wear scars and pits viewed on specimens subjected to various loading conditions in the presence and absence of the corrosion inhibitor. Figure 7 is a scanning electron photomicrograph showing equidistant pits formed along a wear track in a specimen tested in 10% NaCl solution without inhibitor, under a contact pressure of 60 kPa. It is proposed that the pits were formed by chloride attack. The proposition is based on the fact that chloride pitting occurs along scratching sites [12]. Figure 8(a) is a scanning electron photomicrograph showing a pit along a wear track on the surface of the specimen tested in the absence of the inhibitor and subjected to a contact pressure of 60 kPa. The pit shown is characterized and distinguished from its surroundings by being deep and having a wavy morphology. This is in contrast with the morphology of steel surfaces subjected only to sliding wear at low loads. Under such conditions, asperities are first removed, and then delamination of the surface takes place. This causes the whole area subjected to sliding wear to be rough and wavy [15]. It can be inferred from the morphology of the pit, that under these testing conditions, mechanical wear provides a nacent surface for corrosion to proceed. The small longitudinal pits along the side of the pit, which are magnified in Fig. 8(b), illustrate the mechanism by which the pit grows. The marginal pits grow during sliding by mechanical wear or dissolution until they become part of the main pit. Figure 9 is a scanning electron photomicrograph showing a wear scar on the surface of a specimen subjected to a contact pressure of 60 kPa, and tested in the presence of the corrosion inhibitor. The wear scar is characterized by being shallow, which is evidence of the diminished role of corrosion action. Figure 10 is a scanning electron photomicrograph of a specimen tested in the presence of the corrosion inhibitor, subjected to a contact pressure of about 362 kPa. The micrograph shows a wear scar deeper than that shown in Fig. 9 owing to the increased contact pressure. However, the morphology of the scar was smooth compared with that of the pit shown in Fig. 8. It can be deduced that the corrosion inhibitor covers the nacent metal surface furnished by mechanical wear with a protective layer against corrosion. This layer is renewable as mechanical wear proceeds. Thus, during the progressive formation of the wear scar, areas of the sliding surfaces come in contact, while others do not. The surfaces that come in contact are subjected to inhibitor film removal and delamination, while the areas that are not in contact maintain the inhibitor film on their surfaces. That explains the intermittent nature of the wear scars marked “X” that are shown in Figs. 5(b) and 6(b).

5. Conclusions

(1) The cojoint action of corrosion and sliding wear is much more destructive than the action of corrosion by itself. (2) The material loss rates due to the combined effects of wear and corrosion of carbon steel experiencing continuous sliding contact in aqueous sodium chloride solutions are much higher than the rates for carbon steel experiencing only intermittent sliding contact. (3) The material loss rates were found to increase monotonically with an increase in the contact pressure for the specimens in continuous sliding contact, with and without inhibitor. The results for the carbon steel experiencing intermittent sliding contact showed such an effect with inhibitor present, but not in the absence of inhibitor. (4) The presence of the inhibitor greatly reduced the rate of material loss for the carbon steel experiencing intermittent sliding contact.

176 Acknowledgments

This project was made possible through funds from Amoco Company Ltd. and in part by NSERC Canada.

Canada

Petroleum

References 1 J. D. Sudbuty

2 3 4 5 6 7 8 9 10 11 12

13 14 15

ted.), Corrosion of Oil- and Gas-Well Equipment, NACE and API, Dallas, TX, 19.58. D. Bellow, D. Owens and I. Smuga-Otto, Wear of standard and hard-metal-coated couplings with oil field tubing, Wear, 133 (1988) 281. K. P. McCaslin, A study on the methods for preventing rodwear leaks in sucker rod pumping wells (Society of Petroleum Engineers), prad Eng., (November, 1988) 615. A. Batchelor and G. Stachowiak, Predicting synergism between corrosion and abrasive wear, Wear, 1.23 (1988) 281. Task Group T-l-3, CO, Corrosion in Oil and Gas Production, NACE, Houston, TX, 1984. R. N. Tuttle and R. D. Kane (eds.), H$ Corrosion in Oil and Gas Production, NACE, Houston, TX, 1981. M. Altunbay, Method determines best way to control downhole corrosion, World Oil, (November, 1985) 63. R. J. Jasinsk and K. D. Efird, Electrochemical corrosion measurements in crude oil, Corrosion, 43 (8) (1987) 477. NACE Group Committee T-1, Corrosion Control in Petroleum Production, NACE, Houston, TX, 1979. C. C. Nathan (ed.), Corrosion Inhibitors, NACE, Houston, TX, 1973. H. Abd El-Kader and S. El-Raghy, Electrochemical kinetics of stainless steel in aerated chloride solutions during wear process, Electrochim. Acta, 30 (7) (1985) 841. S. Bhattacha~a, V. S. Agarwala and K. Y. Kim, Wear corrosion of steels under cyclic operation in corrosive liquid environment. In K. C. Ludema (ed.), Wear of Mareriafs, Znt. Conf. on Wear of Materials, ASME, New York, 1985. K. Miyoshi and G. Rengstroff, Wear of iron and nickel in corrosive liquid environments, Corrosion, 45 (4) (1989) 266. G. T. Burstein and D. H. Davies, The effects of anions on the behaviour of scratched iron electrodes in aqueous solutions, Corros. Sci., 20 (10) (1980) 1143. S. Jahanmir and N. Suh, Surface topography and integrity effects on sliding wear, Wear, 44 (1977) 87.