An investigation of the formation and destruction of corrosion inhibitor films using electrochemical impedance spectroscopy (EIS)

Corrosion Scieme,Vol. 38, No. 9, pp. 1545-1561, 1996 Copyright 0 1-996Published by Elsevier Science Ltd Printed in Great Britain. All rights reserved 0010-938X/% $15.00+0.00

PII: !30010-9~x(%poo47-9

AN INVESTIGATION OF THE FORMATION AND DESTRUCTION OF CORROSION INHIBITOR FILMS USING ELECTROCHEMICAL IMPEDANCE SPECTROSCOPY (EIS) Y. J. TAN, S. BAILEY and B. KINSELLA Western Australian Corrosion Facility, School of Applied Chemistry, Curtin University of Technology, GPO Box U1987, Perth, WA 6001, Australia AIrstract-Electrochemical impedance spectroscopy (EIS) was used to study the film formation and destruction and the corrosion protection mechanism of a carbon dioxide (COs) corrosion inhibitor, imidazoline. Imidazoline is an inhibitor base which is most commonly used for protecting oil wells, gas wells and pipelines from CO1 corrosion. Typical EIS spectral changes were clearly observed during the inhibitor film formation and destruction processes, and this suggests that EIS is a practical technique for evaluating inhibitor film persistency and for monitoring the inhibitor film performance. EIS data were used to calculate corrosion related electrochemical parameters and EIS was shown to be a very useful tool for studying inhibitor mechanisms. EIS results showed that imidazoline is a very effective CO2 corrosion inhibitor which forms a chemically bonded film on the metal surface. The inhibitor film seems to have a multi-layered structure which is a combination of an inner layer, which may be an inhibitor-metal complex layer, and many outer layers of inhibitor molecules. The inhibitor film showed strong self-repairing ability and good persistency under the testing conditions although surface water shear stress gradually removed the inhibitor film and caused inhibitor film failure after around 170 h under the test regime. Copyright 0 1996 Published by Elsevier Science Ltd

INTRODUCTION Inhibition is the most cost effective and flexible means of corrosion control in the oii and gas production industry. There are different methods of inhibitor application such as continuous treatment and batch treatment. Continuous treatment is the most common technique which involves simply injecting an inhibitor into oil or gas well fluid by a chemical proportional pump and maintaining a certain inhibitor concentration in the fluid for protecting against CO2 corrosion in oil and gas wells and flowlines.’ However, in practice, the selection and application of inhibitors are actually complicated because the corrosion environments in oil and gas fields are very variable. An inhibitor that works in one oil or gas well may not work in another well due to the differences in the corrosion environment. Suitable corrosion monitoring techniques which can quickly evaluate and monitor inhibitor efficiency are very important requirements for the proper application of inhibitors. There are many techniques that can be used in the laboratory and in the field for testing inhibitor efficiency, such as weight-loss coupons and the resistance probe. However, there is not a good technique which can be used to evaluate inhibitor film persistency. The current

Manuscript received 11 September 1995. 1545

1546

Y. J. Tan et al.

technology for assessing inhibitor film persistency using the wheel test is a subject of much discussion and, in many cases, disagreement.3 The persistency of an inhibitor film is an important aspect of inhibitor assessment. When applying an inhibitor in oil and gas wells and flowlines, a most important concern is to make sure that an inhibitor film is formed with the anticipated film persistency.4 Even when inhibitor treatment is continuous, an inhibitor may not really be continuously applied on some parts of pipeline, e.g. valves in the pipeline may plug and the well may need to be shut-in for service.4 When an inhibitor is applied at a low concentration (e.g. 10 ppm), it is difficult to maintain an adequate inhibitor concentration everywhere over a long pipeline. Good film persistency therefore remains an important requirement for inhibitors. The understanding of CO2 corrosion inhibitors is limited, especially since very little is known about the inhibitor film formation and destruction mechanisms. Obviously a better understanding about the inhibitor film formation and destruction processes is important for the development of new CO* corrosion inhibitors which have better film persistency. Electrochemical impedance spectroscopy (EIS) as a new tool for corrosion study has already been successfully used in various corrosion and protection fields such as organic coatings studies, passive layer analysis and water treatment corrosion inhibitor evaluation. It has already shown its power in providing information on the corrosion and protection mechanisms, especially when an adsorbed film or an applied organic coating is present.5.6 The application of EIS for the evaluation of oil field corrosion inhibitors is rarely reported in the literature although reports in recent publications have shown the possibility of this application.7*8 Very little attention has been paid to the potential possibility of EIS applications in the monitoring of inhibitor film persistency and in the study of inhibitor mechanisms. In this work, EIS has been used to study the inhibitor film and the corrosion electrochemical kinetics under the inhibitor film. A typical commercial carbon dioxide (COz) corrosion inhibitor, imidazoline, was studied. Imidazoline is an inhibitor base which is most commonly used for protecting oil wells, gas wells or pipelines from COz corrosion although little has been reported about its corrosion protection mechanism and film properties. The possibilities of using EIS as a method for evaluating inhibitor film persistency and for studying inhibitor mechanisms have been explored. EXPERIMENTAL

METHOD

The electrochemical cell for EIS measurements is schematically shown in Fig. 1. A mild steel cylindrical electrode with total surface area of 3.2 cm* was used as a working electrode for all electrochemical tests. The surface of the electrode was polished with 400 and 800 grit Silicon Carbide paper and cleaned with ethanol and isopropanol. An Ag/AgCl electrode was used as a reference electrode. A platinum electrode with large surface area was used as counter electrode. The working electrode was fitted to an electrode rotator which was controlled by a speed controller (Model MSRX, Pine Instrument Company). An EG and G Princeton Applied Research Potentiostat/Galvanostat Model 273A and a 5210 lock-in-amplifier with software M398 were used for EIS measurements. EIS measurements were carried out at open circuit potential with amplitude of 5 mV in the frequency range from about 1 mHz to 100 kHz.

Formation and destruction of corrosion inhibitor films

1547

Electrode Rotator

Counter

Electrode eference electrode

CO2 sparging tube

-Mild

steel cylinder electrode

Stirring Bar Digital Hot Plate I Stirrer

Fig. 1. A schematic diagram of electrochemical cell set-up

The basic testing solution was 640 mL of 3% NaCl brine (made by adding AR NaCl into milli-Q water). AI1 testing fluids were pre-sparged with carbon dioxide before use. The inhibitor used in this study was a commercial grade imidazoline which has a molecular structure as

N\

R = Cl6

- CIB

Chain

CH2 / c2 B NH2

The working electrode was filmed in a filming fluid (made by adding 50 ppm imidazoline inhibitor into the base 3% NaCl brine). EIS measurements were carried out regularly in the filming fluid for monitoring the film formation process (47 h testing period). The filming fluid was then diluted to 5 ppm and the inhibitor filmed electrode was tested for investigating

Y. J. Tan et al

1548

the inhibitor film behaviour when inhibitor concentration is very low. EIS measurements were carried out regularly in the 5 ppm inhibitor fluid (22 h testing period). The electrode was then quickly transferred into a inhibitor-free fresh brine for investigating inhibitor film destruction by immersing the electrode in the pure brine (20 h testing period). The electrode was stationary during all of the above processes. Finally the electrode was rotated at 1000 rpm for investigating the effects of fluid shear stress on inhibitor film (85 h testing period). The whole experiment took about 170 h. EXPERIMENTAL

RESULTS

AND DISCUSSION

The formation of an inhibitor film The mild steel electrode was placed in a brine solution containing 50 ppm of imidazoline inhibitor for filming. EIS measurements were carried out regularly for monitoring the inhibitor film formation process. Figure 2(a),(b) shows EIS plots (in Nyquist format and in Bode 6 vs log f format) recorded on the electrode after different filming times. As shown in Fig. 2 (a), the continuous increase in the diameter of Nyquist semi-circles during the 42 h of filming process suggests that the presence of inhibitor greatly but gradually changed the corrosion kinetics on the electrode surface. The Bode 8 vs logfplots which were plotted using the same experimental data in the Nyquist format, as shown in Fig. 2 (b), show a new phase angle shift at higher frequency range and a continuous increase in the phase angle shift with filming time. This new phase shift means that the formation of inhibitor film changed the electrode interfacial structure and resulted in an extra time constant. The continuous increase in the phase angle shift obviously correlates with the inhibitor film growth. Two peaks in the 6 vs logfplots mean that there are two major electrochemical kinetic processes on the electrode surface. In the Nyquist plots, as shown in Fig. 2 (a), two semicircles appear, one of which is much smaller than the other one and is not clearly resolved. The semicircle at lower frequency, which shows peaks around 1 Hz in Fig. 2 (b), is due to the corrosion electrochemical process. This is consistent with the suggestions of other researchers5,9 and has been confirmed using weight-loss data by the authors’e The semicircle at high frequency range would be due to the inhibitor film because a surface dielectric film normally has a small time constant and so has a phase angle shift in the high frequency range. 1’-13 The impedance characteristics of this electrode surface could be simulated by an electrical circuit used by a number of workers for coated metal electrodes,‘4.i5 as shown in Fig. 3. However, as shown in Fig. 4(a) and (b), the semicircle fitting according to the model in Fig. 3 obviously does not fit the data points between the high and low frequency semicircles although the poor fitting of very low frequency data points can be explained as a pseudoinductance effect which was suggested to be due to an adsorption-desorption phenomenon of the inhibitor in the vicinity of the corrosion potential.‘d-‘8 The poor fitting of the data points at intermediate frequencies suggests that there are some other interfacial structures on the electrode surface which have intermediate time constants. The large phase angle shift in the higher frequency range in Fig. 2 (b) also suggests a complicated interfacial structure. A likeiy explanation is that the inhibitor film is a multi-layered film and so shows more than one time constant. An equivalent circuit is suggested in Fig. 5 for an electrode with a fourlayer inhibitor film.

Formation and destruction

ofcorrosioninhibitor

1349

Urns

Zlm (ohms)

4txM 3500 3ow 2!500 2QOO 1500 mxl

so0 0 -500

-mm Zrr (ohms)

-10

I

1 E-2

lE-4

J

( lE+O

lE+Z

lE+4

FrocEueneywz, krg =m

Fig. 2.

EIS plots during the filming process of the inhibitor.

Y.J. Tan et

1550

al.

Cfilm

Rt Fig. 3.

Equivalent

circuit for an electrode

filmed with a non-conducting

inhibitor

film.

The spectra were fitted satisfactorily using the four-layer model by a non-linear least square fit (NLLS-fit) and simulation computer program.” Figure 6 shows the fitting of the typical Nyquist plot in Fig. 4. The electrode surface electrochemical kinetics parameters, including the resistances (RtR4) and capacitances (Ct-Cd) of inhibitor layers l-4. The electrochemical charge transfer resistance (R,), and double layer capacitance (Cd,) can be deduced using the NLLS-fit and simulation program, which provides an insight into the properties of the inhibitor film and the electrochemical double layer. Table 1 shows the electrode surface electrochemical kinetics parameters during the formation of the inhibitor film. The satisfactory simulation of the impedance characteristic of the inhibitor-filmed electrode surface by a four-layer model, as shown in Fig. 6 and Table 1, indicate that the inhibitor film has a multi-layered structure. The resistance values obtained for the four inhibitor layers are different from each other, which suggests that each inhibitor layer has a different inhibitor molecular density because the inhibitor layer resistance value is a

Table 1. The continuous charge transfer resistance

Filming (h)

Before filming 0.5 2 3 4 8.5 13 15 24 28 33 31 42

changes of the resistances and capacitances of inhibitor layers (RI-R4 and Cl-Cd). (R,), and capacitance (Cd,) during inhibitor film formation. All measurements were carried out at 30°C and with CO> sparging

R, (ohms)

C‘dl (PF)

RI (ohms)

Cl (/IF)

R2 (ohms)

C2 (PF)

R3 (ohms)

c3

R4

G

(PF)

(ohms)

W)

15 122 207 460 483 1119 2458 3506 3933 4751 4741 5020 5246

1690 951 746 I 1678 4003 1717 283 241 237 186 223 220 213

159 398 649 845 2082 I736 2110 2312 1619 2538 2690 2732

193 407 211 244 169 98 91 89 57 77 74 72

63 257 205 321 600 565 653 719 424 827 875 880

119 160 103 Ill 63 34 34 33 19 28 27 27

10 57 49 77 200 178 198 221 102 235 250 249

83 96 39 36 22 13 13 13 10 11 II 11

2 9 20 31 62 61 63 69

22 I5 12 12 11 11 10 11

69 72 72

10 10 10

Formationand destructionof corrosioninhibitorfilms 4.0

3.0

2.0

E ._

1.0

N

0.0

-1.0

7.0

5.0

3.0 Zre (kohm) I

500.0

l-

I

I

I

.

W

Data Points 400.0 lz S E ._ N

.

300.0

. .

200.0

/

,A

/-~----~-----~--

--‘\.,_

Fitted Semicircle Y \

100.0

.\ \

0.0 L

:

I

100.0

I

I

300.0

500.0

\

% \ :, l

700.0

L

900 .a

Zre (ohm)

Fig. 4.

(a) Semicircle fitting of a typical Nyquist plot using the equivalent circuit in Figure 3. (b) Enlargement of semicircle fitting of high frequency data points in Figure 4(a).

reflection of the penetration of the inhibitor layer by electrolyte and so is related to the inhibitor molecular density in the layer. The first layer has the largest resistance and the fourth layer has the lowest resistance, i.e. the first inhibitor layer should have the densest inhibitor molecular structure and the fourth layer has the lowest inhibitor molecular density. Figure 7 shows linear relationships between the resistance of the first layer (RI) and

1552

Y. J. Tan er al.

Rt Fig. 5.

An equivalent circuit for an electrode filmed with a four-layer, non-conducting inhibitor film.

the resistances of other layers (R,, R3, R4) during the inhibitor filming process. These linear relationships suggest that the molecular structure in each layer is cIosely related. During the first 8.5 h inhibitor filming, as shown in Table 1, the resistances of inhibitor layers quickly increased (10 times increase). After 8.5 h, the resistances of the inhibitor layers only increased slightly. The capacitances of inhibitor layers also quickly changed in the first

Zim (ohms)

3500 3ooo 2500

m l!Too

1000 500 0 -500 -1000

Fig. 6.

Computer simulation of a typical Nyquist plot using a four-layer model.

Formation and destruction ofcorrosion

inhibitor films

1553

Rl (ohms) Fig. 7.

The relationships between the resistance of the first iayer (R,) and the resistances of other layers (R2, Rj, R4) during the inhibitor filming process.

8.5 h and remained relatively stable after 8.5 h. These results suggest that the main structure of the inhibitor film was built up in the first 8.5 h. However, after the first 8.5 h quick increase, the charge transfer resistance (R,) maintained a continuous increase (from 1119 ohms to 5245ohms) during the 42 h inhibitor film formation. Figure 8 shows the relationship between Rt and the first layer resistance (RI), which shows a linear relationship for the first 8.5 h after which Rt showed a much faster increase than RI. These results suggest a continuous change in the electrochemical double layer. An explanation is that a slow chemical reaction occurs at the electrode surface, forming a chemicaily bonded thin surface film which results in a continuous increase in charge transfer resistance. The fact that the electrochemical equilibrium is achieved only very slowly may suggest that a slow surface chemical or chemisorptive reaction rather than an adsorption equilibrium process is involved because an adsorption equilibrium would be likely to be achieved much faster. As discussed above, a physical model can be suggested as shown in Fig. 9. The inhibitor film is supposed to be a combination of an inner layer (the first layer) which would basically be an inhibitor-metal complex and several outer layers (the second, the third and the fourth layers) which would be inhibitor layers of different molecular density with possible inhibitor molecular cross-linking. The changes in double layer capacitance (Cd) and inhibitor layer capacitances also give information about the inhibitor film formation. The capacitance of a parallel-plate

Y. J. Tan et al.

1554

Rt (ohms) 6000 T

Rl (ohms)

Fig. 8.

The relationship

between double layer resistance and the resistance during the inhibitor filming process.

of the first layer (RI)

__~~____-___-------___________--__----

Solution

4th Layer

3rd Layer 2nd Layer 1st Layer

I

I--._

Mild Steel Electrode ._._ -___________-----_----

Fig. 9.

A physical

__---------______~__;--_ --,-----._________J

model for an electrode

filmed with a four-layer,

non-conducting

I

inhibitor

film.

Formation and destruction of corrosion inhibitor films

1555

capacitor is related to the dielectric constant of the insulator, the area of the plate and the distance of separation between plates. Because the area of the electrode surface and the thickness of layers can be assumed to remain unchanged so the changes in capacitance should be due to the changes in the dielectric constant of the layers, which is related to the formation of chemical bonded film and the molecular density in inhibitor layers. For example, a decrease in the capacitances of inhibitor layers suggests a denser inhibitor molecular structure in the layers. The inhibitor film destruction in a diluted inhibitor brine (5 ppm) and in an inhibitor-free fluid

After about 48 h filming, the inhibitor fluid was diluted from 50 to 5 ppm inhibitor concentration for studying the behaviour of the inhibitor film at low inhibitor concentration. The electrode was always kept in the solution phase. This test simulates the situation where inhibitor concentration drops for some reason such as blockage of the inhibitor supply valve. Figure 10(a), (b) show EIS plots (Nyquist plots and Bode 0 vs logf plots) recorded on the electrode after different immersion times. As expected, the inhibitor layer resistances (RI-&) and electrochemical charge transfer resistance (R,) immediately dropped after diluting the inhibited brine, as shown in Table 2. However, it was not expected that the electrode impedance increases again, as shown in Fig. 10 (a), (b) and Table 2, which may mean the inhibitor film is recovered by some means. A similar result also was observed after moving the testing electrode into an inhibitor-free 3% NaCl solution, as shown in Fig. 11 and Table 3. Although the inhibitor layer resistances and charge transfer resistance experienced large changes, as shown in Tables 2 and 3, the inhibitor layer capacitances and electrochemical double layer capacitance only had relatively small changes after the electrode was transferred to 5 ppm and 0 ppm inhibitor brine. This suggests that the main structures of inhibitor film and electrochemical double layer have not been changed. A possible explanation for the sudden decrease of film resistance and electrochemical charge transfer resistance may be localised breakdown of the inhibitor film by an unknown mechanism. In that case, the resistances in ‘pore’ areas of the inhibitor film, which is much smaller than the inhibitor film resistance, will be measured, resulting in a very small resistance value. The inhibitor film recovery, which is indicated by the increase of inhibitor film and charge transfer resistance in both 5 ppm and 0 ppm inhibitor brine, may be due to a selfTable 2. The continuous changes of the resistances and capacitances of inhibitor layers (RI-R4 and Cr-C,), charge transfer resistance (R,), and capacitance (Cd,) after diluting the inhibited brine to Sppm inhibitor concentration. All measurements were carried out at 30°C and with CO2 sparging Testing Hours

Before diluting 2 3 4 8 13 18 22

RI

Cl

R2

(pF)

(ohms)

(p)

(ohms)

(pF)

R3 (ohms)

213 419 861 246 250 247 251 293

2132 266 889 901 793 847 868 835

880 102 396 274 276 309 321 329

27 31 56 35 31 31 30 31

249 38 115 76 83 92 94 99

cdl (otk)

5246 225 504 1383 1104 1150 1160 966

12 97 134 17 70 71 71 76

c2

c3 WI

11 14 24 13 12 13 13 14

% (ohms)

12 4 39 28 29 33 32 34

G W

10 14 12 12 12 12 12

Y.J. Tan et 01.

1556

Zlm (ot\msl 4000 3500 efore dilutingthe inhibi 3000 2500 2000 1500 1000 500 0 -500 Zre (ohms)

-1000 1

- Phase (deg) 60 Before di!utiig the inMitnr 50

Frequency (Hz, log scale)

Fig. 10.

EIS plots after exposing

the fihned electrode

into a 5 ppm inhibitor

fluid.

Formation and destruction of corrosion inhibitor films Zlm

1557

(ohms)

?A

11 hrs

Zre (ohms)

Fig. 11. EIS Nyquist plots after transferring the filmed electrode into an inhibitor-free brine.

Table 3. The continuous changes of the resistances and capacitances of inhibitor layers (RI-R, and C,-C,), charge transfer resistance (R,), and capacitance (Cd,) after transferring the inhibitor filmed electrode into an inhibitor-free fresh brine. All measurements were carried out at 30°C and with CO2 sparging Testing

RI (ohms)

cdl

RI

Cl

R2

c2

R3

c3

R4

c4

(pF)

(ohms)

(gF)

(ohms)

@F)

(ohms)

(PF)

(ohms)

(PF)

Before diluting 0.5 2 7

966 225 739 1138

293 340 299 284

835 208 557 833

76 103 93 77

329 77 159 251

31 40 39 35

99 27 45 59

14 13 14 13

34 11 14 18

12 13 14 13

11 16 20

1177 1141 1106

282 300 268

867 888 814

74 73 67

268 280 258

34 34 34

62 62 57

13 13 12

20 21 20

12 12 12

@I

repairing (self-sealing) effect of the inhibitor film. The multi-layer characteristic of the inhibitor film, as discussed above, makes the self-repairing possible. The self-repairing characteristic is evidence of inhibitor multi-layer structure. The CO1 corrosion product may also contribute to the sealing of the inhibitor film. As the inhibitor film is broken down locally, the corrosion rate increases greatly in these small areas and a protective corrosion product may thus form in these local ‘pore’ areas. The fact that the charge transfer resistance and inhibitor layer resistances in 5 ppm inhibitor brine after about 22 h, as shown in Table 2, are similar to those in inhibitor-free brine after about 20 h, as shown in Table 3, suggests that the electrochemical kinetics and inhibitor surface coverage of the electrode surface are independent of inhibitor concentration. Thus, it can be concluded that the inhibitor does not obey adsorption laws and the inhibitor film is not a adsorptive film, but rather that a chemisorbed layer is formed. Inhibitor film destruction by surface shear stress

Finally, the electrode was rotated at 1000 rpm for investigating inhibitor film destruction by surface shear stress. Figure 12(a), (b) shows EIS Plots (Nyquist plots and bode 8 vs log f plots) recorded on the electrode at different times.

Y.

15%

J.Tan

erui

Zlm (ohms)

-before +I1

hrs

+20

hrs

+39hrs

Before rotating the electrode

\

-53

hrs

-+--85

hrs

400 300 MO 100

4

0 -100

2000

1500

2500

3Oa Zre

(ohms)

(b) - Phase (deg)

-+-before +ll

0

0

1

10

100

1000

Frequency Fig. 12.

-&-

20 hrs

+

39 hrs

-

53 hrs

-t

05 hrs

/

-10 4

0

hrs

EIS plots after rotating

the filmed electrode

loo00

100000

(Hz, log scale)

at 1000 rpm in the inhibitor-free

brine

The continuous decrease in diameter of the Nyquist plots and the gradual disappearance of the high frequency phase angle shift, as shown in Fig. 12(a) and (b), means that the inhibitor film is gradually removed by surface shear stress. The film destruction process was very slow, suggesting a strong interaction between the electrode and the inhibitor. On

Formation and destruction of corrosion inhibitor films

1559

comparison of Figs 12(b) and 2(b), very similar trends are seen, but in the opposite time series. This suggests that the film destruction and film formation follow a ‘reversible’ process. Corrosion rates can then be estimated using Rt data and the Stern-Geary equation5: I

1 1 bob, b,b, ‘Orr= 2.303(b, + b,) x i& = 2.303(b, + b,) x R,

In this equation, the non-steady state parameter R, was used instead of the steady state parameter R, because the R, is not always determinable in the frequency range used. From the point of view of industrial application, the error due to this is negligible. The Tafel slopes, b, and b, were estimated by corresponding linear polarisation measurements using the PARCalc routine which extracts the b, and b, values by performing a non-linear leastsquares fit of the linear polarisation data to the Stern-Geary equation.20 The electrochemical parameters and corrosion rates are calculated and listed in Table 4. The inhibitor failure processes is clearly shown in Fig. 13(a), (b) and in Table 4. This means that the inhibitor film persistency (the time before inhibitor failure) can be assessed using EIS to give data such as those shown in Fig. 13. CONCLUSIONS 1. EIS has been shown to be a valuable technique for studying the mechanism of inhibitor film formation and destruction, and for evaluating the film persistency of CO2 corrosion inhibitor. EIS can be used to measure corrosion related electrochemical parameters such as the resistances and capacitances of inhibitor layers, charge transfer resistance and double layer capacitance. These parameters can be used to analyse the inhibitor mechanism and to calculate the corrosion rate.

Table 4. The continuous changes of the resistances and capacitances of inhibitor layers (RI-& and Cr-C,), charge transfer resistance (R,), and capacitance (Cdl) after rotating the electrode at 1000 rpm in the fresh brine. All measurements were carried out at 30°C and with CO2 sparging Testing (h)

Rt (ohms)

cdl

RI

C,

R2

C2

R3

C3

R4

C4

(PF)

(ohms)

(PF)

(ohms)

(PF)

(ohms)

(PF)

(ohms)

(PF)

Corr. rate (mm/y)

Before stirring 2 6 11 15 20 29 34 39 43 53 70 75 80

1106 428 353 356 322 283 219 173 128 105 68 56 38 24

268 424 396 347 386 447 526 712 941 1122 1497 1609 2136 2894

814 459 373 341 312 291 226 198 168 142 99 92 59 41

67 82 86 81 87 97 120 142 162 186 231 261 340 445

258 181 130 92 73 62 42 42 43 34 25 23 17 13

34 36 45 47 53 59 69 90 114 139 185 213 280 336

57 46 35 27 22 17 10 7 7 5 4 3 2 2

12 13 14 12 15 17 30 42 59 65 99 111 143 230

20 19 15 9 9 7 5 4 3 2 2 1 1 0.5

12 12 13 12 12 12 12 12 13 13 14 14 19 33

0.02 0.04 0.05 0.05 0.06 0.06 0.08 0.10 0.29 0.35 0.54 0.66 0.97 1.54

1’. J. Tan PI al.

Corrosion

Rate (mm/y)

1.6T 1.4-1.2-1 -0.8-0.6--

0

40

20

60

80

Electrode Rotation Time (hrs)

Fig. 13. Inhibitor film failure after rotatmg the filmed electrode at 1OOOrpmin the inhibrtor-free brine.

2. lmidazoline was shown to be a very effective CO2 corrosion inhibitor. It forms a chemically bonded film on the metal surface. The inhibitor film seems to be a combination of an inner-layer which is likely to be an inhibitor-metal complex and several outer-layers which are likely to be inhibitor layers with possible inhibitor molecular cross-linking. The inhibitor film showed a strong self-repairing ability, but surface water shear stress can gradually remove the inhibitor film and cause inhibitor film failure. Ackno~,,led~entmr.s-The authors thank Jean Edward for her assistance in the preparation of experiments, and Yadran Marinovich for his assistance with computer packages.

REFERENCES I. S. Webster. Corrosion Inhibitor Selection for Oil field Pipelines. (hrrasion 93 /AfACE/. Paper 109 (1993). 2. A. Nestle. Corrosion Inhibitors in Petroleum Production Primary Recovery. Corrosion 1nhibitor.r. C. C. Nathan (ed.) NACE (Houston) (1973). 3. NACE Task Group T-l D-8, Whppl Test Procedure Used/or Ew/uuting Film Persistetzt Inhibitors for Oilfield Applications. NACE Publication 1I3182, item No. 54238 (1982). 4. Harold Hilliard, Use qf’lnhk5itor.r For Downhole Corrosion Control In Gus Wells, CO2 Corrosion in Oil and Gas Production - Selected Papers, NACE Task Group T-l-3 (ed.), P. 357 (1978). 5. D. C. Silverman and J. E. Carrico, Corrosion 44(S). 284 (1988). 6. F. Mansfeld. Corrosion 4d(lZ), 856 (1988). 7. Huey J. Chen. Evaluation of Oilfield Corrosion Inhibitors by EIS, Corro.rion 94 (NACE), paper No. 92, 1994. 8. J. L. Dawson ef al.. ‘Electrochemical Measurements for Inhibitor Assessments, Corrosion 93. paper No. 108. NACE (1993). 9. D. C. Silverman, Corrosion 6(7), 589 (1990).

Formation and destruction of corrosion inhibitor films

1561

10. Y. J. Tan, S. Bailey and B. Kinsella, An Experimental Comparison of Corrosion Rate Measurement Techniques: Weight-loss measurement, Linear Polarization, Electrochemical Impedance Spectroscopy and Electrochemical Noise Analysis, Corrosion and Prevention 95, Australasian Corrosion Association, Perth (1995). 11. S. Feliu, J. C. Gahan and M. Morcillo, Corros. Sci. 30(10), 989 (1990). 12. C. H. Tsai and F. Mansfeld, Corrosion 9(49), 727 (1993). 13. I. Thompson and D. Campbell, Corros Sci 1(36), 187 (1994). 14. L. Beaunier, I. Epelboin, J. C. Lestrade and H. Takenouti, Surf. Technol. 4(3), 237 (1976). 15. F. Mansfeld and C. H. Tsai, Corrosion 12(47), 958 (1991). 16. I. Epelboin, M. Keddam and H. Takenouti, J. Appl. Elecfrochem. 2(l), 71 (1972). 17. D. C. Silverman, Corrosion 4S(lO), 824 (1989). 18. W. J. Lorenz and F. Mansfeid, Corros. Sci. 21(9), 647 (1981). 19. B. A. Boukamp, Equivalent Circuit (EQUIVCRT. PAS), Version 3. 96, Second edition (1989). 20. EG & G Princeton Applied Research, ‘Model 352/252 SoftCorr TM II Corrosion Measurement and Analysis Software: User’s Guide’, EG and G Instruments Corporation (1993).