WO2000034759A1

WO2000034759A1 - Corrosion monitoring

Info

Publication number: WO2000034759A1
Application number: PCT/GB1999/004048
Authority: WO
Inventors: David Anthony Eden
Original assignee: Integriti Investments Limited
Priority date: 1998-12-07
Filing date: 1999-12-07
Publication date: 2000-06-15
Also published as: GB2367630A; AU1666300A; GB0113862D0; CA2353937A1

Abstract

A method of monitoring general and localised corrosion of plant structures using a single diagnostic electrode probe referenced to the plant structure and analysing its responses to imposed stimuli to provide indications of the corrosion behaviour of the plant structure.

Description

CORROSION MONITORING

This invention relates to corrosion monitoring of process or production plant in circumstances where the metal surfaces of the plant are, or may in the course of plant operations be, exposed to a liquid environment that is corrosive to them. The liquid environment may be contained, conveyed or stored process liquid or it may be heat-transfer media or even transient condensate. Usually, the liquid will be aqueous. More particularly the invention is directed to electrochemical methods of monitoring such corrosion in which the liquid environment to which plant surfaces are or may become exposed serves at pre-selected locations as an electrolyte of an Electrochemical Corrosion Cell . Electrochemical methods known in the prior art investigate responses of corroding electrodes under chosen conditions of imposed potential or current flow and include those known as polarisation; linear polarisation resistance; electrochemical impedance; electrochemical noise (potential or current noise); and harmonic analysis. All of these investigative methods are part of the armoury of the skilled corrosion engineer and the underlying science of each is now well established. For practical applications in industrial plant preference will be given to methods that intrude as little as possible, both in the sense of avoiding the imposition of irreversible changes in reference material and in the sense of minimal effect on plant conditions or operations.

For the study of general corrosion in operating plant, linear polarisation resistance and electrochemical impedance have been favoured, whereas for the study of pitting or crevice corrosion electrochemical noise methods have been chosen in preference to polarisation which now-a-days is a method only used in the laboratory. However, harmonic analysis has come increasingly to the fore as a practical tool with the advent of instruments capable of performing precise, sensitive measurement and data manipulation, and of PCs able to run sophisticated computer software so as to derive values of derivatives and functions and then to printout or display pertinent information graphically, such as trends or sudden sporadic excursions. The theory underlying harmonic analysis was developed in the late 70s, early 80s but practical field use is more recent.

Harmonic analysis is an extension of electrochemical impedance determination. It involves applying a small ac sinusoidal voltage of selected frequency to a corroding electrode and observing the current response, which here comprises not only the fundamental (i.e. the impedance measurement) but also upper harmonics. Measurement of the fundamental and first and second harmonic components of the Faradaic impedance response is possible with an analyser with harmonic selection. The practical challenge is to apply a perturbation signal of sufficient amplitude to produce a measurable second harmonic without excessively disturbing the system, as the second harmonic may be found to lie within the noise level of the instrumentation. The resulting measurements (which relate the corrosion current i_Co_rr and the Tafel Slopes with the Stern-Geary Constant) provide the information needed to derive values for i_corr/ the charge transfer resistance Ret, and the Stern-Geary constant. (Note: three equations and three unknowns) . By way of example, the applied perturbation might be a voltage amplitude of around 25 mV at a frequency of around 0.1 Hz.

For rigorous academic studies of corrosion phenomena, the corrosion scientist would prefer in many cases to use a three-electrode cell arrangement.

Contrary to current practice and belief, we have established the practical adequacy of a two electrode arrangement for measurements of impedance, linear polarisation resistance, current noise and potential noise. The information derived is meaningful in terms of corrosion measurement and monitoring and provides bases for an informed plant corrosion management policy.

It should be noted here that while early awareness of an increase in the general corrosion rate occurring in a process- or production plant or of the incidence of pitting and crevice corrosion is usually desirable, it is not ordinarily needed. Provided measurements and monitoring outputs can be related retrospectively to an operating time-frame or period and the required analysis then made in a reasonable time in relation to the accumulating net effect of corrosion activity occurring, that usually will suffice. So-called "real time corrosion monitoring" is the monitoring of very recent events, allowing quick response to sudden excursions of corrosion behaviour, but it is not always a practical need. As used herein, the term "corrosion monitoring" has its ordinary meaning of relating changing or stable corrosion behaviour to a known time-frame, either by showing when particular conditions existed or changed or by showing time- windows within which corrosion events occurred.

The present invention stems from a finding that it is in fact possible to monitor both general corrosion rates and the occurrence of pitting or crevice corrosion in an operating plant by using a single introduced sensor electrode and the plant itself (including within the term "plant" the probe structure that may in chosen configurations carry the sensor electrode) as the second electrode of the required electrochemical cell.

For harmonic analysis measurements it is usually considered necessary to use a three-electrode arrangement for the measurement. Thus, for example, a test electrode of the same or electrochemically similar material to that of the plant would form one electrode (the working electrode) , a second electrode (an auxiliary electrode) which may be the same or a different material to the working electrode used to support the current flow, and a third electrode (a reference electrode) against which the polarisation of the working electrode is controlled. The plant liquid would provide the electrolyte. The use of two electrodes of the same/similar metals would not be considered because a symmetrical response would be expected with loss of the second harmonic (rectified) component.

However, we have now shown that if a single sensor electrode of substantially the same (electrochemically equivalent) material as the metal of which the plant is constructed (at the site of corrosion interest) is introduced so that it is in contact with the liquid medium to which relevant plant surface is exposed, an electrochemical corrosion monitoring cell can be established comprising the introduced sensor electrode, the local plant structure, as a second electrode, and the liquid medium as the electrolyte. In such an arrangement the surface area of the sensor electrode is (and is typically required to be) at least an order of magnitude smaller than the effective or apparent local surface area of the plant electrode. (It is of no material consequence that the plant structure is earthed) . The effect of this is that electrochemical responses from the corrosion- monitoring cell are, for practical purposes, due to the smaller of the two electrodes i.e. almost solely responses of the sensor electrode. These, by design, essentially correspond to what must be happening at the plant structure surface. Thus, by correspondence, the corrosion behaviour of the plant structure at the relevant location is monitored by, in fact, monitoring the behaviour of the equivalent metal sensor electrode in the same liquid environment.

Accordingly the present invention, in its broadest aspect, provides a method of establishing the probable, or equivalent, corrosion behaviour of plant metal surface in contact with corrosive liquid (e.g. process/ production liquid media) , the method comprising the following procedures:

(1) establish a corrosion monitoring electrochemical cell comprising the plant metal structure at the reference location as one electrode, a sensor electrode of known effective electrode surface area that is of the same metal as (or is electrochemically equivalent to) the plant metal whose effective surface area is small relative to the effective (or apparent) electrode surface area of the plant, as a second electrode, and the corrosive liquid as electrolyte in contact with both electrodes; and,

(2) monitor, by processing data from the cell in response to imposed stimuli (as by methods known per se) both the rate of general corrosion and the occurrence, if any, of pitting or crevice corrosion of the sensor electrode.

A preferred method of monitoring comprises at least one of the following procedures: -

(1) periodically (say, at half-hourly or hourly intervals so that the amount of data to be processed is reasonably manageable) monitor harmonic distortions using an oscillating potential applied between the sensor electrode and the plant metal surface that has a low frequency (say 0.01 to 1 Hz) sine wave form of amplitude in the range of 10 to 30 mV rms;

(2) determine by harmonic analysis the prevailing corrosion current values and Stern-Geary constant values;

(3) stop harmonic distortion measurement and apply a small polarising voltage between the sensor electrode and the plant metal surface that has a pre-set magnitude which equals, or has value which is a pre- selected fraction of the prevailing Stern-Geary constant value determined under (2);

(4) measure current flow as a function of elapsed time; and

(5) analyse, by statistical methods, filtering, or frequency domain transformation, the current signal obtained under (4) to establish the variability or stability of the signal and assess the rate and active mechanism of corrosion activity occurring at the sensor electrode.

Preferably the polarising voltage has a magnitude that is less than, but a substantial fraction (equal to or greater than 10%) of the prevailing Stern-Geary constant value.

The analysis under (5) will reveal general corrosion rates on a constant basis, by equating the current measured to the value that would be obtained if the system were to be polarised to a potential value equivalent to the Stern-Geary constant .

Optionally, the current noise measured may be correlated with the general corrosion rate. It can reveal localised corrosion (for example pitting or crevice attack) by correlation of the background general corrosion rate (per harmonic analysis) with transient excursions in the current signal.

For many or most situations faced in practice, measurement of current noise will provide sufficient information for plant corrosion management but, if desired, potential noise may be studied in the method of the invention.

The value of the current flow measured under (4) is equal to, or a fraction of, the corrosion current in the same proportion as the applied voltage bears to the Stern - Geary constant.

The sensor electrode should preferably have an effective surface area that is, at least, an order of magnitude (e.g. at least 1/lOth) smaller than the apparent, or 'effective', surface area of the plant electrode.

Two possible sensor electrode configurations will now be described with reference to the accompanying drawings in which;

Figure 1 shows schematically an in-line arrangement of a sensor electrode for a stretch of pipework, the electrode being part of a gasket at a flange coupling in the pipe-line, and Figure 2 shows schematically a cut-away view of a sensor electrode mounting at a boundary wall of a vessel and in the form of a lateral probe.

Referring to Figure 1, two end-to-end pipe sections la and lb are shown joined by flanges 2a and 2b that are clamped by bolts 3 and separated by a gasket 4. The gasket 4 comprises an annular sensor electrode disc (or ring) 4 electrically insulated from the flanges 2a and 2b by spacing insulating material 5. As represented for ease of understanding, the insulating material is shown as two separate plates on opposite sides of a plate electrode 4. In practice the electrode 4 might equally be a ring, or part ring, embedded in the radially inward surface of a cylindrical block of insulating material 5 such that the innermost surface of the electrode 4 is exposed to fluid in the pipeline. Electrical connecting leads are shown at 6, one coupled to the electrode 4 and the other to the flange 2a.

Referring to Figure 2, a localised part of a vessel wall is indicated by arrow 7. It is in contact with fluid indicated by arrow 8. An intrusive probe assembly shown generally by arrow 9 is affixed to the vessel wall 7 at a flanged port 10 in the vessel wall 7 such that a sensor electrode 11 is exposed at its inner end surface 11a to the fluid 8 at the vessel wall boundary. The electrode 11 is coupled to an end plate 12 that is clamped by bolts 13 to the flanged port 10. The electrode 11 is insulated from the vessel wall 7 and from the flanged port 10 by a layer of suitable insulating material (not numbered) . Electrical connections are shown at 14.

In both the illustrated arrangements the exposed surface areas of electrodes 4 and 11 are precisely known and are very small relative to the apparent, or effective, areas of the reference 'plant' electrodes. The materials from which the electrodes are made are the same as, or are electrochemically correlated to, the metals of the pipework or vessel, respectively. The measured responses of the electrochemical cells which these arrangements form are essentially (or may be taken to be) the responses of the sensor electrodes to conditions encountered and these by design mirror those of the plant structures at the monitored locations.

If the continuity of the relevant local plant surface is such, or perceived to be such, that the required ratio of electrodes surface areas may not be achieved, or if it is practically convenient to do so notwithstanding, reference 'plant' electrode surface may be introduced at sensor electrode installation either as a specific item or as a feature of sensor probe design.

In a further example of the method the electrochemical current noise of the sensor versus the plant metal surface was measured for a period of not less than 300 seconds, with a sampling period of 1 second. We then calculated statistical moments of the current signal, i.e. mean, variance, third and fourth moments. The skew and kurtosis of the current signal were then calculated from the moments. We also calculated a value for the localisation index of the signal (I mean divided by I r.m.s.) . This step was repeated later in the procedure .

We then stopped current noise measurement and monitored harmonic distortion of the current using an applied potential sine wave of amplitude 50 millivolts peak to peak, and of frequency 0.01 Hz for a period of 100 seconds. We then calculated the corrosion current and Stern Geary constant (B) from the harmonics.

After stopping harmonic measurement we also monitored solution resistance/conductivity by applying a potential square wave amplitude 10 mV r.m.s. and frequency of 2 kHz, and measuring the applied voltage and the current response, for a total period of 100 seconds.

Finally, we stopped solution resistance/conductivity measurement, and repeated the step mentioned above.

Using this method a continuous estimation of Electrochemical noise corrosion rate, harmonic analysis corrosion rate, Stern Geary value B, and solution resistance was obtained.

Since the measurements were all made on the same electrode, direct comparison of the electrochemical noise and harmonic analysis corrosion currents was possible. The harmonic analysis revealed details of the general corrosion behaviour and the current noise revealed evidence of localised corrosion of the metal. In yet a further example, we monitored harmonic distortion of the current using an applied potential sine wave with an amplitude of 50 millivolts peak to peak, and a frequency of 0.01 Hz for a period of 100 seconds. We calculated corrosion current and Stern- Geary constant (B) from the harmonics. This step was repeated later in the procedure.

We then applied a dc potential signal of a value 50% of the Stern-Geary value, measured electrochemical current noise of the sensor versus the plant metal surface for a period of not less than 300 seconds, with a sampling period of 1 second, and calculated statistical moments of the current signal, i.e. mean, variance, third and fourth moments. We then calculated the skew and kurtosis of the current signal from the moments, and calculated a value for the localisation index of the signal (I mean divided by I r.m.s.) . The mean current flow was of a value of ca 50% of the general corrosion current, and the higher moments indicated the stability of the corrosion current which relates to the probability of localised corrosion. We then repeated the first step referred to above.

This example describes an alternative method which involves polarisation of the sensor to a known fraction of the Stern-Geary constant B. References:

The following documents are incorporated herein by reference: Basic Corrosion 1. Henthorne, M. "Corrosion - Causes and Control" 1971-1972; available from the Carpenter Technology Corporation, PO Box 6621 Reading, PA 19603. 2. Wranglen, G. "Intro, to the Corrosion and Protection of Metals" 1971. 3. Fontana, M.G., and Green, N.D.I "Corrosion Engineering" 1967. 4. West, J.M. , "Basic Corrosion and Oxidation" 1981. 5. Ailor, W.M., (Ed.) Handbook on Corrosion Testing and valuation (1971) . Electrochemical Textbooks 6. Bockris, J.O.M., and Reddy, A.K.N., Modern Electrochemistry Vol. 2 1970. 7. Bard, A.J. and Faulkner, L.R. Electrochemical Techniques, 1980. 8. Southampton Electrochemistry Group, Electrochemical Methods 1985. Corrosion Papers: Linear Polarization 9. Stern M. , and Geary A.L., J. Electrochem. Soc . 104, 56 (1957) . 10. Callow L.M., Richardson J.A. and Dawson J.L., B.Corr.J., II, 123 (1976) . 11. Grauer R., Moreland P.J. and Pini G. A Literature Review of Polarization Resistance Constants (B) NACE, Houston (1982) . Electrochemical Impedance 12. Sluyters J.H., and Sluyters - Rehback M. , Electroanal Chem. Vol. 4, Chpt. 1, pp.1-128. 1969. 13. Schuhman D., J. Chim. Phys . 60, 359 (1963) . 14. Dawson J.L., Gill J.S., Al-Zanki, I. A. and Woollam R.C., Dechema-Monograph Vol. 101, p.235 (1986) . 15. Dawson J.L., Callow L.M., Hladky K. and Richardson J.A. ; On-line Surveillance and Monitoring of Process Plant pp. 33.1-33.9. 1977. 16 Hladky K. , Callow L.M. , and Dawson J.L., Br. Corr. J. 15, 20, (1980) . 17. Harumaya S., and Tsuru T., ASRM STP 727, pp.176- 186 (1981) . 18. Epelboin I., Gabrielli C, Keddam M. , and Takenouti H. , ASTM STP 727, pp. 150-160, (1981) . 19. Walter G.W., Corr. Sci . 26 681 (1986). 20. Gabrielli C, "Identification of Electrochemical Processes by Frequency Response Analysis", Solartron Monograph 1980. 21. Armstrong R.D. and Henderson M. , J. Electroanal, Chem. 48, 150 (1973) . 22. Rowlands J.C. and Chuter D.J., Corr. Sci 23 331 (1983) . Harmonic Analysis 23. Prabharahara Rao, G. and Mishra, A.K., J. Electroanal. Chem., Vol 77, p323 1977. 24. Rangarajan, S.K., J Electroanal. Chem., Vol. 62, p.31 1975. 25. Devay, J. , and Meszaros, L., Acta Chim. Acad.Sci., Hungary, Vol 100, p.183 (1979) . 26. Devay, J., and Meszaros, L., Acta Chim Acad. Sci., Hungary, Vol 104, No. 13, p.311 (1980). 27. Meszaros, L. , Korrosion Figyelo, Vol 21, No. 2, p.30 (1981) . 28. Gill, J.S., Callow, L.M., and Scantlebury, J.D., Corrosion 39, 61 (1983) . 29. Xu Naixin, Dawson, J.L., Thompson, G.E., and Wood, G.C., J. Chinese Soc . , Corr Prot . 159, 167, 4, (3), (1984). 30. Gabrielli, G., Keddem, M. and Takenouti, H. , Symp. on Electrochemical Methods in Corrosion Research, Toulouse, France (1985) . Electrochemical Noise 31. Hagyard, T. and Williams, J.R., Trans. Faraday Soc. , 57, 2288 (1961) . 32. Iverson, W.P., J. Electrochem. Soc, 115, 617 (1968) . 33. Barker, G.C., J. Electroanal. Chem., 21, 127 (1969); 39, 484 (1972), 82, 145 (1977) . 34. Tyagai, V.A. , "Electrocal Phenomena at the Biological Membrane Level" pp. 325-335. Ed. E. Roux, Pub. Elsevier (1977) . 35. Hooge, F.N., etal Physics 42, 331, (1969); 45, 386, (1969); 60, 130, (1972) . 36. Okamoto, G. , Tachibana, K. , Nishiyama, S., Sugita, T., "Passivity and its Breakdown on Iron and Iron Base Alloys" p.106, Pub. NACE, Houston (1976) . 37. Tachibana, K. , and Okamoto, G., Reviews on Coatings and Corrosion IV, No. 3, 229, (1981) . 38. Blanc, G., Gabrielli, C, and Keddam, M. , Electrochim Acta 20, 687, (1975) . 39. Blanc, G. , Epelboin, I., Gabrielli, C, and Keddam, M. , J Eleotroanal . Chem. 62, 59, (1975),; 75, 97, (1977) . 40. Blanc, G., Gabrielli, C, Ksouri, M. , and Wiart, R., Electrochimica Acta, 23, 337 (1978) . 41. Epelboin, I., Gabrielli, C, Keddam, M. and Raillon, L., J. Electroanal. Chem., 105, 389 (1979). 42. Bertocci, U.J., Electrochem. Soc, 127, 1931 (1980) . 43. Bertocci, U. , and Kruger, J. Surface Science, 101, 608 (1980) . 44. Bertocci, U.J., Electrochem. Soc, 128, 520 (1981) . 45. Hladky, K. , and Dawson, J.L., Corr. Sci. 22, 231, (1982) . 46. Hladky, K. , European Pat. 084 404 A3, USA Pat. 455709, Canadian 418938. 47. Eden, D.A., Dawson, J.L., and John, D.G., UK Pat. Appl. 8611518, May 1986. 48. Williams, D.E., Westcott, C, and Fleischmann, M., J. Electrochem Soc. 132 1976 and 1804 (1985). 49. Al-Zanki, A.I., PhD Thesis, UMIST, (1987). 50. Dawson, J.L., Cox, W.M., Eden, D.A., Hladky, K. , and John, D.G., G. Prove non Destructive (Italy) 2, 49, (1986) . 51. Al-Zanki, I.A., Gill, J.S., and Dawson, J.L., Materials Sci. Forum., 8, 463 (1986) . 52. Shih, C.C., PhD Thesis, UMIST, (1989). 53. Bertocci, U. , 7th Int. Congress in Metallic Corrosion, Ass. Brasi ler de Corr., Rio de Janeiro, (1979) . 54. Brown, S., Msc Project, UMIST, (1986) . 55. Dawson, J.L., Eden, D.A., and Brown, S., Paper presented at Symposium on Expt. Tech. in Corr. Sci., Oxford, January 1987. 56. Dawson, J.L., and Ferreira M.G.S., Corr. Sci., 26, 1009 (1986) . 57. Williams, D.E., Electrochemical Corrosion Testing, Dechema Monoqraph 101, 253 (1986) .

Claims

CLAIMS : 1. A method of establishing the probable or equivalent corrosion behaviour of plant metal surface in contact with corrosive liquid wherein a corrosion monitoring cell is established of which the plant metal structure at a reference location where said surface is forms a first electrode, and a sensor electrode of the same or an electrochemically equivalent metal as the plant metal structure forms a second electrode of the cell and has an exposed surface in contact with the corrosive liquid at the reference location, characterised in that: the second electrode has a pre-determined effective electrode surface area which is small relative to the effective, or apparent, surface area of the first electrode, and data from the cell obtained in response to imposed stimuli are monitored to provide an indication of the occurrence, if any, of either pitting or crevice corrosion of the sensor electrode, said indication also being an indication of the local corrosion behaviour of the plant metal structure.

2. A method as claimed in claim 1, wherein the rate of general corrosion is also measured.

3. A method as claimed in claim 1 or claim 2, including the further step of periodically measuring harmonic distortions.

4. A method as claimed in claim 3, wherein the harmonic distortions are measured using an oscillating potential with a low frequency sine-wave form applied between the electrodes.

5. A method as claimed in claim 3 or claim 4, wherein the harmonic distortions are of 0.01 to 1 Hz and 10 to 30 mV rms amplitude.

6. A method as claimed in any one of claims 3-5, including the further step of determining by harmonic analysis values of the prevailing corrosion current and the Stern-Geary constant .

7. A method as claimed in claim 6, including the further step of stopping harmonic distortion measurement and applying a polarising voltage between the electrodes that has a pre-set magnitude equalling or being a pre-selected fraction of the determined Stern-Geary constant value.

8. A method as claimed in any one of claims 3-7, including the further step of measuring current flow as a function of elapsed time.

9. A method as claimed in claim 8 including the further step of analysing by statistical methods, filtering, or frequency domain transformation the obtained current signal to establish the variability or stability of the signal and assess the rate and active mechanism of corrosion activity occurring at the sensor electrode.

10. A method as claimed in any one of claims 3-9, in which the polarising voltage has a value that is less than the Stern-Geary constant value.

11. A method as claimed in any one of claims 3-10, wherein the polarising voltage has a value of at least 10% of the Stern-Geary constant value.

12. A method as claimed in any preceding claim wherein potential noise is additionally monitored.

13. A method as claimed in any preceding claim wherein said ratio of electrodes surface areas is not greater than 1:10.

14. A method as claimed in any preceding claim, wherein the sensor electrode is part of a gasket assemblage at a flanged coupling of the plant metal structure.

15. A method as claimed in any one of claims 1 to 13, wherein the sensor electrode is incorporated in an intrusive lateral probe attached to the plant metal structure at a port in a vessel wall.

16. A method as claimed in any preceding claim, wherein the first electrode comprises surface regions of the probe that are insulated from the sensor electrode, also exposed to the corrosive liquid at said location, and are similar or equivalent in electrochemical response to adjacent plant metal structure.