Go To On-Line Conferencing Forum
M.A.A. Tullmin1, C.M. Hansson2 and
P.R. Roberge3
1Materials
and Metallurgy Engineering Department, Queen's University, Kingston,
ON Canada, K7L 3N6
2University
Research, University of Waterloo, Waterloo, ON Canada, N2L 3G1
3Department
of Chemistry and Chem. Eng., Royal Military College, Kingston,
ON Canada, K7K 5L0
email: mt7@knot.queensu.ca
Abstract
The problem of the concrete infrastructure degradation due to rebar corrosion processes is particularly serious with major economic implications. The ability to assess the severity of rebar corrosion in existing structures and to predict their remaining service life is thus becoming increasingly important. A number of electrochemical corrosion measurement techniques are available for these purposes. They can be broadly divided into two categories, the first of these being measurements on the actual structural rebar and the second involving measurements on smaller, strategically positioned, embedded rebar probes. Since all these techniques have certain advantages and limitations, an approach of using a combination of these is advocated. Significant benefits can be derived from the application of electrochemical rebar corrosion measurement technology, provided certain pitfalls are avoided.
Introduction
Increasingly it is becoming apparent that the steel reinforced concrete infrastructure of North America and many other regions is suffering large scale degradation and the economic implications of this problem are alarming. The principal cause of the degradation of structures such as bridges, parking garages, transit system tunnels, piers and residential buildings is corrosion damage to the reinforcing steel (rebar) which is embedded in the concrete. In turn, this corrosion damage is largely related to the use of de-icing salts and chlorides found in marine environments, as well as carbonation of the concrete from carbon dioxide in the atmosphere.
Inattention to corrosion control as part of an overall maintenance program for infrastructure facilities has been reported to cost the US more than $ 250 billion annually (1). The US Department of Transportation recently projected the rehabilitation costs of existing bridges at $ 155 billion (2). In Canada, with the large scale use of de-icing salts dictated by the cold climate, the situation is correspondingly serious. Canada's concrete infrastructure, of which a significant portion is near the end of its design life, has a replacement value of over half a trillion dollars (3).
The ability to assess the severity of corrosion in existing concrete structures for maintenance and inspection scheduling and the use of corrosion data for predicting the remaining service life is becoming increasingly important. In this paper, a number of electrochemical techniques for measuring rebar corrosion are described. There are many pitfalls in applying these techniques to the steel in concrete system, due to "complications" not usually found in other measuring environments.
Electrochemical Rebar Corrosion Measurements
These measurements can be performed completely non destructively on the actual reinforcing steel or on separate probes that are embedded in the concrete structure, preferably at different depths of cover. There are a number of electrochemical techniques for measuring the severity of rebar corrosion, each with certain advantages and limitations. This data can provide early warning of structural distress and for evaluating the effectiveness of corrosion control strategies that may have been implemented. Once rebar corrosion has proceeded to an advanced state, where its effects are apparent on the outside concrete surface, it is too late to implement effective remedial measures such as cathodic protection or chloride extraction treatments.
A classification of the severity of rebar corrosion rates is presented in Table 1, and a proposed guideline (4) for relating the measured corrosion rates and remaining service life is shown in Table 2. However, caution needs to be exercised in using data of this nature, since constant corrosion rates with time are assumed. Rebar corrosion rate data measured on a particular structure can also be correlated to other parameters such as chloride concentration profiles to determine critical chloride levels, which are useful for life prediction of a particular structure.
Generally, the value of isolated "once-off" rebar corrosion rates is limited, because rebar corrosion rates can change considerably with time. Such variations are the result of fluctuations in temperature, humidity, degree of aeration, microstructural changes in the concrete, the development of cracks in structures etc.
Corrosion Measurements on the Actual Rebar
Rebar potential measurements
Perhaps the simplest assessment technique for rebar corrosion damage is the measurement of the corrosion potential (or rest potential, half cell potential). A measurement procedure is described in the ASTM C876-80 standard (5), the basis of which is that the corrosion potential of the rebar will shift in the negative direction if the surface changes from the passive to the active state. The simplified interpretation of the potential readings is presented in Table 3.
The advantages of this method are its simplicity and the possibility of "mechanized" measurements to cover large concrete surface areas in short time frames. Such devices are described in more detail in References 6 and 7. The disadvantages are that the results are only qualitative, without the establishment of an actual rebar corrosion rate. At times, very negative rebar potential readings can be misleading and may be recorded in submerged concrete, where the corrosion rate remains negligibly low due to unavailability of oxygen stifling the cathodic reaction rate.
A more recent derivative of this methodology is the measurement
of potential gradients using two reference electrodes, eliminating
the need to make direct electrical contact with the rebars (8).
Guard Ring Devices
The main challenge of applying well known corrosion rate measurement techniques such as the Linear Polarization Resistance (LPR) technique to actual rebars embedded in concrete has been confinement of the applied potential (or current) perturbation to a well defined rebar area, in order to minimize measurement errors. The development of guard ring devices has addressed this issue.
The guard ring is maintained at the same potential as the counter electrode to prevent the current from the counter electrode to flow beyond the confinement of the guard ring. The counter, reference and guard ring electrodes can be conveniently located in a sensor placed directly above the rebar of interest (Figure 1), necessitating only one electronic lead attachment to the rebar for corrosion rate measurements, as in the case of the simple potential measurements.
Guard ring devices will display a certain corrosion rate reading expressed as thickness loss per time unit (say mm/year) after the polarization cycle is completed but there are many simplifying assumptions in the derivation of this corrosion rate. Important limitations include the assumption of uniform corrosion over the rebar surface (this rarely applies to chloride induced rebar corrosion), simplified models for the electrochemical reactions and charge transfer processes, assumed values for the Tafel constants as well as possible inaccuracies in "IR" drop resistance corrections. A further fundamental source of inaccuracies is that no allowance is made for the effects of macrocell corrosion which are inherent to actual rebar grids. In addition, the applicability of these type of measurements to cracked concrete is presently not clear. The importance of the influence of structural cracks on rebar corrosion damage is discussed in Reference 9. In the authors' experience, the above limitations are all too often ignored and the corrosion rate values are mistakenly quoted to too many significant figures.
For the steel in concrete system, it is imperative that sufficient time is allowed for a current value to stabilize at a certain potential (or vice versa) (10). For example, in the potentiostatic LPR technique, it will typically take several minutes for the current to reach a stable level, after the polarizing voltage is applied. Shorter polarization could lead to significant measurement errors.
A commercial LPR based guard ring type corrosion measuring system has been evaluated on laboratory concrete slabs with and without chloride contamination of the concrete. Higher corrosion rates were consistently measured in the chloride bearing concrete. The guard ring corrosion rate results were similar to those determined by separate conventional LPR corrosion rate measurements obtained from smaller embedded rebar probe elements.
The guard ring system was subsequently used for corrosion measurements on a number of bridge structures. On the Kennedy Bridge of the 401 Highway in Toronto, the lower corrosion rates were generally measured at points where the rebar corrosion potential was found to be more "passive", as shown in Figure 2. However, significant scatter in the data is apparent. In an application on the Oak Street Bridge in Vancouver, the guard ring device indicated very little corrosion activity, except for isolated readings of higher corrosion rates. In this bridge deck it was possible to inspect the actual rebars visually, subsequent to the corrosion measurements, after the concrete cover was removed by hydra milling. Good agreement was found between the actual condition of the rebar and the guard ring corrosion data, in that corrosive attack was limited to isolated "hot spots" on the surface of the rebars. The results of other comparative corrosion tests with guard ring type devices may be found in References 11 and 12.
Unfortunately, the guard ring technology does not lend itself for assessing large concrete surface areas in short time periods. To reduce evaluation times to acceptable, practical levels it may be advisable to map the corrosion potential values, followed by selective application of the guard ring device to critical areas.
Galvanostatic Pulse Technique
In this technique a short term anodic current pulse is imposed onto the rebar, from a counter electrode positioned on the surface of the concrete. The resultant rebar potential change (E) is recorded by means of a reference electrode, also located on the concrete surface. Typical current pulse duration (t) and amplitude have been reported at 3s and 0.1 mA respectively (7).
The slope of the potential vs. time curve (E/t) , measured during the current pulse, has been used to provide information on the rebar corrosion state. Passive rebar reportedly has a relatively high slope, while rebar undergoing localized corrosion has a very small slope. In the latter case, the rebar potential only shifts by a few millivolts under the applied current pulse (7). The potential data has also been used to obtain a measure of the concrete resistivity (for a given depth of cover) (7). The technique is reportedly very rapid and may facilitate more unambiguous information on the rebar corrosion state than is possible by simple potential mapping (7).
Corrosion Measurements on Embedded Probes
Potentiodynamic Polarization Curves
These well known dc type measurements can be performed on rebar steel in concrete using a three electrode system. Typically the working electrode is a small section of rebar, while stainless steel or graphite may be used for counter electrodes. Embedded reference electrodes need to be resistant to the highly alkaline concrete pore solution, such a manganese dioxide electrodes. External reference electrodes such as the Cu/CuSO4 variety can also be used but due cognizance of possible "IR" drop errors is required.
The rebar potential and current density data can provide useful
information such as:
A fundamental drawback of these measurements is that "artificial" corrosion damage and rebar surface changes can be induced at relatively high polarization levels. For steel in concrete, the potential scan rates used should be extremely low, making the tests time consuming. Specialist personnel is required for data interpretation.
Electrochemical Impedance Spectroscopy (EIS) and Harmonic Analysis
One method of obtaining corrosion rates from EIS data involves the use of equivalent circuit models. A number of different types of equivalent circuit models have been proposed for reinforcing steel in concrete (13). By accounting for the concrete "solution" resistance and the use of more sophisticated modeling, a more accurate corrosion rate value should "theoretically" be obtained compared with the more simplistic dc LPR analysis. However, EIS data generation and analysis generally requires specialist electrochemical knowledge and can be rather lengthy, making it unsuitable for rapid evaluation of corrosion rates. To derive the corrosion rates, the Tafel constants also still have to be estimated or assumed.
A further more recently evolving technique closely related to EIS measurements is harmonic analysis. An advantageous feature of harmonic analysis is that mathematical data treatment facilitates direct computation of the Tafel constants and the corrosion rate. Since harmonic analysis is performed in a narrow frequency range, it can provide for practical and rapid rebar corrosion rate determination (13).
A severe restriction of EIS and harmonic analysis is that, as in the LPR technique, the fundamental assumption of uniform rebar corrosion has to be made in the calculation of penetration rates. If localized corrosion damage is actually taking place, the data is at best of a qualitative nature, indicating the breakdown of passivity and the possibility of localized attack.
A number of initiatives have been launched aimed at applying EIS measurements directly to structural rebars, rather than small scale embedded probes. These developments have involved guard ring concepts (14) and modeling of signal transmission along the length of the rebar (14,15).
Zero Resistance Ammetry
The macro cell current measured between embedded rebar probes can serve as an indication of the severity of corrosion. This principle is used in the ASTM G102-92 corrosion test procedure (16), where the current flows between rebar embedded near the surface and rebar at greater depths of cover. A similar approach has been adopted in a rebar corrosion monitoring system (17), with currents being measured between strategically placed carbon steel rebar probe elements.
Interestingly, the concept of measuring a macro cell current as indicator of corrosion severity can also be applied to probe elements of identical materials and exposed to the same environment. It may be somewhat surprising that a significant current will flow between nominally identical probe elements but this principle has been used in commercial corrosion monitoring and surveillance systems for many years. It can be argued that such measurements are mainly relevant to detecting the breakdown of passivity and the early stages of corrosion damage. If extensive corrosion damage is occurring on both the probe elements, the macro cell current measured will not accurately reflect the severity of attack, as indeed reported by Berke et al.(18). The increase in macro cell current flowing between identical rebar probe elements exposed to a simulated concrete pore solution, after the addition of 2% chlorides, is shown in Figure 3. Current measurements between rebar elements embedded at the same level have also been used successfully as an indicator of rebar corrosion activity (19). In some cases a crack was induced in the concrete over one of the probe elements and the other element in an uncracked area, in order to determine the degree of preferential corrosion of the "cracked" probe.
Electrochemical Noise
Unlike other electrochemical techniques, noise measurements do not rely on any "artificial" signal imposed on the rebar probe elements. Rather, natural fluctuations in the corrosion potential and current are measured to characterize the severity and type of corrosive attack. For these measurements, three nominally identical rebar probe elements can be conveniently embedded in the concrete.
There may be some reluctance to using electrochemical noise for rebar corrosion measurements in the field, due to a perceived "over sensitivity" of the equipment and fears of external signal interference. While such concerns may be justified in certain cases, and this technology is relatively new to the rebar field, it has recently been used successfully in rebar corrosion measurements in the Vancouver harbor and in clarifier tanks of the Paper and pulp industry in British Columbia. In these applications, the rebar noise probes were embedded in large (up to 4 meters long) concrete prisms. These prisms were partially submerged and exposed to sea water at the United Grain Growers Terminal (Vancouver harbor) and to the effluent solution in the clarifiers. The probe arrangement and nature of the specimens in the former application is described in Reference 19.
Electrochemical noise data showing increased corrosion activity, associated with the tidal cycle in the Vancouver harbor is described in Reference 19. In this case, the highest corrosion activity on the probe element, located in cracked concrete, occurred as the tide level approached that of the crack. Under these conditions sea water is available to penetrate the concrete as corrosive electrolyte, together with a high degree of aeration, a combination expected to stimulate electrochemical corrosion processes. Current noise data from the exposure to the paper and pulp effluent is shown in Figure 4, for rebar probes embedded in cracked and uncracked concrete respectively. This noise data confirmed that, as expected, the presence of cracks facilitated more rapid diffusion of corrosive species to the rebar surface and led to higher electrochemical corrosion activity. To date, no significant interference or other problems have been encountered in the noise measurements of these two field exposure programs.
A distinct advantage of the electrochemical noise techniques is that the initiation and propagation of corrosion pits can be clearly identified. Distinct pit initiation transients, comprised of a sharp signal increase with passive film breakdown and a more gradual recovery of the signal to base line levels as the passive film repairs itself are shown in Figure 5, for a carbon steel rebar probe exposed to chloride containing concrete pore solution. These distinct signatures of the initiation of corrosion pits were evident long before the attack was observable by visual means, indicating the "early warning" capabilities of this technology.
Apart from the interpretation of the "raw" noise data, it has become customary to conduct further statistical data processing, spectral analysis and chaos theory analysis (20). Such data treatment serves to reduce the volume of data and to assist in distinguishing different forms of corrosion from one another.
Benefits
The benefits derived from applying the electrochemical rebar corrosion
measuring technology include the following:
If these benefits can be documented in present and future corrosion
measurement programs, steady increases in the application of electrochemical
rebar corrosion measuring technology are anticipated, given the
background of the large scale concrete infrastructure degradation.
Summary and Conclusions
A number of electrochemical rebar corrosion measurement techniques is available presently, each with certain advantages and limitations. The complexity and required specialist know-how in applying these techniques also varies significantly. To obtain maximum information about the corrosion state of rebar in a particular structure, a combination of measuring techniques is recommended. Furthermore, decisions should not be based on isolated corrosion readings.
Many measuring instruments will display a certain rate of corrosion penetration but caution needs to be exercised in interpreting these values, since they are often based on many simplifying (in some cases unrealistic) assumptions. Although the electrochemical corrosion measurements are usually qualitative, or at best semi quantitative, significant benefits can be derived from them.
Acknowledgements
The first two authors gratefully acknowledge financial support from Concrete Canada and the Ontario Centre for Materials Research. Some of the experimental data presented in this paper was obtained by collaborative work with Levelton Associates based in Richmond, British Columbia.
References
1. S.L. Nadel: in "publicly Speaking", Materials Performance, December 1994, p.59.
2. O. Chaix, W.H. Hartt, R. Kessler and R. Powers: "Localized Cathodic Protection of Simulated Prestressed Concrete Pilings in Seawater", Corrosion, Vol.51, No.5, May 1995, pp.386-398.
3. "High Performance Concrete: Improving Canada’s Infrastructure and International Competitiveness", Concrete Canada, Vol.1, October 1993.
4. K.C. Clear: "Measuring Rate of Corrosion of Steel in Field Concrete Structures", Transportation Research Record 1211, 1989, pp.28-37.
5. ASTM C876 Standard Test Method for Half Cell Potentials of Reinforcing Steel in Concrete, Annual Book of ASTM Standards, Vol. 04.02, 1983.
6. J.P. Broomfield, P.E. Langford and A.J. Ewins: "The Use of a Potential Wheel to Survey Reinforced Concrete Structures", Corrosion Rates of Steel in Concrete, ASTM STP 1065, N.S. Berke, V. Chaker and D. Whiting Eds., 1990, pp.157-173.
7. B. Elsener and H. Bohni: "Potential Mapping and Corrosion of Steel in Concrete", Corrosion Rates of Steel in Concrete, ASTM STP 1065, N.S. Berke, V. Chaker and D. Whiting Eds., 1990, pp.143-156.
8. A.K. Suryavanshi, S. Syam Sunder and B.U. Nayak: "A comparison of surface potentials of r.c. structures using reference electrodes", Corrosion Prevention and Control, August 1991, pp.105-107.
9. B. Borgard, C. Warren, S. Somayaji and R. Heidersbach: "Correlation Between Corrosion of Reinforcing Steel and Voids and Cracks in Concrete Structures", Transportation Research Record 1211, 1989, pp.1-11 (including discussion C.F. Crumpton)
10. K.R. Gowers, S.G. Millard, J.S. Gill and R.P. Gill: "Programmable linear polarisation meter for determination of corrosion rate of reinforcement in concrete structures, British Corrosion Journal,Vol.29, No.1, 1994, pp.25-32.
11. J. Flis, H.W. Pickering and K. Osseo-Asare: "Assessment of data from Three Electrochemical Instruments for Evaluation of Reinforcement Corrosion Rates in Concrete Bridge Components", Corrosion, Vol.51, No.8, August 1995, pp.602-609.
12. J.A. Gonzalez, M. Benito, S. Feliu, P.Rodriguez and C. Andrade: "Suitability of assessment Methods for Identifying Active and Passive Zones in Reinforced Concrete", Corrosion, Vol.51, No.2, February 1995, pp.145-152.
13. M.I. Jafar, J.L. Dawson, D.G. John: "Electrochemical Impedance and Harmonic Analysis Measurements on Steel in Concrete", Electrochemical Impedance: Analysis and Interpretation, ASTM STP 1188, J.R. Scully, D.C. Silverman and M.W. Kendig Eds, 1993, pp.384-403.
14. L. Lemoine, F. Wenger and J. Galland: "Study of the Corrosion of Concrete Reinforcement by Electrochemical Impedance Measurement", Corrosion rates of steel in Concrete, ASTM STP 1065, N.S. Berke, V.Chaker and D. Whiting Eds., 1990, pp. 118-133.
15. D.D. Macdonald, M. Urquidi-Macdonald, R.C. Rocha-Filho and Y. El-Tantawy: "Determination of the Polarization Resistance of Rebar in Reinforced Concrete", Corrosion, Vol.47, No.5, May 1991, pp.330-335.
16. ASTM G102-89 Standard Practice for Calculation of Corrosion Rates and Related Information from Electrochemical Measurements, Annual Book of ASTM Standards, Vol. 03.02, 1992.
17. P. Schiessl and M. Raupach: "Macrocell Steel Corrosion in Concrete Caused by Chlorides", Second CANMET/ACI International Conference on Durability of Concrete, Montreal Canada, 1991, pp.565-583.
18. N.S. Berke, D.F. Shen and K.M. Sundberg: "Comparison of the Polarization Resistance Technique to the Macrocell Corrosion Technique", Corrosion rates of Steel in Concrete, ASTM STP 1065, N.S. Berke, V. Chaker and D. Whiting Eds., 1990, pp.38-51.
19. R. Weiermair, C.M. Hansson, P.T. Seabrook and M. Tullmin "Corrosion Measurements on Steel Embedded in High Performance Concrete, Third CANMET/ACI Conference on Performance of Concrete in Marine Environment, Aug. 4-9, 1996, New Brunswick, Canada.
20. A. Legat and V. Dolecek: "Corrosion Monitoring System Based on Measurement and Analysis of Electrochemical Noise", Corrosion, Vol.51, No.4, April 1995, pp.295-300.
Table 1 Corrosion Rates of Steel in Concrete
| Rate of Corrosion | Corrosion Current Density (icorr) A/cm2 | Corrosion penetration
m/year |
| High | 10-100 | 100-1000 |
| Medium | 1-10 | 10-100 |
| Low | 0.1-1 | 1-10 |
| Passive | <0.1 | <1 |
After Reference 15
Table 2 Proposed Relationship between Corrosion rate and Remaining
Service Life
| icorr (A/cm2) | Severity of Damage |
| <0.5 | no corrosion damage expected |
| 0.5-2.7 | corrosion damage possible in 10 to 15 years |
| 2.7-27 | corrosion damage expected in 2 to 10 years |
| >27 | corrosion damage expected in 2 years or less |
After Reference 4
Table 3 Likelihood of Corrosion Damage as a Function of the Corrosion
Potential
| Corrosion Potential (Volts vs Cu/CuSO4) | Probablility of Corrosion |
| >-0.200 | <10 % |
| -0.200 to -0.350 | Uncertain |
| <-0.350 | >90 % |
After Reference 5
Your comments and/or suggestions are encouraged. After you fill out the form below, click on the "Post Message" button. Then, click on the reload button on your browser to see your comment and any other new ones.
On-Line Conferencing Forum