Comparison of Half-Cell Potential and Corrosion Rate Measurement Methods-Experience on Site Evaluation of Rebar Corrosion

1. Background

Since 1978 Half Cell Potential (HCP) mapping [1] has been used for detecting corroding areas on concrete structures in Denmark. In the beginning (after balcony gangway has totally collapsed) this method was mainly used on carbonated structures and balconies exposed to de- icing salts. Later the method was used for all kind of structures and the experiences were discussed in Newsletters published by the Danish Corrosion Centre [2]. It was early recognized that the interpretation of the HCP results were difficult or misleading in wet and semi- wet structures where lack of oxygen as well as corrosion would lead to potential gradients.

As early as 1978, Denmark began to use the half-cell potential map method to determine the eroded range of concrete structures. At first, this method was mainly used for the detection of carbonized concrete structures and balconies prone to corrosion by deicing salts. Later, this method was popularized for all structures, and the experience was discussed in the Danish Corrosion Center publication. For a concrete structure that lacks oxygen and is wet or semi-humid, corrosion will result in a potential difference. It is difficult to interpret the half-cell potential structure, and the results of the half-cell potential are difficult to interpret or mislead.

A typical potential map of a highway bridge pillar is shown in fig. 1. The pillar is exposed to de-icing salts splashed from the passing cars up to a level of 2 meters, but also has a high humidity at the ground level caused by capillary suction. The water filled pore system in the concrete makes the potential drop because the oxygen necessary to maintain the passive film will not be able to diffuse into the concrete fast enough.

Figure 1 shows a typical potential diagram of a highway bridge column. The 2-meter-high column is exposed to the environment that is susceptible to ice and salt splashes from passing vehicles, and due to the siphon effect, the humidity at the surface is very large, and the concrete pores are filled with water, making The passivation film necessary for maintenance cannot be quickly diffused and filled into the concrete structure, causing the potential of the concrete structure to drop.

Fig.1. Typical potential map of a high way concrete pillar exposed to de-icing

salts

The ongoing corrosion process is described by the chemical reaction:

2Fe + O2 + 2H2O↔2Fe +++ 4OH-

The Fe ++ forms at the high pH in concrete a complex protective film with oxygen. This film can be broken down by the chloride ions from de-icing salts or by neutralising of the high pH eg by carbonation.

The interpretation problem is that the potential will drop either because the loss of passive film due to lack of oxygen (the corrosion process will not be able to proceed without oxygen) or because chlorides break down the passive film and start corrosion.

The ongoing corrosion process can be expressed by the following chemical reaction formula:

2Fe + O2 + 2H2O↔2Fe +++ 4OH-

The iron with high PH value is oxidized to form a composite passivation film, which will be destroyed due to the neutralization of deionized salt chloride or carbonate ions.

It is difficult to explain that the damage to the passivation film caused by the lack of oxygen will also reduce the potential, and the chloride ions will also destroy the passivation film and cause the steel bars to begin to corrode.

In 1994 it was decided to develop an equipment for concrete structures based on well know techniques for determination of corrosion rates to be able to distinguish between active corrosion and the lack of oxygen situation.

In 1994, it was decided to develop a mature and advanced equipment for determining the corrosion rate of concrete structures, and to distinguish between the active corrosion state and the lack of oxygen.

The results presented in this paper are all based on Galvanostatic Pulse Measurements (GPM). This polarisation technique makes it possible within a short time (typically 10 seconds) to calculate the corrosion rate [3,4]. The equipment gives both the corrosion rate and the half-cell potential as well as the resistance between the hand-held electrode placed on the examined concrete surface and the reinforcement. Four different examples from on-site investigations are described below, one where there is a good correlation between the HCP method and the GPM method and 3 where the HCP measurements are misleading.

The results presented in this paper were all measured by Galvanostatic pulse method. Polarization technology enables calculation of corrosion rate in a short time of 10 seconds. In addition to measuring the corrosion rate, this instrument can also measure the half-cell potential, as well as the resistivity between the electrode and the steel bar placed on the measured concrete surface. Four different on-site test cases will be explained below. In one case, the results of the half-cell potential (HCP) and the pulse current test (GMP) are very relevant, and in the other three cases, the results of the half-cell potential measurement are easy. Misleading.

2. EXAMPLES OF ON-SITE INVESTIGATIONS

2. On-site test case

2.1 Example no. 1

2.1 Case One

Two parallel bridges were built in 1965-67 in the Copenhagen area, where the highway crosses over a railway line, a parking lot and two minor roads. The eastern bridge was rehabilitated extensively at a very high price in 1978, after which the western bridge have only received much less rehabilitation, but substantial inspection, test- loadings, probabilistic assessment etc., which essentially have kept the bridge in function at a much less cost that the eastern part.

From 1965 to 1967, two parallel bridges were built in the Copenhagen district. The highway line spans the railway line, with a parking lot and two branch roads. In 1978, the bridge in the east was expensively overhauled, and the bridge in the west was relatively strong after inspection, testing and evaluation, so only a small simple repair with a lower cost was adopted.

Initial inspections, core investigations and chloride profiling in 1999 (fig. 2) pointed out column No. S303 to be suitable for corrosion rate measurements.

In 1999, a coring and chloride ion content test (Figure 2) was performed, which showed that No. S303 cylinder was suitable for corrosion rate test.

Fig 2. Chloride profiles at level 0.3m and level 1 m

Figure 2. Chloride ion content at 0.3 meters and 1 meter

Electrical continuity in the reinforcement was checked and a permanent connection was welded to the reinforcement. The vertical reinforcement (Ø35 mm) is typically in 60 mm depth and the horizontal (Ø14 mm) in 40 mm depth. Already in 1999 the chloride content in level 0.3 m is so high that active corrosion can be expected. In September 2000 and in April

2001 corrosion rates were determined together with the half- cell potentials and resistance measurements [5] (fig 3).

One end of the device is fixedly connected to the steel bar, and a continuous pulse current is applied. The vertical rebar (Ø35 mm) is at a depth of 60 mm, and the horizontal rebar (Ø14 mm) is at a depth of 40 mm. In 1999, the chloride ion content at a depth of 0.3 meters was already very high, and there should be active corrosion. In September 2000 and April 2001, the corrosion rate was determined using half-cell potential and resistivity test methods. (image 3)

In this case there is a rather good correlation between resistance, half-cell potential and corrosion rate mapping.

In this case, there is a good correlation between the corrosion rate graph, resistivity, and half-cell potential.

To verify the corrosion state a break-up was made at the position 90 degrees south near ground level, see fig. 4.

In order to verify the corrosion situation, the structure close to the ground was opened in the direction of 90 degrees to the south.

Fig. 4 Corroding reinforcement. Cross section loss: 1-2 mm

Figure 4. Corroded rebar, loss of section is about 1-2mm

As the constructions have been examined close during the last 20 years it is possible to make a good estimate of the initiation of corrosion. Calculation of the average corrosion rate from the cross section loss of app. 2 mm and assuming the corrosion was initiated after app . 10 years gives an average corrosion rate of 9µA / cm22, which is with in the range of corrosion rates determined at this position by the GPM. The very low half-cell potentials agree with the high corrosion activity.

In the past 20 years, due to the very precise inspection and testing of the building, it is possible to make a relatively good assessment of the initial corrosion. The calculated average corrosion rate of the cross section is close to 2mm, assuming that the average corrosion rate will reach 9µA / cm2 after about 10 years, at which time corrosion will begin. The corrosion rate range of this area is obtained by GPM test. The very low half-cell potential is suitable for highly active corrosion state testing.

2.2 Example 2

2.2 Case 2

The next two examples are from a bridge foundation and a bridge deck in Greenland. The foundation was investigated in the tidal zone as shown in fig. 5 [6].

The next two cases are the pier and deck of a bridge in Greenland. The bridge piers under investigation are in the intertidal zone, see Figure 5.

Fig. 5. The investigated area and the location of the chloride profile.

Figure 5. Survey area and location of chloride profiles

The chloride concentration in the depth of the reinforcement is in the range between 0,3% and 0,7% of the concrete weight. As indicated by the half-cell potential measurements corrosion should therefore be expected. However the measured corrosion rates are low and the verification by visual inspection (fig. 7) shows no damage to the reinforcement.

The chlorine content of the concrete at the reinforcement is approximately 0.3% to 0.7%. According to the half-cell potential measurement, there should be corrosion, but the measured corrosion rate is very high, and through visual observation of the steel, no corrosion or damage has occurred.

Fig. 6 HCP and Icorr of bridge foundation half-cell potential and corrosion current

Fig. 7 Photo of break-up

Figure 7 The picture after opening the structure

2.3 Example 3

2.3 Case 3

The bridge deck from the same bridge in Greenland showed very different results as shown on fig. 8 and fig. 9.

Figures 8 and 9 show completely different results for the deck of the same bridge in Greenland.

Fig.8. The investigated area and the location of the chloride profile.

The typical dept of the reinforcement is here minimum 40-50 mm and the chloride concentration in this depth is near 0, 3% of the concrete weight. At this high chloride concentrations the half-cell potentials are expected to be low but the most negative values ​​measured are all above -100 mV vs. Ag / AgCl (fig. 9).

When the steel bar is at least 40-50mm deep in the protective layer and the chloride ion content is close to 0.3%, the half-cell potential should be relatively low, but the potential values ​​measured with the Ag / AgCl reference electrode are all above -100 mV (Figure 9).

Fig. 9. The read circle indicates the location of the chloride profile and the break-up shown at fig. 11.

Figure 9. The circle shows the chloride profile. Figure 11 is an image of the open structure.

The corrosion rate map fig.10 shows a completely different picture and indicates active corrosion at several locations.

The corrosion rate diagram of Figure 10 shows completely different pictures and indicates several areas where corrosion is active.

Fig. 10. Corrosion rate of the bridge deck. The read circle indicates the location of the

chloride profile and the break-up shown at fig. 11

Figure 10 Corrosion rate of the bridge deck, the circle indicates the chloride profile. Figure 11 is the picture after opening the structure

Fig.11. The first picture shows the corroded reinforcement and some water from cooling

the diamond-cutting blade. The second picture shows the damage and the mortar repairs.

Figure 11. The first photo shows corroded rebar and some water cooling the diamond cutting blade. The photos in Chapter 2 show the damage of bridge deck and the repair of mortar.

General comments to examples 2 and 3

Summary notes for case 2 and case 3

This bridge is located in a very cold environment. During the measurements described above the temperature was 15 ° C at midday and this explains the rather high corrosion rates at the bridge deck. Further there were some damages to the concrete surface due to the traffic directly on the concrete surface and a lot of mortar repairs.

The bridge is in a very cold environment. During the measurement, the temperature at noon was 15 ° C. There will be a higher corrosion rate on the bridge deck, because the traffic load directly acts on the concrete surface, so the concrete surface is more damaged and there are many mortar repairs.

2.4 Example 4

2.4 Case 4

In this example of swimming pool wall the conditions for performing corrosion rate measurements were not ideal. Tiles covered the inside of the swimming pool but preliminary performed GalvaPulse measurements showed that the joint filler was porous. Due to this fact it was possible to conduct the corrosion rate measurements by means of GalvaPulse equipment. The outside reinforcement was corroding at the casting joint between the pool floor and the pool walls and it was found necessary to investigate the inside reinforcement although no rust stains were visible.

The corrosion rate test environment of the swimming pool wall in the case is not ideal. The inner walls of the swimming pool are covered with ceramic tiles, but the initial test results using GalvaPulse show that the joint material is porous. Based on this fact, it is feasible to use GalvaPulse to test the corrosion rate. The external reinforcement at the joint between the bottom and the wall of the swimming pool has been corroded. Even if rust spots are not visible, it is necessary to measure the internal reinforcement.

The results were projected to a plane plot where the cast joint is in the centre of the plot fig. 12.

The test results are reflected in a plan view, the center of Figure 12 is the location of the junction.

Figure 12 Corrosion rate diagram along the bottom of the pool, through the junction and then to the wall

Even at these un-ideal measuring conditions the GalvaPulse pointed out the reinforcement corrosion points which was confirmed by visual inspections in breakups fig. 13.

Even in an undesirable test environment, GalvaPulse can accurately indicate the location of the steel bar corrosion, as shown by the actual photo in Figure 13.

3. CONCLUSIONS:

3. Conclusion:

1. Two techniques for evaluation of reinforcement corrosion, half-cell potential (HCP) and galvanostatic pulse measurements (GPM) are presented and discussed.

1. There are currently two techniques for evaluating steel corrosion, the half-cell potential (HCP) and the Galvanostatic pulse current test (GPM).

2. The evaluation of corrosion by means of the traditional half-cell potential technique using the existing standards may lead to mistakes in cases where the concrete is water saturated, carbonated and also exposed to the very low temperature.

2. Concrete structure under low temperature, carbonation and saturated water environment, if the current standard half-cell potential method is used for corrosion assessment, it will easily lead to erroneous results.

3. Complimentary measurements by means of galvanostatic pulse technique determining the corrosion rate contribute to the unambiguous evaluation of reinforcement corrosion also under conditions where the results obtained by the HCP technique could be misleading.

Using Galvanostatic Pulse Current Test (GPM) to test the corrosion rate is an admirable test method. Even in the case where it is easy to obtain inaccurate results with the HCP half-cell potential method, the corrosion of the steel bar can be clearly evaluated.

4. Four examples from on-site measurements are presented. Three of them show the need of using corrosion rate measurements together with half-cell potential for reliable evaluation of the actual corrosion state.

The four field test cases introduced in the article, three of which illustrate that to properly evaluate the corrosion status, in addition to using the HCP half-cell potentiometric method, simultaneous corrosion rate test is very necessary.

5. Passive areas are defined by galvanostatic pulse measurements as areas where the potential curve on the instruments computer screen has not reached a steady-state after pulsing over 5-10 seconds. In areas with active corrosion, areas where the potential curve exhibit a steady -state potential after 5-10 seconds, the corrosion current is measured as accurate as it can be expected from an on-site measurement taking into account the variation of the area of ​​the reinforcement polarized over, the actual corroding area of ​​the reinforcement and the inherent variations in moisture condition of the concrete and the temperature.

After using the GPM method to pulse an area for more than 5-10 seconds, when the voltage curve displayed on the computer screen of the instrument cannot reach a stable and constant level, it indicates that the area is an inactive corrosion area. In the active corrosion area, the voltage curve will stabilize after applying the pulse for 5-10 seconds, and the corrosion current measured at this time will be accurately measured. During the on-site measurement, attention should be paid to the changes in the polarization area of ​​the steel bar, the actual steel bar corrosion area and the inherent temperature change in the wet concrete environment.

6. It is not possible to estimate the actual loss of cross sectional area of ​​the reinforcement from a single GPM measurement. If multiple GPM measurements are taken over a period of time, an average value can be estimated. Alternatively the reinforcement must be exposed in the most corrosion active areas as done in these 4 examples.

It is impossible and inadequate to use only one GMP measurement to evaluate the actual loss in the cross-reinforcement area. If multiple GPM measurements are performed over a period of time, the average value can be used for evaluation. During the measurement, as in the four cases, the steel bars must be exposed to the active corrosion area.

4. REFERENCES:

[1] American Society of Testing and Materials: “Standard Test Method for Half-Cell Potentials of uncoated Reinforcing Steel in Concrete” ASTM C876, 1987.

[2] H. Arup: "Potential Mapping of Reinforced Concrete Structures", The Danish Corrosion Centre Report, January 1984

[3] B. Elsener, O. Klinghoffer, T. Frølund, E. Rislund, Y. Schiegg and H. Böhni: ​​"Assessment of Reinforcement Corrosion by Galvanostatic Pulse Technique, Proc. Int. , Norway, pp 391-400,1997.

[4] J. Mietz and B. Isecke: "Electrochemical Potential Monitoring on Reinforced Concrete Structures using Anodic Pulse Technique", in "Non destructive Testing in Civil Engineering" ed. Bungey, H., The British Institute of NDT, 2, 567 , 1993.

[5] T. Frølund. FM Jensen and R. Bässler: "Determination of corrosion rate by means of the galvanostatic pulse technique", First International Conference on Bridge Maintenance, Safety and Management, IABMAS 2002, Barcelona 2002.