U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
202-366-4000


Skip to content
Facebook iconYouTube iconTwitter iconFlickr iconLinkedInInstagram

Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations

Report
This report is an archived publication and may contain dated technical, contact, and link information

Federal Highway Administration > Publications > Research > Structures > Electrochemical Chloride Extraction: Influence of Concrete Surface on Treatment

Publication Number: FHWA-RD-02-107
Date: September 2002

Previous | Table of Contents | Next

Chapter 3. Results and Discussion

Currently, this study has involved a number of specimens to obtain the following results. ECE studies have been completed on 20 of the Type I and 12 of the Type II test blocks. These blocks included different cover depths, which ranged from 3.5 cm to 6.4 cm. To evaluate the affect of the w/c ratio on ECE, test blocks with ratios ranging from 0.40 to 0.60 were designed. Descriptions of the block designs are presented in tables 7 through 12 and illustrated in figures 4 and 5.

Changes in the Current and Voltage during ECE

During ECE, the current was monitored, while internal and external voltage measurements were made. The designations in the following graphs to the different measurement points that were monitored during ECE for the Type I and Type II test blocks are listed in tables 11 and 12, respectively. Figures 4 and 5 can provide additional insight into the connections used to make measurements.

The typical change in voltage during ECE for a set of Type I specimens with different w/c ratios, but the same cover depth is shown in figure 10. The benefit of these specimens was the ability to measure the voltage changes in different layers between the anode and cathode. Under constant voltage conditions, it can be seen that the voltage between the anode and the upper titanium rods increases. At the same time, the voltage between the lower titanium rods and the reinforcing steel is decreasing. In each case, the rate of change of the voltage is greatest initially, i.e., during the first 10 to 15 days, and then the rate of change decreases for the remaining extraction period. All of the specimens evaluated exhibited this type of behavior. Even with larger applied voltages, the voltage difference in the top layer of concrete increased during ECE. A typical example of this is shown for two specimens in figure 11. As the applied voltage to the slab increases under constant current conditions, the voltage across the top layer of concrete increases.

This figure is composed of four plots showing the voltage vs. elapsed time for Type I specimens, each for a different water to cement ratio.  It demonstrates that as the voltage in the region between the anode and rebar remains constant (9 V), the voltage in the region between the anode and upper titanium rods increases and the voltage in the region between the lower titanium rods and rebar decreases.

Figure 10. Measured voltage changes during ECE in a set of Type I specimens (Maximum voltage was 9V)

Part 1: This figure shows that for two Type II specimens with different water to cement ratios the voltage between the anode and the rebar closely follows the voltage in the region between the anolyte and the titanium strip in the concrete.  This is shown to occur under constant current and constant voltage conditions.

Part 2: This figure shows that for two Type II specimens with different water to cement ratios the voltage between the anode and the rebar closely follows the voltage in the region between the anolyte and the titanium strip in the concrete.  This is shown to occur under constant current and constant voltage conditions.

Figure 11. Comparison of the voltages measured between (1) the anode and rebar and
(2) between the anolyte and the titanium strip in the concrete. Top: for a 0.55-w/c Type II
specimen; Bottom: for a 0.50-w/c Type II specimen (Maximum voltage was 40V)

Influence of Concrete Surface on Voltage and Current

To better understand the influence of the surface-layer concrete on the voltage and current, one of the Type I specimens was subjected to an extended ECE experiment and analysis. A timeline illustrating the sequence of events that took place in this extended experiment is shown in figure 12. First, the specimen was ECE treated or polarized for 26 days and then depolarized for 38 days. The voltage and current data for this 26-day ECE treatment are shown in figure 13a. Then, the specimen was re-energized for 12 hours. The voltage and current data for this first 12-hr polarization are shown in figure 13b. This was followed by removal and storage of the electrolyte used during the ECE test period. The surface of the concrete was then sandblasted and the stored electrolyte was poured back into the reservoir. After allowing twenty hours for the solution to soak into the concrete, ECE was initiated again for another 12 hours and the same measurements were made (figure 13c). Finally, the sample was depolarize for 36 hours and then re-energized for a final 12-hour duration and a final set of measurements was made (figure 13d).

Comparison of figures 13a and 13b would indicate that the voltages and current of the system at the start of the first 12-hour re-polarization were practically the same as where the system was at the end of the 26-day of polarization. Comparison of figures 13b and 13c would reveal the effect of the sandblasting the surface layer concrete on the voltages and current. It is clear that sandblasting the surface of this specimen increased the current density by over 0.3 A/m2. This current density was even greater than the initial current density value (at the beginning of the 26-day treatment). Following sandblasting, the voltage between the anode and the reinforcing steel also switched from constant-voltage mode to constant-current mode, but then returned to a constant-voltage mode. In addition, sandblasting resulted in voltage changes within the different concrete regions, which are on the order of approximately 2 V. This indicates that the surface of the concrete appears to have a significant influence on the voltage and current during ECE. However, additional testing on other specimens will be required to confirm these observations.

This figure is a timeline showing the sequence of events that took place during the concrete surface study on a Type I specimen.  This figure displays how the block was subjected to the following sequence of events:  ECE for 26 days, depolarized for 36 days, reenergized for 12 hours, sandblasted and reponded for 20 hours, reenergized for 12 hours, depolarized for 36 hours, and finally reenergized for 12 hours.

Figure 12. Timeline of concrete surface study

Figure 13. Influence of concrete exterior surface on the voltage and current in a 0.45-w/c Type I specimen

This figure contains four graphs, each representing a different phase of the surface study on a Type I specimen (0.45 water to cement ratio).  All of the graphs have multiple Y-axes, corresponding to voltage and current density measurements, which are both plotted against the treatment time period.  13(A) demonstrates how the voltage in the layer between the titanium strip and the top titanium rod increases during treatment as the voltage in the layer between the bottom titanium rod and rebar decreases.  During this period all other voltage values remain constant and the current density is below the preset maximum.  In 13(B) all of the voltage and current density values begin where they ended prior to the depolarization step.  The voltage values remain constant thoughout this treatment period and the current density decreases in the first two hours and then remains constant.  13(C) shows a significant increase in current density (above that even seen initally) and increases in voltage in the layer between the top and bottom titanium rod as well as the layer between the bottom titanium rod and rebar.  Concurrently, the voltage between the anode and rebar as well as the region between the top titanium strip and top titanium rod are inially lower but increase as the treatment continues.  When compared to 13(C), the voltage and current density exhibit a similar response in figure 13(D).

Changes in the Concrete Resistance During ECE

The resistance between different points in the concrete was determined using the IR Drop technique, which was discussed in the earlier section "IR Drop Measurements" on page sixteen. During ECE, the resistance was observed to increase in the region between the anode and the upper layer of titanium rods, as shown in figure 14. In contrast, the solution (concrete) resistance decreased in all other regions during ECE. A comparison, based on w/c ratios, of the resistance between the anode and the upper titanium rods did not indicate an obvious relationship. This pattern was consistent in all of the samples studied.

This figure is composed of four resistance vs. elapsed time plots, each corresponding to Type I specimens with a different water to cement ratio.  It demonstrates that while the resistance in the layers between the anode and upper titanium rod increases, the resistance in the other layers decreases.

Figure 14. Example showing the change in resistances for a single set of Type I specimens during ECE

Resistivity

It was apparent that, immediately before ECE, the resistivity of the top layer of concrete, as measured by the upper row of four titanium rods, was less than that of the lower concrete that surrounded the lower set of four titanium rods, as shown in figure 15. This difference was observed in the other specimens prior to ECE. It is also apparent in figure 15 that the resistivity of the lower region of the concrete undergoes only a relatively slight to moderate change during ECE. In contrast, the resistivity of the top layer of concrete increased and exceeded that of the lower concrete and then remained at a level that is greater than that of the lower concrete region. However, a clear relationship between w/c ratio and resistivity for either the upper or lower titanium rod region was not evident. This is shown in figures 16 and 17, respectively.

This figure is composed of four resistivity vs. elapsed time plots, each corresponding to Type I specimens with a different water to cement ratio.  It demonstrates that while the resistivity in the upper layer of concrete is initially lower that the resistivity in the lower layer, the upper layer resistivity becomes greater than the lower layer resistivity.

Figure 15. Change in Resistivity for Type I specimens of various w/c ratios (cover thickness of 4.4 cm)

This figure of the resistivity vs. elapsed time plot is for the upper layer concrete measurements on the Type I specimens and it compares the different water to cement ratio.  It demonstrates that a trend based on water to cement ratio is not evident.  The water to cement 0.55 line clearly separates from the other three lines and exhibits a lower resistivity value when compared with the other water to cement ratios.  The water to cement 0.40 line displays the highest resistivity values and rate of increase when compared with the other water to cement ratios.  However, the water to cement 0.50 and water to cement 0.45 lines overlap each other as they increase in resistivity with continued treatment time.

Figure 16. Resistivity change in the upper layer of concrete, for Type I Specimens with various w/c ratios (cover thickness of 4.4 cm)

This figure of the resistivity versus elapsed time plot is for the upper layer concrete measurements on the Type I specimens and it compares the different water to cement ratio.  It demonstrates that a trend based on water to cement ratio is not evident.  The water to cement 0.55 line clearly separates from the other three lines and exhibits a lower resistivity value when compared with the other water to cement ratios.  The water to cement 0.45 line displays the second lowest resistivity values and rate of increase when compared with the other water to cement ratios.  The water to cement 0.50 and water to cement 0.40 lines overlap each other as they increase in resistivity with continued treatment time.

Figure 17. Resistivity change in the lower layer of concrete, for Type I Specimens with various w/c ratios (cover thickness of 4.4 cm)

Chloride Concentration in Concrete with ECE

The previous data have shown that the concrete specimens studied behaved in a stratified manner during ECE. Generally, the resistivity in the upper layers of concrete increased while that at the lower layers decreased. Upon determining the chloride concentration before and after ECE in Type I specimens, it is evident that a large change in chloride concentration occurs near the surface adjacent to the anode, which is shown in figures 18 and 19. (The negative values in figures 18 and 19 indicate a decrease in chloride concentration; conversely, positive values signify an increase in chloride concentration.) It is apparent that significant removal of chlorides during ECE are occurring in the concrete layer near the anode. In contrast, the treated Type II specimens exhibited a more even removal distribution at deeper depths in the concrete, which is shown in figures 20 and 21.

Unlike the Type II specimens, the Type I specimens exhibited a decrease in chloride removal at deeper depths within the concrete and some specimens even display an increase in chloride concentration near the reinforcing steel. This is attributed to two factors: (1) the use of admixed chlorides in the Type I specimens and (2) the difference in the steel surface area in the different specimen types (table 13). It is possible that the admixed chlorides below the bar were drawn upward during ECE. In addition, these results support earlier studies that indicate the importance of available cathode area. The decreased cathode surface area in the Type I specimens appears to decrease the total current flow, which decreased the percentage of chloride removed. In the Type I specimens, 0% - 52% of the chlorides were removed during treatment, whereas 33% - 76% of the chlorides were removed from the Type II specimens.

Cover thickness appears to influence the quantity of chloride removed. However, additional samples must be evaluated to provide statistical validity to the observed trend.

This figure is a bar graph that shows the average change in chloride concentration at three depth ranges (0.0 to 0.6 centimeter, 0.6 to 0.41 centimeter, and 4.1 centimeter to rebar) as well as the overall change for Type I specimens with 4.4 centimeter of concrete cover.  All four water to cement ratios are plotted on the same figure.  The water to cement 0.40 catagory sampled from 4.1 centimeter to rebar, displays an increase whereas all other categories show decreases in chlorides following treatment.

Figure 18. Average change in chloride concentrations due to ECE in Type I specimens with
4.4 cm of concrete cover over rebar

This figure is a bar graph that shows the average change in chloride concentration at three depth ranges (0.0 to 0.6 centimeter, 0.6 to 0.41 centimeter, and 4.1 centimeter to rebar) as well as the overall change for Type I specimens with 5.7 centimeter of concrete cover.  All four water to cement ratios are plotted on the same figure.  The water to cement 0.40 catagory sampled from 4.1 centimeter to rebar shows no change in chloride concentration following treatment.   The water to cement 0.40 catagory sampled from 0.6 to 4.1 centimeter and the water to cement 0.45, water to cement 0.50, and water to cement 0.55  catagories sampled from 4.1 centimeter to rebar display increases. All other categories show decreases in chlorides following treatment.

Figure 19. Average change in chloride concentrations due to ECE in Type I specimens with
5.7 cm of concrete cover over rebar

This figure is a bar graph that shows the average change in chloride concentration at three depth ranges (0.0 to 0.6 centimeter, 0.6 to 0.41 centimeter, and 4.1 centimeter to rebar) as well as the overall change for Type II specimens with 3.8 centimeter of concrete cover.  All four water to cement ratios are plotted on the same figure.  All categories show decreases in chlorides following treatment.

Figure 20. Change in chloride concentrations due to ECE in a single set of Type II specimens
with 3.8 cm of concrete cover over rebar

This figure is a bar graph that shows the average change in chloride concentration at three depth ranges (0.0 to 0.6 centimeter, 0.6 to 0.41 centimeter, and 4.1 centimeter to rebar) as well as the overall change for Type II specimens with 6.4 centimeter of concrete cover.  All four water to cement ratios are plotted on the same figure.  The water to cement 0.50, water to cement 0.55, and water to cement 0.60  catagories sampled from 0.0 to 0.6 centimeter display increases. All other categories show decreases in chlorides following treatment.

Figure 21. Average change in chloride concentrations due to ECE in Type II specimens
with 6.4 cm of concrete cover over rebar

Visual Observations

After completing ECE, each specimen's surface was examined for physical changes. In each case, a tightly adhering substance had formed on the surface. Images of this formation are shown in figure 22. Attempts to remove the unknown material from the surface proved difficult. Upon trying to cleave it, a portion of the concrete was dislodging in addition to the unknown material. This provided a cross section of the interface, which is shown in figure 23. A surface formation following treatment has been found on actual treated bridge decks.[70] As shown in figure 24, the formation parallels the reinforcing steel mat below. It is evident from these photographs that the surface formation creates perpendicular lines that crisscross the older concrete, as well as the recently repaired concrete (areas covered with more white material than their surrounding). Surface formation samples were collected and are currently being analyzed.

This picture shows the white substance on the surface of the concrete.  The substance completely covers the surface in some locations while other areas appear bare with the concrete clearly visible.  In some areas, the white substance exhibits tiny cavities that resemble little craters.

Figure 22. Tightly adhering layer of white material formed on the concrete surface during ECE

This picture shows four profile images of a piece of concrete dislodged from the surface following ECE.  The white substance on the surface of the concrete completely covers the surface and in some locations it penetrates into the larger surface pores.

Figure 23. Various views of surface layer that formed on the concrete during
ECE: (Left) top view, (Middle) edge views, (Right) bottom view

These two pictures show an actual bridge deck surface following ECE.  The white substance on the surface of the concrete creates perpendicular lines on the surface of the concrete.  These white lines are above the reinforcing steel that is embedded in the concrete.  These lines even stripe the recently placed concrete repair patches

Figure 24. Layer of white material formed on the concrete surface directly
above the reinforcing steel following ECE on an actual bridge deck [70]

The increased current flow because of sandblasting the concrete surface, which removed the residue, was discussed earlier. That particular specimen is shown in figure 25. It is interesting to note that white deposits formed in the exposed pores, which is shown in the bottom photograph in figure 25. As indicated earlier, the study of new samples (shown in figure 27) will add statistical validity to the observed improvement in current flow following sandblasting. It is expected that removal of the tightly adherent surface formation will improve the ECE process.

Part 3: These three pictures show the concrete surface at different stages during the surface study outlined in figure 12.  The top photograph shows a white substance covering the surface of the concrete with tiny cavities that resemble little craters in some areas.  The middle photograph shows a bare concrete surface (concrete pores exposed), with the white substance completely removed. The bottom picture exhibits the formation of white material inside the large concrete pores that were exposed during sandblasting.Part 3: These three pictures show the concrete surface at different stages during the surface study outlined in figure 12.  The top photograph shows a white substance covering the surface of the concrete with tiny cavities that resemble little craters in some areas.  The middle photograph shows a bare concrete surface (concrete pores exposed), with the white substance completely removed. The bottom picture exhibits the formation of white material inside the large concrete pores that were exposed during sandblasting. Part 3: These three pictures show the concrete surface at different stages during the surface study outlined in figure 12.  The top photograph shows a white substance covering the surface of the concrete with tiny cavities that resemble little craters in some areas.  The middle photograph shows a bare concrete surface (concrete pores exposed), with the white substance completely removed. The bottom picture exhibits the formation of white material inside the large concrete pores that were exposed during sandblasting.

Figure 25. Type I specimen surface appearance: (Top) after ECE but prior to
sandblasting, (Middle) after sandblasting but prior to second application of
ECE, (Bottom) following a second application of ECE for 12 hours

Surface Deposit Analysis

To identify the white material covering the concrete surface following ECE, which is shown in figures 22 and 23, XRD was performed on two different Type II specimens. Calcium carbonate was clearly identified by XRD to be present in both specimens. Figure 26 is an overlay of a calcium carbonate spectrum over the top of the spectrum for one of the unknown samples. It is apparent that all of the peaks locations in the standard match the peaks locations in the unknown sample. However, the intensities differed, which indicates additional crystalline materials are most likely present in minute amounts in the sample. In addition, not all of the peaks in the unknown sample diffraction pattern are accounted for by the calcium carbonate standard. Based on these observations, calcium carbonate appeared to be the major component in the white residue; however, the presence of other components in trace quantities is evident. The XRD spectrum from the other sample analyzed displayed the same characteristics as those found in the sample shown in figure 26. Table 15 lists the intensities and peak positions for the two samples examined.

XPS was also performed on the white material deposited on one Type I specimen and one Type II specimen. The Type I specimen used calcium hydroxide as the electrolyte, whereas industrial lime was used as the electrolyte for the Type II specimen. The XPS data indicated that calcium chloride was present on the surface of both samples. In addition, XPS detected magnesium in the sample exposed to a solution of industrial lime, but magnesium was not present in the sample that used reagent grade calcium hydroxide as the electrolyte. Work is underway to evaluate the significance of these finding through ongoing analysis on additional samples. Table 16 lists the elements and binding energies for the two samples examined.

This figure shows an overlay of the XRD spectrum for the unknown material on top of a calcium carbonate spectrum.  The two spectrums show strong agreement with each other over the entire range scanned (5.0 - 75.0 degrees).

Figure 26. Surface deposit XRD pattern from a Type II specimen comparing the
unknown material to calcium carbonate

Table 15. Surface deposit peak data using x-ray diffraction

Unknown #1 Surface Deposit

Unknown #2 Surface Deposit

2θ, Deg

Intensity, CPS

2θ, Deg

Intensity, CPS

13.080

61.000

12.380

78.333

13.320

54.000

12.980

95.000

14.040

56.000

13.220

88.333

14.340

84.000

13.320

71.667

14.500

56.000

13.580

88.333

14.800

59.000

13.660

131.667

15.120

63.000

13.849

81.167

15.369

63.467

14.140

128.333

15.760

56.000

14.300

103.333

15.880

63.000

14.460

80.000

16.160

56.000

14.680

76.667

16.340

54.000

15.040

80.000

16.760

69.000

15.240

68.333

17.120

71.000

15.460

56.667

17.340

48.000

15.800

116.667

17.660

52.000

16.160

56.667

18.061

59.950

16.460

113.333

18.700

56.000

16.620

76.667

19.520

63.000

17.120

75.000

20.180

59.000

17.440

93.333

20.900

60.000

17.620

61.667

23.119

162.533

17.820

106.667

29.461

2347.350

18.145

213.700

29.980

49.000

18.320

126.667

30.400

53.000

18.740

75.000

31.500

91.000

19.380

61.667

31.620

56.000

19.680

75.000

31.920

74.000

20.280

80.000

32.160

46.000

21.620

65.000

32.700

64.000

23.160

188.333

35.993

266.550

27.440

51.667

36.300

70.000

28.020

80.000

39.458

553.950

29.020

60.000

39.700

79.000

29.393

2408.000

43.201

415.500

29.690

95.000

47.190

190.050

30.460

63.333

47.555

498.533

30.620

88.333

47.900

54.000

31.500

83.333

48.557

614.350

31.560

103.333

56.627

126.150

32.520

51.667

57.446

248.483

33.020

86.667

60.708

180.717

33.560

53.333

61.051

132.933

34.040

55.000

61.412

70.233

34.186

81.850

64.699

162.000

34.300

93.333

65.644

77.183

36.051

251.367

70.320

56.000

39.471

405.367

72.960

93.000

43.226

448.333

   

47.228

202.267

   

47.585

467.133

   

48.585

488.733

   

56.624

58.000

   

57.436

178.167

   

60.739

137.117

   

61.480

50.000

   

64.714

120.733

   

65.718

80.733


Table 16. Surface deposit peak data using x-ray photoelectron spectroscopy

Unknown #1,
Ca(OH)2 electrolyte

Unknown #2,
Industrial lime electrolyte

Element
(Photoelectron Line)

Binding Energy, eV

Element
(Photoelectron Line)

Binding Energy, eV

C (1s)

285.9

C (1s)

286.3

Ca (2p 3/2)

348.6

Ca (2p 3/2)

349.0

Cl (2p 3/2)

199.1

Cl (2p 3/2)

199.1

O (1s)

533.2

O (1s)

533.4

Mg (2p 3/2)

None Present

Mg (2p 3/2)

51.0

Si (2p)

102.7

Si (2p)

102.9

Previous | Table of Contents | Next

Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000
Turner-Fairbank Highway Research Center | 6300 Georgetown Pike | McLean, VA | 22101