The specimens were prism-shaped reinforced concrete elements, with a dimension of 18×18×8cm3, which were reinforced by a grid 16×16cm2 composed of six steel bars (5mm diameter) welded symmetrically forming squares of 5cm side, and placed 2cm under the anode, Fig. 1. The concrete dosage was as shown in Table 2. Two mixes were prepared: one without added chloride for preparing the specimens intended for pure cathodic prevention (CPre) treatments, and another mix containing 2% Cl− relative to cement mass for the specimens used in cathodic protection (CP) applications or combined treatments (ECE+CP) and (ECE+CPre), see Table 1. Once the formwork was removed, the specimens were moist-cured at 95–98% relative humidity (RH) for 28 days. The characteristics of the hardened concrete were as follows: compressive strength 37.8N/mm2[32], porosity 11.1% [33], and bulk density 2.38 T/m3[34].

Figure 1.

Sketch of reinforcement of samples and connection of the cathodic system (steel reinforcement).

(0,17MB).

Adapted from [24].

Fig. 1 also shows the system used for connecting the reinforcement (cathode system) to the negative pole of the electric power source, through plastic isolated copper connectors screwed to the rebar.

All the electrochemical tests were performed using a CCP anode. The anodic overlay system was prepared by hardening a graphite–cement paste (GCP) obtained by mixing graphite powder and Portland cement at 50%–50% in mass. Water to solid mix ratio was 0.8. Secondly, a 2mm thick layer of this paste was applied on the surface of each specimen, and then all of them were moist-cured for 10 days. After that, two grooves were performed lengthwise onto the anodic overlay, without reaching the concrete surface, in order to receive both graphite rods to connect to the positive pole of the electric source. To finish up, these rods were overlaid with graphite–cement paste in order to join with the anode system perfectly, but avoiding any contact between graphite rods and concrete. A PVC receptacle was assembled on the top of the samples to retain both ECE electrolyte (distilled water) and/or the dissolution used during CP or CPre applications to simulate a continued chloride contamination (65ml NaCl 0.5M applied weekly), Fig. 2. The ratio between the surface of concrete covered by the anodic overlay and the surface of the primary anodes (graphite rods) was 9.6; the ratio between the surface of concrete covered by the anodic overlay and the total surface of the steel bars was 1.7, Fig. 2. All current density and electric charge density values reported in this work are relative to surface of concrete, (equal to the anode surface), except otherwise stated.

Figure 2.

Sketch of specimen assembly.

(0,2MB).

Adapted from [24].

The resistivity of the graphite–cement paste was measured through the four-probe method. To this end, paste specimens were cast in 4×4×16cm3 moulds, and moist-cured at 95–98% RH during 14 days. The experimental details of the measurements can be found elsewhere [35]. The average obtained resistivity was 1.5Ωm.

The measurements of steel corrosion potential (Ecorr) and all the single electrode potentials, were performed using Ag–AgCl reference electrodes. These electrodes were housed in respective holes drilled from the exposed surface of the concrete specimen (that bearing the graphite–cement anode) to the vicinity of the rebar, Fig. 2. For this purpose, the holes were sheathed with a plastic tube, and filled with a KOH 0.2M aqueous solution, trying to approach the physical-chemical conditions of the concrete's inner pore solution.

In this way nine specimens were prepared, seven of them with salt in the mixing water and two without it. Two of the salted specimens were used only for determining the efficiency of the ECE process. Concrete cores were extracted from them, and their chloride content profiles were determined, in one case before and in the other case after ECE. These two specimens were discarded after coring. The seven remaining specimens were intended for the tests corresponding to CP, CPre, and the combined treatments (ECE+CP) and (ECE+CPre), see Table 1.

The ECE efficiencies were calculated as the percentages of reduction of the initial chloride content. Obtaining the chloride content profiles before and after the ECE trials allowed calculating the local and overall efficiencies. At the end of the 24 weeks of exposure to a severe Cl− load, i.e. the first phase of CP and CPre treatments, see Section 2.4, chloride content profiles were obtained from all the reinforced concrete specimens. The Cl− profiles were measured following essentially RILEM recommended procedures [36]. Cylindrical concrete cores, 95mm diameter and 20mm height (up to the rebar depth), were extracted. From these cores, concrete dust samples were obtained by grinding thin (2mm thick) successive parallel layers to the exposed surface [36]. In this way 10 concrete dust samples were gained from each core, thus allowing the obtention of sufficiently detailed chloride content profiles. The determination of the samples’ acid soluble chloride contents was carried out by potentiometric titration [37,38]. All the chloride content values are expressed in this work as % Cl− relative to cement mass.

Four of the specimens made with saline mixing water were subjected to ECE. For this purpose, the specimens were electrically connected in series by pairs to a direct current source. The relevant parameters of the ECE treatments are shown in Table 3. A low charge density was applied, only 1.5 MC/m2 relative to concrete surface (2.6MC/m2 relative to steel surface). The current source feeding voltage (ΔEfeed) was controlled at a level below 40V during the processes, for safety reasons. It was necessary in this respect to interrupt the current passage two times (two pauses of 24h each one) along the treatment. The whole processes were performed inside a fume hood to eliminate the chlorine, Cl2(g), produced by electrochemical oxidation of the Cl− ions extracted from the concrete. When the ECE processes were finished, the pH values of the electrolytes were measured in order to check the acidifying effect caused by the electrochemical anodic reactions.

This section describes the details of the experiments carried out to demonstrate the applicability of the GCP anodes for combined protective treatments of ECE+CP and ECE+CPre, please refer to Table 1.

The CP treatments were applied with 15mA/m2 of current density (relative to concrete or anode surface) to two specimens, one of those previously treated with ECE (EA in Table 1) and the other one without previous treatment (A in Table 1). On the other hand the application of CPre with 2mA/m2 of current density (relative to concrete surface) to a sample previously ECE treated (EB in Table 1) and to another manufactured without salt and not ECE treated (B in Table 1). The values of the current densities relative to the steel bars surface are 25.5mA/m2 and 3.4mA/m2 for the CP and Cpre treatments, respectively. The direct current source was programmed to maintain a constant current density along the whole processes. Each application has its reference sample without treatment of CP or CPre to compare the results. The reference sample for the specimens subjected to combined treatments (EA for ECE+CP, and EB for ECE+CPre) is ER. R is the reference for A, and P is the reference for B.

The application of CP and CPre consisted of two phases:

Phase 1. First 24 weeks. The aforementioned treatments CP and CPre were continuously applied during the first 13 weeks. Then, the current was switched off for 4 weeks and after that, treatments were resumed to the end. Chloride contamination was continuously applied, even during the switch off periods.

While applying the CP and CPre treatments some electrochemical parameters were measured. During the current passing periods the feeding voltage of each specimen, ΔEfeed, was obtained as the potential difference between cathode and anode; and the individual anodic and cathodic potentials, Ea and Ec, respectively, were measured against the reference electrode Ag–AgCl. Finally, in order to check the efficiency of CP and CPre as maintainers of protection conditions of steel, the “100mV decay” criterion was used, as is specified in ISO 12696:2012 [39]. This criterion has been also extensively employed for this purpose by several researchers [21,40]. The method consists in obtaining the 4h potential decay (ΔEdecay), that is the difference between Ec4h (the value of Ec 4h after the current switch off), and the instant-off cathodic potential Ecio, which in this case was measured 1s after the current switch off. The minimum value of the 4h depolarization must be 100mV for an adequate corrosion protection of steel [39]. The values of Ecio were monitored with an automatic data logger able to obtain and record 500 measurements in 6s, after the current switch off.

Once the 24 weeks processes were fulfilled, cores were extracted from all specimens of Table 1, and their respective Cl− content profiles were obtained. This was done with the purpose of evaluating the net effect of the electrochemical treatments on the Cl− ion uptake by the reinforced concrete specimens during the continued exposure to a very aggressive environment.

Phase 2. At the end of Phase 1 it was observed that all the specimens had lost the steel protection condition, evidenced by ΔEdecay values lower than 100mV. Then, it was decided to start this second phase with the objective of recovering the protection conditions of steel by adjusting the current density of the CP treatments. The procedure was to increase progressively the current density during 4 weeks, starting with a value of 20mA/m2, until obtaining the protection conditions.

Four of the reinforced concrete specimens were subjected to an ECE treatment before starting the first phase of the CP and CPre treatments, see Section 2.3. Once finished the ECE process with the settled parameters, Cl− content profiles were obtained, corresponding to the states before and after the ECE treatment. The local ECE efficiencies, understood as percentage of Cl− content removed, are plotted in Fig. 3. The average of removed Cl− was 51% of the initial content, i.e. the residual Cl− content of concrete after ECE was approximately 1% referred to cement mass. This indicates a good performance of the ECE process applied on a conventional ordinary Portland cement concrete with the GCP anodic overlay system, for a relatively low charge density of 1.5×106C/m2 relative to concrete surface. This result can be compared to the 41% efficiency obtained for a very similar reinforced concrete element, with the same initial amount of Cl−, subject to an ECE treatment, using a Ti–RuO2 mesh anode, and passing a total charge density of 1MC/m2 relative to concrete surface [26].

Figure 3.

Chloride concentration profiles before ECE (initial) and after ECE (final), and local efficiencies of the extraction process. ECE details: current density: 2A/m2, charge density: 1.5MC/m2, both related to exposed concrete surface.

(0,09MB).

Adapted from [24].

The pH values determined for the electrolytes, after the ECE experiments, were in the range 5–5.5. This means that the acidity of the electrolytes had increased slightly, (the initial electrolyte was distilled water). This acidifying effect may be due to several of the electrochemical anodic reactions, namely the oxidation of hydroxide ions or water to O2(g) and the oxidation of carbon to CO2(g). It is possible that the alkaline character of the graphite–cement paste anode may have reduced the acidity produced by those electrochemical processes. It is worth noting that no sharp increase of the feeding voltage was observed during the ECE treatments. This suggests that no important damage of the GCP overlay has been produced during ECE [21], even though these trials are performed with a high current density of 2A/m2. We must recall that the ECE trials are performed under continuous liquid ponding of the anode, and its duration is short, about eight days.

In this section we describe the results of tests carried out to investigate the performance of the GCP anodes during protective electrochemical treatments to reinforced concrete elements affected by steel corrosion due to chloride contamination, such are CP, CPre and combined treatments of ECE+CP or ECE+CPre, see Table 1.

To verify the effectiveness of CP and CPre treatments for protecting steel from corrosion, the “100mV decay” criterion [39] was used, as stated in Section 2.4. Fig. 4 shows the evolution of the ΔEdecay values for the specimens in Table 1, during the 24 weeks experiments. The ΔEdecay values of the A specimen, treated only with CP, practically never reached the threshold value of 100mV. It seems that a 15mA/m2 current density was not sufficient to provide protection to steel in such harsh conditions: initial Cl− content of 2% plus the continued salting regime (65ml NaCl 0.5M weekly sprayed onto the anodic overlay surface). Regarding the EA specimen (ECE+CP), the protection conditions of steel were kept during 11 weeks because of the current circulation, despite external Cl− loading. Cathodic protection of 15mA/m2 current density, relative to concrete surface, was able to keep protective conditions for the steel reinforcement if the initial Cl− content of the specimen was about 1% (EA). In the case of the specimen with initial Cl− content of about 2% (A), a higher current density would be needed for reaching the protection conditions [9,11]. These observations corroborate the main hypothesis of the present research, i.e. in cases of reinforced concrete structures with a high Cl− contamination subject to very harsh chloride environments, it would be advantageous to apply successively an initial ECE treatment to reduce the Cl− content, and then maintain protective conditions to steel through a continuous CP treatment without the need of using a too high CP current density, which could eventually impair the performance of the anodic system [21]. These combined treatments, ECE+CP, would be more conveniently implemented with the CCP coatings, since the same anode can serve both for the ECE and for the CP treatments.

Figure 4.

Evolution of ΔEdecay during the CP or CPre treatments. A: CP; B: CPre; EA: ECE+CP; EB: ECE+CPre. All of them subjected to Cl− contamination during the 24 weeks. The electrochemical treatments were interrupted between week 13 and week 17.

(0,09MB).

Adapted from [24].

Regarding the CPre treated specimens (cathodic current density 2mA/m2 relative to concrete surface), specimen B showed ΔEdecay values higher than 100mV up to the current switch off at 13th week, confirming the capacity of continuously applied CPre treatments to keep steel protection conditions for an initially Cl− free reinforced concrete, despite extensive external Cl− load [9,11,41,42]. On the other hand such a low current density is unable to protect steel if concrete is previously contaminated at a level of about 1% relative to cement mass, and steel has started to corrode, as shown for specimen EB (ECE+CPre) in Fig. 4.

At the end of Phase 1 all the reinforced concrete specimens had reached a very high degree of Cl− contamination, as can be appreciated in Table 4. Nevertheless, some comparisons can be done on the behaviour of the different cases. For instance, Fig. 5 shows that the specimens treated with ECE+CP (EA) or ECE+CPre (EB) during Phase 1, have experienced less Cl− ingress than the reference specimen ER, which after the ECE trial was left untreated during the Phase 1. This represents a further evidence of the “chloride barrier effect”, mentioned by Pedeferri [9], as one of the beneficial secondary effects of CP, as the polarity of the electric field induces a repellent effect of the negative ions, thus reducing the chloride uptake by concrete. The effect can be quantified by using a Cl− barrier effect parameter (Cl− BEP), which is defined here as the percentage of reduction of the Cl− uptake by concrete, due to CP or CPre actions, referred to the Cl− uptake of the reference, (not treated) specimens, see Eq. (1).

Figure 5.

Profiles of Cl− content in specimens treated previously with ECE and after with CP (EA), with ECE and after CPre (EB), and in the reference sample only treated with ECE (ER), all of them subjected to Cl− contamination process, after ECE and first phase of CP or CPre (24 weeks).

(0,08MB).

Adapted from [24].

Table 5 shows the net values of the Cl− barrier effect parameter found for the specimens tested in this work, putting in evidence that this parameter reaches values of about 20% in the experimental conditions of this work.

Given that after the 24 weeks of treatments of Phase 1, including rest periods between the 13th and 17th weeks, the steel reinforcements in all concrete specimens had completely lost their protection conditions, and having been demonstrated that a CP of 15mA/m2 was unable to restore the protective conditions (Fig. 4), the Phase 2 was started. The external Cl− load was discontinued since all the specimens had reached very high Cl− contents, see Table 4. In these conditions, CP was applied with higher current densities. The question was if it would be possible to recover the protection conditions of steel by increasing the current density to the appropriate value. In the beginning of this last phase, the current density was set at 20mA/m2. After 4 weeks in operation the threshold value of 100mV ΔEdecay was not reached, i.e. the protection conditions were not obtained, Fig. 6. Neither success was obtained in a second attempt at 25mA/m2 (data not shown in Fig. 6). Finally, a third step of 40mA/m2 was set. In this case, after 4 weeks, the rule of 100mV of ΔEdecay was achieved for EA, A and B specimens.

Figure 6.

Evolution of ΔEdecay during Phase 2 of CP. First step of 4 weeks with 20mA/m2 of current density, and third step with 40mA/m2. Specimens A, B, EA and EB.

(0,1MB).

Adapted from [24].

Moreover, protection conditions were verified with the measurement of depolarization potential difference values 7 days after switch off [39]. In fact, more than 150mV of ΔEdecay was reached after 7 days (209mV for EA, 211mV for A and 153mV for B). This efficiency of CP is similar to that obtained by other researchers [40,43]. However, this was not the case for EB. After 4 weeks with a cathodic current density of 40mA/m2 the protection conditions were not recovered for this latter specimen.