- Email: [email protected]
Effectiveness of corrosion inhibitors – a field study
18 Effectiveness of corrosion inhibitors – a field study Y. S C H I E G G, F. H U N K E L E R and H. U N G R I C H T, Swiss Society for Corrosion Protection (SGK), Switzerland and Technical Research and Consulting on Cement and Concrete (TFB), Switzerland
18.1
Introduction
The number of reinforced concrete structures showing signs of deterioration or damage due to corrosion of rebars has increased dramatically over the last 20 years. There is, therefore, an obvious and urgent need by the owners of reinforced concrete structures for simple, quick, durable and cost-efficient repair techniques. The application of corrosion inhibitors might be such a rehabilitation method since the chloride-contaminated or carbonated concrete does not have to be removed. Thus, this repair method seems to be very promising. Although inhibitors are used in practice [1–2] and some field trials are underway [3] there is still a lack of results from well monitored long-term field studies as well as established practical experience.
18.2
Field study in the Naxbergtunnel
18.2.1 Goals of the study The goal of this three-year field study was to determine and to compare the effectiveness of two inhibitors, sodium monofluorophosphate (MFP) and FerroGard-903 (FG), in the case of chloride-induced rebar corrosion and to verify the results of laboratory experiments [4–5] as well as to evaluate an appropriate monitoring system for such a repair technique. The study was started in 1997 with the condition survey.
18.2.2 Description of the Naxbergtunnel The 550 m-long Naxbergtunnel, built 1972–79, is a part of the highway A2 from Lucerne through the Gotthardtunnel to Italy. It has two normal lines and one emergency line and is located about 1000 m above sea level, thus, in an area where much deicing salt is used during the winter season. The side walls of the tunnel are covered with prefabricated, 2.10 m wide elements 226
Effectiveness of corrosion inhibitors – a field study
227
(panels) with a thickness of only 40 to 50 mm. The reinforcement of the elements consists of one mat of rebars (∅ 4 mm) in the centre.
18.2.3 Condition survey Based on results of a previous condition survey (potential and chloride measurements) of the whole tunnel, 16 elements, situated approximately in the middle part of the tunnel, were chosen for this investigation. A more detailed condition survey of these 16 elements was undertaken in 1997. The following steps were carried out: (a) Potential mapping (measurements grid: 0.15 × 0.15 m), (b) Chloride analysis, (c) Removal of the concrete in small areas (openings) and visual inspection of the corrosion state of the rebars. Figure 18.1 shows the chloride profiles at different heights above the ground (0.45–3.0 m) and the relation between the chloride content and the corrosion potential. At the lower part of the elements (<2.0 m), the chloride content is very high (1.5–2.5% by mass with respect to the mass of cement). At a height of about 2.0 to 3.0 m, the chloride content near the rebars (cover of the rebars: mean value 19 mm) is lower than 0.5% by mass with respect to the mass of cement. Obviously, the corrosion potentials decrease and the intensity of the corrosion process increases with increasing chloride contents.
18.3
Investigation
18.3.1 Test fields and instrumentation The 16 elements used as test fields were as follows: • 4 elements as reference/4 elements treated with MFP/4 elements treated with FerroGard-903, • 2 elements treated with FerroGard-903 and Sikagard-701 W (hydrophobic impregnation), • 2 elements treated with Sikagard-701 W. In the autumn of 1997 these test fields were instrumented with the following monitoring components: • • • •
Electrically isolated rebars, Instrumented cores for resistivity measurements, Sensors for humidity and temperature of the air and concrete, Installation of the data loggers and cabling.
Details of these monitoring components, as well as additional and general information on the monitoring of concrete structures after a repair, are given
228
Corrosion of reinforcement in concrete Autumn 1997 2.5
Chloride content (mass %/c)
0–1 m 2
1–2 m
1.5
2–3 m
1
0.5
0 0
10
20 30 Depth (mm) (a)
Autumn 1997
40
50
–100
0
2.5
Chloride content (mass %/c)
KG3 2 KG4 1.5
1 KG2 0.5
0 –500
–400
–300 –200 Potential [mV (CSE)] (b)
18.1 (a) Chloride profiles at different heights above ground and (b) relation between chloride content and potential and corrosion state of the rebar (KG: grade of corrosion) in openings: KG 0: blank, no corrosion; KG 1: slight signs of corrosion; KG 2: small areas with corrosion; KG 3: corrosion on the whole surface; KG 4: pitting corrosion.
elsewhere [6]. At the same time cores were taken out of the elements to determine the ohmic resistivity of the concrete as a function of the relative humidity under controlled laboratory atmospheres.
18.3.2 Application of the inhibitors The application of the inhibitors was carried out in June 1998, about eight months after the instrumentation of the test fields. This procedure should
Effectiveness of corrosion inhibitors – a field study
229
have allowed the concrete and the new mortar to regain the equilibrium moisture content. Before the application of the inhibitors the surface of the concrete was washed and cleaned with water. The inhibitors were then applied in several steps. This work was carried out by the providers of the inhibitors under the supervision of the project leader. The amount of FerroGard-903 applied during the treatment was more than 500 g per m2 of concrete surface and thus higher than the recommended target value of 300–500 g m–2. The applied amount of MFP was about 2.4 l m–2. The concentrations of the inhibitors were analysed on concrete cores and concrete dust samples. The first analyses were made directly after the application and the second in the autumn of 1999.
18.3.3 Measurements A full set of measurements was executed after the instrumentation of the test fields, then just before and after the application of the inhibitors (June 1998) and thereafter approximately every 6 months. It included the following measurements: • Potential mapping of all elements, • Potential as a function of depth of concrete (potential profiles), • Macrocell current, potential difference, polarisation and ohmic resistance of the isolated rebars, • Macrocell current, potential difference and ohmic resistance of the embedded stainless steel bars, • Ohmic resistance of the instrumented cores (embedded wires), • Temperature and relative humidity. Some of the above mentioned parameters were continuously registered by data loggers.
18.4
Results
The continuous recording of the temperature and the relative humidity (rh) in the tunnel gave mean values of 9 °C and 71%. The temperature varied from –12.3 to 28.5 °C. During rainfall outside of the tunnel, the rh was comparatively high; there were often values between 90 and 100%.
18.4.1 Potential mapping In Fig. 18.2 the results of the statistical analysis of the corrosion potentials measured in the different test fields are shown. The following conclusions can be drawn:
99.9
99
99
95 90 80 70
95 90 80 70
Percentage (%)
Percentage (%)
99.9
99.99 R 97 R 99 R 00
50 30 20 10 5 1
99.99 MFP 97 MFP 99 MFP 00
99.9
50 30 20 10 5 1
.1 0
100
.01 –600 –500 –400 –300 –200 –100 Potential [mV (CSE)] (b)
95 90 80 70 50 30 20 10 5 1
.1
.01 –600 –500 –400 –300 –200 –100 Potential [mV (CSE)] (a)
FG 97 FG 99 FG 00
99
Percentage (%)
99.99
.1 0
100
.01 –600 –500 –400 –300 –200 –100 Potential [mV (CSE)] (c)
0
18.2 Statistical analysis of the corrosion potentials measured 1997, 1999 and 2000 in the reference fields (R), the fields treated with MFP (MFP) and with FerroGard-903 (FG): (a) 1997: Measurements before the application of the inhibitors (September); (b) 1999: approximately one year after the application of the inhibitors (July); (c) 2000: approximately half a year after the elements were placed outside of the tunnel (August).
100
Effectiveness of corrosion inhibitors – a field study
231
• There were only minor changes of the corrosion potentials of the test fields between 1997 and 1999. • In the reference- and MFP-fields a slightly positive shift is apparent in the potential range above –50 mV (CSE). This might be due to a slight reduction in the moisture content of the concrete. • The fields with FerroGard-903 show a slight increase of the potentials in the upper part and slight decrease in the lower part of the curve. This might be caused by the increase of the concrete conductivity of the concrete cover due to the application of the inhibitor (alkaline solutions with salts). • In 2000, the corrosion potentials were generally more negative. The conditions (temperature, humidity) in the elements were different from the conditions during the measurements for 1997 and 1999 in the tunnel. Therefore, they can not be compared directly. • The potential mapping provides reproducible results. The potential profiles of the MFP-elements showed a slight decrease of the potential in the first 10–20 mm. This effect was not recognised in case of FerroGard-903.
18.4.2 Macrocell currents The course of the macrocell currents over time of some electrically isolated rebars are given in the Fig. 18.3 and 18.4. The cleaning of the surface and the 0.12 Cleaning + MFP application Cleaning + FerroGard application
0.1
Macrocell current (mA)
E87, E87, E82, E82,
2.3 0.3 0.5 2.2
mFG mFG mMFP mMFP
June 2./3. 0.08
0.06
0.04
0.02
0 May/30
Jun/2
Jun/4
Jun/6 Jun/9 Date
Jun/11
Jun/13
Jun/16
18.3 Macrocell currents of some electrically isolated rebars during the time of the cleaning process of the surface and the application of the inhibitors. MFP: MFP fields, FG: FerroGard-903 fields.
Corrosion of reinforcement in concrete 0.12 0.10
20
0.08
0
10
0.06 0.04 0.02 0.00 0.12 0.10
Concrete temp. [°C]
Macrocell current (mA)
232
–10 E82 MFP, 0.5 m E82 MFP, 2.23 m Concrete temp. E87 FG, 2.32 m E87 FG, 0.33 m
Treatment
0.08 0.06 0.04 0.02 0.00 Jul/1/1998
Jan/1/1999 Date (a)
Jan/1/2000
R (2.09 m) R (0.97 m) MFP (0.96 m) MFP (1.45 m) FG (0.33 m) FG (1.44 m)
0.05
Macrocell current (mA)
Jul/1/1999
0.04 0.03 0.02 0.01 0.00
–0.01 Jul/1/1998
Jan/1/1999
Jul/1/1999 Date (b)
Jan/1/2000 Jul/1/2000
18.4 Concrete temperature and macrocell currents of some electrically isolated rebars of the differently treated test fields located at different heights above ground [(a) data logger; (b): single measurements]. R: reference field, MFP: MFP fields, FG: FerroGard903 fields.
application of the inhibitors led to a sharp increase of the currents by a factor of 1.5 to 3 (increased conductivity of the concrete). This effect is more pronounced in the lower part of the elements, where higher macrocell currents were measured and probably a larger amount of water reached the surface during the cleaning. The short term transients (peaks) are caused by some single steps of the whole process (cleaning, prewetting, application).
Effectiveness of corrosion inhibitors – a field study
233
The more or less regular variations of the currents within hours or days corresponds to temperature changes. This is more pronounced at higher currents. The highest currents are measured during the summer season. The corrosion process does not stop at temperature below 0 oC. The macrocell currents of the isolated rebars in the lower part of the elements are generally higher than those in the upper part due to the higher chloride content and, thus, more active anodic areas. A clear decrease of the currents after the application of the inhibitor could not be detected.
18.4.3 Concrete resistances Cleaning of the concrete surface with water and the application of the inhibitors (June 1998) led to a decrease of the ohmic resistances (Fig. 18.5 and 18.6), which is more pronounced in the concrete cover than in the middle of the elements (Fig. 18.6). Significant differences between the test fields can not be seen. The variation over time corresponds to the seasonal changes of the temperature (winter/summer). The concrete resistances over the depth (resistance profile) measured with the instrumented concrete cores are shown in Fig. 18.7. The measurements were taken in summer 1998 (after the treatment) and in summer 1999 at similar air temperatures. In most cases, the resistances are higher at the exposed side than in the middle or backside of the elements because of the carbonation of the concrete (increases the resistances). Compared with other structures the profiles are rather flat. The highest resistances are measured in the field with the hydrophobic impregnation, which probably reduced the moisture content of the concrete. Apart from the field with the hydrophobic
Normalized resistance
2.5
R (m) MFP (m) FG (m)
2.0
1.5
1.0
0.5 Jul/1/1998
Jan/1/1999
Jul/1/1999 Date
Jan/1/2000
Jul/1/2000
18.5 Normalized concrete resistances (averages) over time measured with the electrically isolated rebars. R: reference field, MFP: MFP fields, FG: FerroGard-903 fields.
234
Corrosion of reinforcement in concrete
Normalized resistance
2.5
2.0
R MFP FG
5–12.5 mm
1.5
1.0
0.5 Jul/1/1998
Jan/1/1999
Jul/1/1999 Date (a)
Jan/1/2000
Jul/1/2000
Normalized resistance
2.5
2.0
R MFP FG
20–27.5 mm
1.5
1.0
0.5 Jul/1/1998
Jan/1/1999
Jul/1/1999 Date (b)
Jan/1/2000
Jul/1/2000
18.6 Normalized concrete resistances over time measured with the instrumented cores at different depths: (a) 5–12.5 mm; (b) 20–27.5 mm. R: reference field, MFP: MFP fields, FG: FerroGard-903 fields.
treatment, the test fields do not show significant changes of the concrete resistances. The values are similar in the lower and upper parts of the elements. Figure 18.8 shows the correlation between the macrocell current, measured with the electrically isolated rebars, and the concrete resistance, measured with the instrumented cores (depth 12.5–20 mm), in a logarithmic scale. The linear correlation between macrocell current and concrete resistance indicates an ohmic control of the corrosion process (slope ≈ –1). The different intensity of the macrocell currents is mainly related to the various anodic areas on the rebars. The application of the inhibitors did not result in a clear decrease of the corrosion activities.
Effectiveness of corrosion inhibitors – a field study Height over ground >100 cm
2.5 × 104
R MFP FG SG
R MFP FG SG
2 × 104
28.7.99
30.6.98
Temperature approx. 15 °C
1.5 × 104
Road side
Resistance (Ω)
3 × 104
1 × 104
235
5000 0 0
5
10
15
20 25 Depth (mm)
30
35
40
18.7 Concrete resistance over the depth, measured with the instrumented cores. R: reference fields; MFP: MFP fields; FG: FerroGard-903 fields; SG: Sikagard 701-W fields.
Macrocell current (mA)
0.1 E82 2.11 m
7 6 5 4 3 2
E87 2.32 m
0.01 7 6 5 4
MFP before application MFP after application FG before application FG after application
3 2
0.001 3
4
5
6 7 8 Resistance [Ω]
9 ×104
18.8 Correlation of the macrocell current, measured with the electrically isolated rebars, and the concrete resistance, measured with the instrumented cores (depth 12.5 to 20 mm). MFP: MFP fields, FG: FerroGard-903 fields.
18.4.4 Galvanic pulse measurements The galvanic pulse measurements were carried out with equipment which was developed by IBWK, ETH Zürich [7]. Figure 18.9 shows the variation of the polarisation resistances measured over the isolated rebars (in this case connected
236
Corrosion of reinforcement in concrete 800 E78 E80 E82 E85 E86 E87 E92 E93
Polarisation resistance [Ω]
700 600 500 400
R MFP MFP R FG FG FG/SG FG/SG
300 200 100
0 Apr/15/1998
Nov/2/1998
Apr/7/1999
Jul/28/1999
Date
18.9 Polarisation resistance over time of some electrically isolated rebars of the differently treated test fields: R: reference field, MFP: MFP fields, FG: FerroGard-903 fields, SG: Sikagard 701-W fields.
to the rebar mat) in differently treated test fields over time. The higher resistances of E78R, E85R and E93FG/SG were measured at a height of 2.0 to 3.0 m above ground, where the chloride content is lower than 0.5 mass %/c. There is almost no change in the polarisation resistances of the different elements at the end of the field study. No effect of the treatment is recognised. As with the macrocell currents and the ohmic resistance, the polarisation resistance depends strongly on the temperature too. The ohmic resistances, determined from the pulse measurements, showed the same behaviour.
18.4.5 Chemical analysis of the inhibitors Tables 18.1 (FerroGard) and 18.2 (MFP) contain the results of the analysis from 1998 and 1999. The analysis methods of the two promoters and TFB are the same. Therefore, the results can be compared directly. The analysis gave, in both cases, inhibitor concentrations near the rebars which are higher Table 18.1 Inhibitor content, analysed from concrete cores [FerroGard-903: organic component (ppm)] Depth (mm) 0–7 10–17 20–27 30–37 40–47
1998 (Analysis by the promoter) 5088 ± 146 168 ± 19 16 ± 1 < 13 < 13
5780 413 41 16
± 194 ± 28 ±1 ±1 < 13
6709 556 80 26 15
± 342 ± 68 ± 10 ±2 ±2
1999 (Analysis by TFB) 1180 340
760 110
Effectiveness of corrosion inhibitors – a field study
237
Table 18.2 Inhibitor content, analysed from concrete cores and concrete dust (m.%/concrete)] samples [MFP: in PO 2– 4 Depth (mm)
0–10 10–20 20–33 33–44 50–60
1998 (Analysis by the promoter, dust samples) 0.406 0.214 0.237 0.056
0.920 1.093 0.150 0.117
0.229 0.240 0.171 0.090
0.203 0.097 0.055 0.044
1999 (concrete cores) promoter 0.047 0.034 0.055 0.058
TFB 0.250 0.119
0.232 0.081
0.099
0.079
than the target values (MFP: 0.05 mass % in respect to the mass of concrete, FerroGard: approximately 13 ppm organic component). The profiles are steep, especially in the elements treated with FerroGard. The content of FerroGard decreased from 1998 to 1999 (possibly because of evaporation). The mentioned target value near the rebars of 13 ppm corresponds to the limit of the chemical analysis to detect FerroGard. There is no reason why the promoter did not find any MFP in the concrete core in 1999. The mentioned target value near the rebars of 0.05 mass % in respect to the mass of concrete corresponds to the natural ground level of the phosphate content of the concrete.
18.5
Conclusions
The extensively instrumented and monitored field study on the effectiveness of corrosion inhibitors was started in 1997. The side elements (panels) of the walls of the Naxbergtunnel were chosen as test fields. The following conclusions might be drawn: • In both cases were the inhibitor concentrations near the rebars higher than the target values. • No significant effects of the inhibitors MFP and FerroGard-903 on the corrosion of the rebars could be detected either with potential mapping or with the measurements on the electrically isolated rebars or with the instrumented concrete cores. • The hydrophobic impregnation led to an increase in concrete resistance due to the reduced moisture content of the concrete. • Potential measurements are generally possible and give useful results. But they are not sufficient to evaluate the effectiveness of inhibitors. Measurements of the corrosion currents of isolated rebars are a good method to get information about the changes of the corrosion rate. • The instrumentation as well as the monitoring (combination of manual measurements with data logging) has proven to be appropriate.
238
18.6
Corrosion of reinforcement in concrete
Acknowledgements
The authors gratefully acknowledge the Construction Department of the Canton Uri, Altdorf, and the Swiss Federal Highway Administration, Bern, for the financial support of this field study as well as SIKA AG, Zürich, Switzerland and MFP SA, Divonne les Bains, France, for the application of the inhibitors and the chemical analysis.
18.7
References
1. M. Haynes and B. Malric, ‘Use of migratory corrosion inhibitors’, Constr. Repair, July/August 1997, 10–15. 2. E. Brühwiler and P. Plancherel, ‘Instandsetzung von Sichtbetonfasaden mit Inhibitoren’, Schweiz. Ing. Architekt, 1999, 26, 583–586. 3. J. Broomfield, ‘The pros and cons of corrosion inhibitors’, Constr. Repair, July/ August 1997, 16–18. 4. B. Elsener, M. Büchler and H. Böhni, ‘Organic corrosion inhibitors for steel in concrete’, Eur. Fed. Corrosion Publ., 2000, 31, 61–71. 5. M. Salta, E. Pereira and P. Melo, ‘Influence of organic inhibitors on reinforcing steel corrosion’, COST 521, Proc. of a European Workshop and Annual Progress Reports, Belfast, 2000, 247–254. 6. Y. Schiegg, L. Zimmermann, B. Elsener and H. Böhni, ‘Electrochemical techniques for monitoring the conditions of concrete bridge structures’, Int. Conf. Repair of Concrete Structures’, Solvear, May 1997, 213–222. 7. B. Elsener, D. Flückiger, H. Wojtas, H. Böhni, ‘Methoden zur Erfassung der Korrosion von Stahl in Beton’, VSS-Ber., 521, 1996.
18.1
Introduction
The number of reinforced concrete structures showing signs of deterioration or damage due to corrosion of rebars has increased dramatically over the last 20 years. There is, therefore, an obvious and urgent need by the owners of reinforced concrete structures for simple, quick, durable and cost-efficient repair techniques. The application of corrosion inhibitors might be such a rehabilitation method since the chloride-contaminated or carbonated concrete does not have to be removed. Thus, this repair method seems to be very promising. Although inhibitors are used in practice [1–2] and some field trials are underway [3] there is still a lack of results from well monitored long-term field studies as well as established practical experience.
18.2
Field study in the Naxbergtunnel
18.2.1 Goals of the study The goal of this three-year field study was to determine and to compare the effectiveness of two inhibitors, sodium monofluorophosphate (MFP) and FerroGard-903 (FG), in the case of chloride-induced rebar corrosion and to verify the results of laboratory experiments [4–5] as well as to evaluate an appropriate monitoring system for such a repair technique. The study was started in 1997 with the condition survey.
18.2.2 Description of the Naxbergtunnel The 550 m-long Naxbergtunnel, built 1972–79, is a part of the highway A2 from Lucerne through the Gotthardtunnel to Italy. It has two normal lines and one emergency line and is located about 1000 m above sea level, thus, in an area where much deicing salt is used during the winter season. The side walls of the tunnel are covered with prefabricated, 2.10 m wide elements 226
Effectiveness of corrosion inhibitors – a field study
227
(panels) with a thickness of only 40 to 50 mm. The reinforcement of the elements consists of one mat of rebars (∅ 4 mm) in the centre.
18.2.3 Condition survey Based on results of a previous condition survey (potential and chloride measurements) of the whole tunnel, 16 elements, situated approximately in the middle part of the tunnel, were chosen for this investigation. A more detailed condition survey of these 16 elements was undertaken in 1997. The following steps were carried out: (a) Potential mapping (measurements grid: 0.15 × 0.15 m), (b) Chloride analysis, (c) Removal of the concrete in small areas (openings) and visual inspection of the corrosion state of the rebars. Figure 18.1 shows the chloride profiles at different heights above the ground (0.45–3.0 m) and the relation between the chloride content and the corrosion potential. At the lower part of the elements (<2.0 m), the chloride content is very high (1.5–2.5% by mass with respect to the mass of cement). At a height of about 2.0 to 3.0 m, the chloride content near the rebars (cover of the rebars: mean value 19 mm) is lower than 0.5% by mass with respect to the mass of cement. Obviously, the corrosion potentials decrease and the intensity of the corrosion process increases with increasing chloride contents.
18.3
Investigation
18.3.1 Test fields and instrumentation The 16 elements used as test fields were as follows: • 4 elements as reference/4 elements treated with MFP/4 elements treated with FerroGard-903, • 2 elements treated with FerroGard-903 and Sikagard-701 W (hydrophobic impregnation), • 2 elements treated with Sikagard-701 W. In the autumn of 1997 these test fields were instrumented with the following monitoring components: • • • •
Electrically isolated rebars, Instrumented cores for resistivity measurements, Sensors for humidity and temperature of the air and concrete, Installation of the data loggers and cabling.
Details of these monitoring components, as well as additional and general information on the monitoring of concrete structures after a repair, are given
228
Corrosion of reinforcement in concrete Autumn 1997 2.5
Chloride content (mass %/c)
0–1 m 2
1–2 m
1.5
2–3 m
1
0.5
0 0
10
20 30 Depth (mm) (a)
Autumn 1997
40
50
–100
0
2.5
Chloride content (mass %/c)
KG3 2 KG4 1.5
1 KG2 0.5
0 –500
–400
–300 –200 Potential [mV (CSE)] (b)
18.1 (a) Chloride profiles at different heights above ground and (b) relation between chloride content and potential and corrosion state of the rebar (KG: grade of corrosion) in openings: KG 0: blank, no corrosion; KG 1: slight signs of corrosion; KG 2: small areas with corrosion; KG 3: corrosion on the whole surface; KG 4: pitting corrosion.
elsewhere [6]. At the same time cores were taken out of the elements to determine the ohmic resistivity of the concrete as a function of the relative humidity under controlled laboratory atmospheres.
18.3.2 Application of the inhibitors The application of the inhibitors was carried out in June 1998, about eight months after the instrumentation of the test fields. This procedure should
Effectiveness of corrosion inhibitors – a field study
229
have allowed the concrete and the new mortar to regain the equilibrium moisture content. Before the application of the inhibitors the surface of the concrete was washed and cleaned with water. The inhibitors were then applied in several steps. This work was carried out by the providers of the inhibitors under the supervision of the project leader. The amount of FerroGard-903 applied during the treatment was more than 500 g per m2 of concrete surface and thus higher than the recommended target value of 300–500 g m–2. The applied amount of MFP was about 2.4 l m–2. The concentrations of the inhibitors were analysed on concrete cores and concrete dust samples. The first analyses were made directly after the application and the second in the autumn of 1999.
18.3.3 Measurements A full set of measurements was executed after the instrumentation of the test fields, then just before and after the application of the inhibitors (June 1998) and thereafter approximately every 6 months. It included the following measurements: • Potential mapping of all elements, • Potential as a function of depth of concrete (potential profiles), • Macrocell current, potential difference, polarisation and ohmic resistance of the isolated rebars, • Macrocell current, potential difference and ohmic resistance of the embedded stainless steel bars, • Ohmic resistance of the instrumented cores (embedded wires), • Temperature and relative humidity. Some of the above mentioned parameters were continuously registered by data loggers.
18.4
Results
The continuous recording of the temperature and the relative humidity (rh) in the tunnel gave mean values of 9 °C and 71%. The temperature varied from –12.3 to 28.5 °C. During rainfall outside of the tunnel, the rh was comparatively high; there were often values between 90 and 100%.
18.4.1 Potential mapping In Fig. 18.2 the results of the statistical analysis of the corrosion potentials measured in the different test fields are shown. The following conclusions can be drawn:
99.9
99
99
95 90 80 70
95 90 80 70
Percentage (%)
Percentage (%)
99.9
99.99 R 97 R 99 R 00
50 30 20 10 5 1
99.99 MFP 97 MFP 99 MFP 00
99.9
50 30 20 10 5 1
.1 0
100
.01 –600 –500 –400 –300 –200 –100 Potential [mV (CSE)] (b)
95 90 80 70 50 30 20 10 5 1
.1
.01 –600 –500 –400 –300 –200 –100 Potential [mV (CSE)] (a)
FG 97 FG 99 FG 00
99
Percentage (%)
99.99
.1 0
100
.01 –600 –500 –400 –300 –200 –100 Potential [mV (CSE)] (c)
0
18.2 Statistical analysis of the corrosion potentials measured 1997, 1999 and 2000 in the reference fields (R), the fields treated with MFP (MFP) and with FerroGard-903 (FG): (a) 1997: Measurements before the application of the inhibitors (September); (b) 1999: approximately one year after the application of the inhibitors (July); (c) 2000: approximately half a year after the elements were placed outside of the tunnel (August).
100
Effectiveness of corrosion inhibitors – a field study
231
• There were only minor changes of the corrosion potentials of the test fields between 1997 and 1999. • In the reference- and MFP-fields a slightly positive shift is apparent in the potential range above –50 mV (CSE). This might be due to a slight reduction in the moisture content of the concrete. • The fields with FerroGard-903 show a slight increase of the potentials in the upper part and slight decrease in the lower part of the curve. This might be caused by the increase of the concrete conductivity of the concrete cover due to the application of the inhibitor (alkaline solutions with salts). • In 2000, the corrosion potentials were generally more negative. The conditions (temperature, humidity) in the elements were different from the conditions during the measurements for 1997 and 1999 in the tunnel. Therefore, they can not be compared directly. • The potential mapping provides reproducible results. The potential profiles of the MFP-elements showed a slight decrease of the potential in the first 10–20 mm. This effect was not recognised in case of FerroGard-903.
18.4.2 Macrocell currents The course of the macrocell currents over time of some electrically isolated rebars are given in the Fig. 18.3 and 18.4. The cleaning of the surface and the 0.12 Cleaning + MFP application Cleaning + FerroGard application
0.1
Macrocell current (mA)
E87, E87, E82, E82,
2.3 0.3 0.5 2.2
mFG mFG mMFP mMFP
June 2./3. 0.08
0.06
0.04
0.02
0 May/30
Jun/2
Jun/4
Jun/6 Jun/9 Date
Jun/11
Jun/13
Jun/16
18.3 Macrocell currents of some electrically isolated rebars during the time of the cleaning process of the surface and the application of the inhibitors. MFP: MFP fields, FG: FerroGard-903 fields.
Corrosion of reinforcement in concrete 0.12 0.10
20
0.08
0
10
0.06 0.04 0.02 0.00 0.12 0.10
Concrete temp. [°C]
Macrocell current (mA)
232
–10 E82 MFP, 0.5 m E82 MFP, 2.23 m Concrete temp. E87 FG, 2.32 m E87 FG, 0.33 m
Treatment
0.08 0.06 0.04 0.02 0.00 Jul/1/1998
Jan/1/1999 Date (a)
Jan/1/2000
R (2.09 m) R (0.97 m) MFP (0.96 m) MFP (1.45 m) FG (0.33 m) FG (1.44 m)
0.05
Macrocell current (mA)
Jul/1/1999
0.04 0.03 0.02 0.01 0.00
–0.01 Jul/1/1998
Jan/1/1999
Jul/1/1999 Date (b)
Jan/1/2000 Jul/1/2000
18.4 Concrete temperature and macrocell currents of some electrically isolated rebars of the differently treated test fields located at different heights above ground [(a) data logger; (b): single measurements]. R: reference field, MFP: MFP fields, FG: FerroGard903 fields.
application of the inhibitors led to a sharp increase of the currents by a factor of 1.5 to 3 (increased conductivity of the concrete). This effect is more pronounced in the lower part of the elements, where higher macrocell currents were measured and probably a larger amount of water reached the surface during the cleaning. The short term transients (peaks) are caused by some single steps of the whole process (cleaning, prewetting, application).
Effectiveness of corrosion inhibitors – a field study
233
The more or less regular variations of the currents within hours or days corresponds to temperature changes. This is more pronounced at higher currents. The highest currents are measured during the summer season. The corrosion process does not stop at temperature below 0 oC. The macrocell currents of the isolated rebars in the lower part of the elements are generally higher than those in the upper part due to the higher chloride content and, thus, more active anodic areas. A clear decrease of the currents after the application of the inhibitor could not be detected.
18.4.3 Concrete resistances Cleaning of the concrete surface with water and the application of the inhibitors (June 1998) led to a decrease of the ohmic resistances (Fig. 18.5 and 18.6), which is more pronounced in the concrete cover than in the middle of the elements (Fig. 18.6). Significant differences between the test fields can not be seen. The variation over time corresponds to the seasonal changes of the temperature (winter/summer). The concrete resistances over the depth (resistance profile) measured with the instrumented concrete cores are shown in Fig. 18.7. The measurements were taken in summer 1998 (after the treatment) and in summer 1999 at similar air temperatures. In most cases, the resistances are higher at the exposed side than in the middle or backside of the elements because of the carbonation of the concrete (increases the resistances). Compared with other structures the profiles are rather flat. The highest resistances are measured in the field with the hydrophobic impregnation, which probably reduced the moisture content of the concrete. Apart from the field with the hydrophobic
Normalized resistance
2.5
R (m) MFP (m) FG (m)
2.0
1.5
1.0
0.5 Jul/1/1998
Jan/1/1999
Jul/1/1999 Date
Jan/1/2000
Jul/1/2000
18.5 Normalized concrete resistances (averages) over time measured with the electrically isolated rebars. R: reference field, MFP: MFP fields, FG: FerroGard-903 fields.
234
Corrosion of reinforcement in concrete
Normalized resistance
2.5
2.0
R MFP FG
5–12.5 mm
1.5
1.0
0.5 Jul/1/1998
Jan/1/1999
Jul/1/1999 Date (a)
Jan/1/2000
Jul/1/2000
Normalized resistance
2.5
2.0
R MFP FG
20–27.5 mm
1.5
1.0
0.5 Jul/1/1998
Jan/1/1999
Jul/1/1999 Date (b)
Jan/1/2000
Jul/1/2000
18.6 Normalized concrete resistances over time measured with the instrumented cores at different depths: (a) 5–12.5 mm; (b) 20–27.5 mm. R: reference field, MFP: MFP fields, FG: FerroGard-903 fields.
treatment, the test fields do not show significant changes of the concrete resistances. The values are similar in the lower and upper parts of the elements. Figure 18.8 shows the correlation between the macrocell current, measured with the electrically isolated rebars, and the concrete resistance, measured with the instrumented cores (depth 12.5–20 mm), in a logarithmic scale. The linear correlation between macrocell current and concrete resistance indicates an ohmic control of the corrosion process (slope ≈ –1). The different intensity of the macrocell currents is mainly related to the various anodic areas on the rebars. The application of the inhibitors did not result in a clear decrease of the corrosion activities.
Effectiveness of corrosion inhibitors – a field study Height over ground >100 cm
2.5 × 104
R MFP FG SG
R MFP FG SG
2 × 104
28.7.99
30.6.98
Temperature approx. 15 °C
1.5 × 104
Road side
Resistance (Ω)
3 × 104
1 × 104
235
5000 0 0
5
10
15
20 25 Depth (mm)
30
35
40
18.7 Concrete resistance over the depth, measured with the instrumented cores. R: reference fields; MFP: MFP fields; FG: FerroGard-903 fields; SG: Sikagard 701-W fields.
Macrocell current (mA)
0.1 E82 2.11 m
7 6 5 4 3 2
E87 2.32 m
0.01 7 6 5 4
MFP before application MFP after application FG before application FG after application
3 2
0.001 3
4
5
6 7 8 Resistance [Ω]
9 ×104
18.8 Correlation of the macrocell current, measured with the electrically isolated rebars, and the concrete resistance, measured with the instrumented cores (depth 12.5 to 20 mm). MFP: MFP fields, FG: FerroGard-903 fields.
18.4.4 Galvanic pulse measurements The galvanic pulse measurements were carried out with equipment which was developed by IBWK, ETH Zürich [7]. Figure 18.9 shows the variation of the polarisation resistances measured over the isolated rebars (in this case connected
236
Corrosion of reinforcement in concrete 800 E78 E80 E82 E85 E86 E87 E92 E93
Polarisation resistance [Ω]
700 600 500 400
R MFP MFP R FG FG FG/SG FG/SG
300 200 100
0 Apr/15/1998
Nov/2/1998
Apr/7/1999
Jul/28/1999
Date
18.9 Polarisation resistance over time of some electrically isolated rebars of the differently treated test fields: R: reference field, MFP: MFP fields, FG: FerroGard-903 fields, SG: Sikagard 701-W fields.
to the rebar mat) in differently treated test fields over time. The higher resistances of E78R, E85R and E93FG/SG were measured at a height of 2.0 to 3.0 m above ground, where the chloride content is lower than 0.5 mass %/c. There is almost no change in the polarisation resistances of the different elements at the end of the field study. No effect of the treatment is recognised. As with the macrocell currents and the ohmic resistance, the polarisation resistance depends strongly on the temperature too. The ohmic resistances, determined from the pulse measurements, showed the same behaviour.
18.4.5 Chemical analysis of the inhibitors Tables 18.1 (FerroGard) and 18.2 (MFP) contain the results of the analysis from 1998 and 1999. The analysis methods of the two promoters and TFB are the same. Therefore, the results can be compared directly. The analysis gave, in both cases, inhibitor concentrations near the rebars which are higher Table 18.1 Inhibitor content, analysed from concrete cores [FerroGard-903: organic component (ppm)] Depth (mm) 0–7 10–17 20–27 30–37 40–47
1998 (Analysis by the promoter) 5088 ± 146 168 ± 19 16 ± 1 < 13 < 13
5780 413 41 16
± 194 ± 28 ±1 ±1 < 13
6709 556 80 26 15
± 342 ± 68 ± 10 ±2 ±2
1999 (Analysis by TFB) 1180 340
760 110
Effectiveness of corrosion inhibitors – a field study
237
Table 18.2 Inhibitor content, analysed from concrete cores and concrete dust (m.%/concrete)] samples [MFP: in PO 2– 4 Depth (mm)
0–10 10–20 20–33 33–44 50–60
1998 (Analysis by the promoter, dust samples) 0.406 0.214 0.237 0.056
0.920 1.093 0.150 0.117
0.229 0.240 0.171 0.090
0.203 0.097 0.055 0.044
1999 (concrete cores) promoter 0.047 0.034 0.055 0.058
TFB 0.250 0.119
0.232 0.081
0.099
0.079
than the target values (MFP: 0.05 mass % in respect to the mass of concrete, FerroGard: approximately 13 ppm organic component). The profiles are steep, especially in the elements treated with FerroGard. The content of FerroGard decreased from 1998 to 1999 (possibly because of evaporation). The mentioned target value near the rebars of 13 ppm corresponds to the limit of the chemical analysis to detect FerroGard. There is no reason why the promoter did not find any MFP in the concrete core in 1999. The mentioned target value near the rebars of 0.05 mass % in respect to the mass of concrete corresponds to the natural ground level of the phosphate content of the concrete.
18.5
Conclusions
The extensively instrumented and monitored field study on the effectiveness of corrosion inhibitors was started in 1997. The side elements (panels) of the walls of the Naxbergtunnel were chosen as test fields. The following conclusions might be drawn: • In both cases were the inhibitor concentrations near the rebars higher than the target values. • No significant effects of the inhibitors MFP and FerroGard-903 on the corrosion of the rebars could be detected either with potential mapping or with the measurements on the electrically isolated rebars or with the instrumented concrete cores. • The hydrophobic impregnation led to an increase in concrete resistance due to the reduced moisture content of the concrete. • Potential measurements are generally possible and give useful results. But they are not sufficient to evaluate the effectiveness of inhibitors. Measurements of the corrosion currents of isolated rebars are a good method to get information about the changes of the corrosion rate. • The instrumentation as well as the monitoring (combination of manual measurements with data logging) has proven to be appropriate.
238
18.6
Corrosion of reinforcement in concrete
Acknowledgements
The authors gratefully acknowledge the Construction Department of the Canton Uri, Altdorf, and the Swiss Federal Highway Administration, Bern, for the financial support of this field study as well as SIKA AG, Zürich, Switzerland and MFP SA, Divonne les Bains, France, for the application of the inhibitors and the chemical analysis.
18.7
References
1. M. Haynes and B. Malric, ‘Use of migratory corrosion inhibitors’, Constr. Repair, July/August 1997, 10–15. 2. E. Brühwiler and P. Plancherel, ‘Instandsetzung von Sichtbetonfasaden mit Inhibitoren’, Schweiz. Ing. Architekt, 1999, 26, 583–586. 3. J. Broomfield, ‘The pros and cons of corrosion inhibitors’, Constr. Repair, July/ August 1997, 16–18. 4. B. Elsener, M. Büchler and H. Böhni, ‘Organic corrosion inhibitors for steel in concrete’, Eur. Fed. Corrosion Publ., 2000, 31, 61–71. 5. M. Salta, E. Pereira and P. Melo, ‘Influence of organic inhibitors on reinforcing steel corrosion’, COST 521, Proc. of a European Workshop and Annual Progress Reports, Belfast, 2000, 247–254. 6. Y. Schiegg, L. Zimmermann, B. Elsener and H. Böhni, ‘Electrochemical techniques for monitoring the conditions of concrete bridge structures’, Int. Conf. Repair of Concrete Structures’, Solvear, May 1997, 213–222. 7. B. Elsener, D. Flückiger, H. Wojtas, H. Böhni, ‘Methoden zur Erfassung der Korrosion von Stahl in Beton’, VSS-Ber., 521, 1996.