Featuring developments in Federal highway policies, programs, and research and technology.
|This magazine is an archived publication and may contain dated technical, contact, and link information.|
|Federal Highway Administration > Publications > Public Roads > Vol. 60· No. 2 > Performance of Epoxy-Coated Rebars in Bridge Decks|
Performance of Epoxy-Coated Rebars in Bridge Decks
by Jeffrey L. Smith and Yash Paul Virmani
The deterioration of reinforced concrete structures is a major problem. The cost of repairing or replacing deteriorated structures -- estimated to be more than $20 billion and to be increasing at $500 million a year -- has become a major liability for highway agencies.(1) The primary cause of this deterioration is the corrosion of steel reinforcing bars due to chlorides. The two main sources of chlorides are deicing chemicals and seawater. During the winter months, many highway agencies use a large amount of salt-based deicing chemicals. The most common of which is sodium chloride. A large number of bridges have also been built in coastal areas and are exposed to seawater.
The corrosion of steel reinforcing bars is an electrochemical process that requires a flow of electric current and several chemical reactions. The rate of corrosion is dependent on the availability of water, oxygen, and chloride ions; the ratio of steel surface area at the anode to that at the cathode; the electrical resistivity of the concrete; and the temperature. The availability of oxygen is a function of its rate of diffusion through the concrete, which is affected by how saturated the concrete is with water. When totally submerged, the diffusion rate is slowed because the oxygen must diffuse through the pore water. When the concrete is dry, the oxygen can freely move through the pores. Alternating wet-dry cycles accelerates the corrosion process. Wet concrete has lower resistivity compared to dry concrete because the water serves as an electrolyte.
Due to the high alkalinity of the concrete pore water, the steel reinforcing bars are protected by an iron oxide film. Chloride ions reach the reinforcing steel by penetrating the concrete via the pore water and through cracks in the concrete. The chloride ions initiate corrosion by depassivating and/or penetrating the iron oxide film and reacting with iron to form a soluble iron-chloride complex.(2) When the iron-chloride complex diffuses away from the bar to an area of greater alkalinity and concentration of oxygen, it reacts with hydroxyl ions to form Fe(OH)2, which frees the complexed chloride ions to continue the corrosion process, if the supply of available water and oxygen is adequate.(3)
The distribution of chlorides in a concrete bridge deck is not uniform. The chlorides typically enter the concrete from the top surface. The top mat of reinforcing steel is then exposed to higher concentrations of chlorides. The chlorides shift the potential of the top mat reinforcing steel to a more negative (anodic) value. Since the potential of the bottom mat has a more positive (cathodic) value, the resulting difference in potentials sets up a galvanic type of corrosion cell called a macrocell. The concrete serves as the electrolyte. Wire ties, metal chair supports, and steel bars serve as metallic conductors. An electric circuit is established. Likewise, the concentration of chlorides is not uniform along the length of the steel bars at the top mat due to the heterogeneity of the concrete and uneven deicer application. These differences in chloride concentrations establish anodes and cathodes on individual steel bars in the top mat and result in the formation of microcells.
The corrosion products resulting from the corrosion of steel reinforcing bars occupy a volume three to six times the volume of the original steel. This increase in volume induces stresses in the concrete that result in cracks, delaminations, and spalls. This accelerates the corrosion process by providing an easy pathway for the water and chlorides to reach the steel.
Several measures have been developed and implemented to prevent the chloride-induced corrosion of steel reinforcing bars and the resulting deterioration. Some of the early measures used included lowering the water-cement ratio of the concrete to reduce permeability and increasing the concrete cover over the steel reinforcing bars. Concrete permeability can also be reduced by the use of admixtures. Corrosion inhibitors are also being used to reduce permeability and protect the passive iron oxide film.
For most corrosion protection measures, the basic principle is to prevent the chloride ions from reacting with the steel surface and, at the same time, increase the time needed for the chloride ions to penetrate through the concrete cover. While these measures generally do not stop corrosion from eventually initiating, they do increase the service life of reinforced concrete structures by slowing the corrosion process.
Epoxy-coated reinforcing steel (ECR) was introduced in the mid 1970s as a means to minimize concrete deterioration caused by corrosion of the reinforcing steel and to extend the useful life of highway structures. The epoxy coating is a barrier system intended to prevent moisture and chlorides from reaching the surface of the reinforcing steel and reacting with the steel. It also serves to electrically insulate the steel to minimize the flow of corrosion current.
In response to reports of poor performance of ECR, most notably the bridges in the Florida Keys, the Federal Highway Administration (FHWA) recommended that states evaluate the performance of ECR in existing bridge decks. As a result, several states initiated investigations and prepared reports documenting their findings and results.
The FHWA report Performance of Epoxy Coated Rebars in Bridge Decks, publication number FHWA-RD-96-092, summarizes the results of those investigations as well as others that have been performed by highway agencies in the United States and Canada, academia, and the Canadian Strategic Highway Research Program (C-SHRP).(4) A total of 92 bridge decks, two bridge barrier walls, and one noise barrier wall located in California, Indiana, Kansas, Michigan, Minnesota, New York, Ohio, Pennsylvania, Virginia, West Virginia, and Wisconsin, and the provinces of Alberta, Nova Scotia, and Ontario were evaluated. At the time of the investigations, the ECR had been in service for three to 20 years.
The investigations typically included field and laboratory evaluation phases. The field evaluation phases consisted of some or all of the following:
The laboratory evaluation phases consisted of some or all of the following:
Summary of Findings and Discussion
The summary of findings and discussion are based on the results of field evaluations of the structures and the laboratory evaluation of concrete cores taken from the various structures as documented in the published and draft reports.(5-18) A brief summary of the investigations, the overall condition of the structures, average concrete cover, average chloride content, and the condition of ECR segments extracted from the cores is contained in Table 1.
Overall Condition of Structures
The bridge decks, bridge barrier walls (parapets), and noise barrier wall were evaluated in the field for cracking, delaminations, and spalls. Overall, the structures were generally found to be in good condition. Concrete deterioration was generally in isolated areas and often not related to corrosion of the ECR.
The extent of deck cracking ranged from very little or none to extensive. Cracking, when present, was generally transverse in nature. Deck cracking was not thought to be a result of any corrosion of ECR. The cracking in the bridge barrier walls consisted of scattered pattern cracking with some vertical cracks. The noise barrier wall consisted of precast concrete panels, and the panels that were closest to the roadway surface were cracked the most and exhibited rust staining and spalling.
Very few spalls or delaminations were found. Delaminations were detected in only 10 of the bridge decks. Approximately half of these delaminations were small, 0.1 square meters (m2) in size. The others varied from 0.3 m2 to approximately 2.8 m2 in size. Several other detected delaminations were associated with expansion devices (uncoated metal) and were not due to any corrosion of ECR.
Depth of Concrete Cover
The depth of concrete cover over the top rebar was measured in each of the cores. Average concrete cover was generally found to be adequate -- at least 51 millimeters. However, some instances of inadequate concrete cover were found. In these instances chloride concentrations were usually higher, and the concrete was typically cracked. As a result, the probability of corrosion occurring is enhanced.
Most investigators determined the total chloride (acid-soluble) content at the rebar level or chloride profiles. The concrete samples were either obtained from the concrete cores or from holes drilled into the concrete. In most cases, the average chloride concentrations at the rebar level were at or above the threshold level to initiate corrosion in black steel. The average total chloride concentration in 44 bridge decks was 2.1 kilograms per cubic meter (kg/m3) with the highest average concentration being 6.8 kg/m3. The average water-soluble chloride concentration in 16 other bridge decks was 0.7 kg/m3 with the highest average concentration being 2.6 kg/m3. These are the averages of all reported total or water-soluble chloride concentrations that may or may not be at the rebar level.
Some of the ECR segments extracted from the cores were examined for visual defects in the coating (holidays), thickness of epoxy coating, and the blast profile on selected bars. Most, if not all, of the segments that were examined contained holidays or bare areas. The thicknesses of the coatings were generally within the limits specified at the time of construction. In most of the instances when the coating thickness did not meet specifications, it exceeded the upper limit. The blast profiles that were evaluated were found to have met applicable specifications.
Rebar Corrosion and Coating Disbondment
ECR segments extracted from the cores were examined to determine the condition of the steel surface under the coating. Out of approximately 202 ECR segments extracted from bridge decks, 162 -- 81 percent -- did not have any corrosion present. For some of the remaining segments that exhibited evidence of corrosion, the corrosion may have been present at the time of construction since chloride contents at the time of the evaluation were below the initiation threshold. Only four ECR segments -- 2 percent -- were reported as having experienced significant corrosion. The areas of corrosion were typically at locations of visible holidays or bare areas. The more heavily corroded ECR segments were also from locations of relatively shallow concrete cover with high chloride concentrations.
Ten ECR segments were extracted from the barrier and noise walls. Out of these segments, eight -- 80 percent -- did not have any corrosion present. Only one ECR segment -- 10 percent -- was reported as having experienced significant corrosion. The areas of corrosion were typically at locations of visible holidays or bare areas. The more heavily corroded ECR segment was also from a location of very shallow, highly permeable concrete with a high chloride concentration.
Some ECR segments extracted from the cores were also examined for any coating disbondment. The extent of coating disbondment varied and was found in both corroded and non-corroded areas. Visible holidays were generally present on ECR segments that experienced coating disbondment. In most cases, the coating was generally still bonded to the steel surface.
California reported coating disbondment on 12 of 32 ECR segments in both corroded and non-corroded areas. Except for one segment, visible holidays were present on all ECR segments that experienced coating disbondment. The extent of coating disbondment varied from 3 to 100 percent of the rebar surface with six segments having coating disbondment of more than 75 percent of their surface. Indiana reported no signs of debonding of the epoxy coating in any of the ECR segments. The coatings were difficult to strip with a knife.
Michigan reported that the epoxy coatings on ECR segments extracted from the experimental decks and with moist concrete were easily removed by hand using a fingernail. Virginia reported the epoxy coatings remained tightly bonded to the steel and could only be removed with a knife.
The results of two separate investigations done in Ontario were reported. In the first investigation, two barrier walls were evaluated. None of the ECR segments in one of the barrier walls showed evidence of any coating disbondment. There was evidence of isolated locations in the second wall with poor bond between the epoxy coating and the ribs of the rebars where the coating could be removed with a knife. However, the coating on the body of the bar could not be removed using a knife after scoring a cross into the coating.
In a second investigation, ECR segments were extracted from 12 bridges in exposed concrete components: barrier walls, end dams, sidewalks, and decks built without waterproofing. The extracted ECR segments were tested for adhesion of the epoxy coating to the steel surface. Of the ECR segments extracted from structures built in 1979 and 1980, 27 percent had a well-adhered coating that could not be lifted from the steel substrate. Of the ECR segments extracted from structures built between 1982 and 1985, 60 percent had a well-adhered coating. Of the ECR segments extracted from structures built in 1990, 88 percent had a well-adhered coating. It appears that adhesion of the epoxy coating decreases with time as ECR segments extracted from bridges with the longest service life exhibited the most adhesion loss.
In the C-SHRP study, the extent of coating disbondment was determined using the dry knife adhesion test. Of the 44 tests performed on ECR segments from the 19 structures, 54 percent had a very well-bonded coating, 14 percent had a coating that was somewhat easy to remove, and 32 percent had a coating that was easy to remove or totally disbonded. The coatings on slightly more than half of the ECR segments still had good adhesion.
The following conclusions are based on the results and findings from the evaluations of the performance of ECR in bridge decks, bridge barrier walls (parapets), and a noise barrier wall.
The overall condition of the bridge decks was considered to be good. Even though deck cracking was prevalent, it did not appear to be corrosion-related. Very few of the decks had any delaminations and/or spalls. Most of the delaminations were not associated with ECR. The maximum extent of delamination reported was less than 1 percent of the deck area. However, the actual extent of delamination was not reported.
A bridge in West Virginia had a total of approximately 3.7 m2 of delaminated area out of a total deck area of 1653.6 m2 -- approximately 0.25 percent of the deck area after 19 years of service life. The largest of these delaminations was centered on a construction joint and was most likely not corrosion-related. Chloride contents are not available for this deck, and the report does not indicate if the delaminations are corrosion-induced. The state of West Virginia indicated in its report that based on previous experience, a typical deck of the same design, but with black steel, would have more delaminations; 5 to 20 percent delamination of the deck area is common. A detailed chloride analysis is also required to determine long-term performance of ECR in aggressive environments.
The chloride concentration at the rebar level for most bridges was at or above the corrosion threshold for black steel. However, the chloride levels in some others were still below the threshold. Most of these decks had not been in service long enough for the chloride levels to reach the threshold level. For these bridges, it may be too soon to determine the effectiveness of ECR.
Corrosion on the extracted ECR segments was determined to be minor in most of the extracted cores. No evidence of corrosion was found on 81 percent of the extracted ECR segments even though chloride concentrations up to 3.8 kg/m3 were well above the chloride threshold level for initiating corrosion in black steel.
ECR did not appear to perform as well when the concrete was cracked as when the concrete was not cracked. There was more corrosion activity on ECR segments extracted from cores taken at locations where the deck was cracked. Even with high chloride concentrations up to 7.6 kg/m3, no visible or negligible corrosion was found on ECR segments extracted from cores taken in uncracked locations. The cracks give both chlorides and moisture easy and direct access to ECR, which appears to accelerate the corrosion process. The lack of cracks appears to hinder the corrosion process.
In California, corrosion on the extracted ECR segments was more severe at locations of heavy cracking and shallow concrete cover -- 15 to 25 mm -- and high chloride concentrations -- 9.7 to 15.0 kg/m3. Moisture/water and a high chloride content present at the rebar level for a considerable length of time are responsible for the observed corrosion.
The Ontario Ministry of Transportation reported that corrosion on the extracted ECR segments was more severe at a location of heavy cracking, shallow concrete cover -- 15 mm -- and a high chloride concentration -- 9.4 kg/m3. This ECR segment was extracted from a noise barrier wall panel that had significant corrosion-induced concrete distress. Moisture/water and a high chloride concentration at the rebar level are once again responsible for the corrosion observed. The concrete in this barrier wall was also very permeable -- 21,293 and 22,722 coulombs. A typical bridge deck does not have such a low concrete cover and/or highly permeable concrete.
Coating disbondment and softening occurred as a result of prolonged exposure to a moist environment. In California, coating disbondment occurred at both corroded and non-corroded areas and was generally detected at visible holidays. In Indiana, ECR segments showed no signs of coating disbondment. In Michigan, coatings on ECR segments extracted from moist concrete could easily be removed. In New York, coating deterioration was not found on any of the ECR segments. Tests performed in Ontario showed that adhesion of the epoxy coating to the steel substrate decreases with time. Approximately 54 percent of the ECR segments evaluated under the C-SHRP program still had good adhesion of the epoxy coating.
The number of defects in the epoxy coating and the amount of disbondment influence the performance of ECR. Many of the extracted ECR segments contained defects: holidays, bare areas, mashed areas, or a combination of one or more of these. In California, high chloride concentrations up to 4.6 kg/m3 did not initiate corrosion when there were no defects (holidays) in the coating, indicating that non-damaged epoxy coatings provide an adequate barrier to chlorides. In Virginia, there were no indications of significant corrosion even though the initial condition of the coating was poor and numerous holidays and bare areas were present.
The use of ECR has reduced, if not completely eliminated, the deterioration of deck concrete resulting from corrosion of reinforcing steel. West Virginia reported that relatively high chloride concentrations -- up to 3.1 kg/m3 -- had not resulted in any significant concrete deterioration due to any possible corrosion of ECR.
A comparison of the performance of ECR in decks with only the top mat of reinforcing steel epoxy-coated and decks with both the top and bottom mat of reinforcing steel epoxy-coated suggests superior performance when both mats are epoxy-coated.
The bridges evaluated in California were originally constructed with black steel. Based on the dates of original construction and first redecking, it appears that the use of black steel only provided 10 to 12 years of service life. However, it is possible that there were other contributing factors shallow cover and a lower quality of concrete besides the use of black steel.
The use of an adequate, good quality, concrete cover; adequate inspection; finishing and curing of the concrete; and the proper manufacturing and handling of ECR complement the use of ECR in providing effective corrosion protection for concrete bridge decks.
ECR has provided effective corrosion protection for three to 20 years of service. Corrosion was not a significant problem in any of the decks evaluated. No signs of distress were found in the first bridge decks built with ECR. There was no evidence of any significant premature, concrete deterioration that could be attributed to corrosion of ECR. Some of the cores were intentionally taken at locations representing a worst case. Therefore, these cores may not be representative or indicative of the overall performance that can be obtained from ECR. Little or no maintenance or repair work has been done on most of the decks.
Table 1 -- ECR Bridge Deck Survey Summary
1This column represents the observed corrosion on the extracted ECR. A report of 12 percent corrosion means that 12 out of 100 ECR segments had some corrosion, which may be due to chlorides or may have been present at the time of construction as some ECR segments showed corrosion where the chloride level was lower than the threshold for black steel.
2This includes the disbondment caused by holidays in the coating on the ECR segments. A report of 24 percent disbondment means that 24 out of 100 segments could be disbonded with a knife. In reality it means that the epoxy coating on 24 out of 100 segments had deteriorated over 8.8 years.
NR -- Not reported, ND -- Not determined.
3Metric conversions: 1 in = 25.4 mm, 1 lb/yd3 = 0.6 kg/m3, 1 ft2 = 0.0929 m2.
1. E.J. Gannon and P.D. Cady. Condition Evaluation of Concrete Bridges Relative to Reinforcement Corrosion, Volume 1: State of the Art of Existing Methods, Publication No. SHRP-S/FR-92-103, Strategic Highway Research Program, Washington D.C.
2. J. Fraczek. "A Review of Electrochemical Principles as Applied to Corrosion of Steel in a Concrete or Grout Environment," Corrosion, Concrete, and Chlorides -- Steel Corrosion in Concrete: Causes and Restraints, SP-102, American Concrete Institute, Detroit, Mich.
3. J.G. Dillard, J.D. Glanville, W.D. Collins, R.E. Weyers, and I.L. Al-Qadi. Concrete Bridge Protection and Rehabilitation: Chemical and Physical Techniques, Feasibility Studies of New Rehabilitation Techniques, Publication No. SHRP-S-665, Strategic Highway Research Program, Washington D.C.
4. J.L. Smith and Y.P. Virmani. Performance of Epoxy Coated Rebars in Bridge Decks, Publication No. FHWA-RD-96-092, Federal Highway Administration, Washington, D.C., August 1996.
5. R.A. Reis. In Service Performance of Epoxy Coated Reinforcement In Bridge Decks, Draft Report No. FHWA/CA/TL-96/01-MINOR, California Department of Transportation, Sacramento, Calif.
6. H.O. Hasan and J.A. Rameriz. Behavior of Concrete Bridge Decks and Slabs Reinforced with Epoxy-Coated Steel, Publication No. FHWA/IN/JHRP-94/9, Purdue University, West Lafayette, Ind.
7. R.L. McReynolds. KDOT Bridges with ECR and ECR Deck Condition Ratings, Kansas Department of Transportation, Topeka, Kan.
8. B.J. Wojakowski. "Epoxy Coated Reinforcing Steel in Kansas," departmental memorandum, Kansas Department of Transportation, Topeka, Kan.
9. R.L. McCrum, B.R. Lower, and C.J. Arnold. A Comparison of The Corrosion Performance of Uncoated, and Epoxy Coated Reinforcing Steel in Concrete Bridge Decks, Draft Research Report No. R-1321, Michigan Department of Transportation, Lansing, Mich.
10. H.J. Gillis and M.G. Hagen. Field Examination of Epoxy Coated Rebars In Concrete Bridge Decks, Minnesota Department of Transportation, St. Paul, Minn.
11. G.R. Perregaux and D.R. Brewster. In-Service Performance of Epoxy-Coated Steel Reinforcement in Bridge Decks, Technical Report 92-3, New York Department of Transportation, Albany, N.Y.
12. R. Weyers. Evaluation of Epoxy Coated Reinforcing Steel in Eight Year Old Bridge Decks, Lafayette College, Easton, Pa.
13. G. Malasheskie, D. Maurer, D. Mellott, and J. Arellano. Bridge Deck Protective Systems, Publication No. FHWA-PA-88-001+85-17, Pennsylvania Department of Transportation, Harrisburg, Pa.
14. W.T. McKeel. Evaluation of Epoxy-Coated Reinforcing Steel, Publication No. FHWA/VA-94-R5, Virginia Transportation Research Council, Charlottesville, Va.
15. D. Lipscomb. Evaluation of Bridge Decks Using Epoxy Coated Reinforcement, MIR #1261603, West Virginia Department of Transportation, Charleston, W.Va.
16. D.G. Manning. Field Investigation of Epoxy-Coated Steel Reinforcement, MAT-89-02, Ontario Ministry of Transportation, Ottawa, Ontario.
17. I. Ip. "Performance of Epoxy-Coated Reinforcement in Bridge Decks -- Ontario Experience -- Draft Summary Report," personal correspondence.
18. K.C. Clear. Effectiveness of Epoxy Coated Reinforcing Steel -- Final Report, Canadian Strategic Highway Research Program, Ottawa, Ontario.
Jeffrey L. Smith is the bridge engineer for FHWA's Kansas Division in Topeka, Kan. He received his bachelor's degree in civil engineering from the University of Toledo and his master's degree in civil engineering from the University of Kansas. He is a licensed professional engineer in Ohio and Virginia.
Yash Paul Virmani is a research chemist in the Structures Division of FHWA's Office of Engineering Research and Development at the Turner-Fairbank Highway Research Center in McLean, Va. Dr. Virmani is the program manager for the high-priority research area "Reinforced and Prestressed Concrete and Cable Stays." He is the co-inventor of conductive polymer concrete, a material that is the basis of several cathodic protection systems.
Page Owner: Office of Corporate Research, Technology, and Innovation Management
Scheduled Update: Archive - No Update
Technical Issues: TFHRC.WebMaster@dot.gov