Fusion-Bonded Epoxy Coatings: a Technology for Rebar Corrosion Prevention

J. Alan Kehr, 3M Corrosion Protection Department, USA

Fikry F. Barouky, Consulting Services Dept., Saudi Aramco, Saudi Arabia

Introduction

Corrosion of steel in concrete has become a costly problem in the United States over the last twenty-five years.[i] Approximately half of the nearly six hundred thousand bridges in the US Federal Aid Highway system have structural deficiencies or are functionally outmoded. According to US Federal Highway Administration (FHWA) estimates, a quarter of US bridge decks are badly deteriorated. Since the beginning of road-salt application, expensive repairs are often required within five to ten years.

It’s a worldwide problem. Research indicates that the service life of buildings in the Arabian Gulf may be five to fifteen years. Reinforced concrete bridges near the seashore in Japan show rapid deterioration within ten years of construction.

The problem is caused primarily by inorganic-salt induced corrosion of steel in concrete. The salt, primarily chloride, penetrates the concrete from sources such as road-deicing salts or sea exposure. It can also be built in through the use of salt-contaminated aggregate, seawater in the concrete, or chloride-based admixtures.

The chloride ion initiates and catalyzes the corrosion reaction. The iron corrosion products resulting from the reaction occupy a much greater volume than iron and cause tremendous pressure on the concrete. The pressure causes the concrete to crack and spall, allowing even greater access of corrodents to the steel and accelerated deterioration of the structure. ²

While most attention is paid to rebar, all steel components are affected as well bridge decks, piers, pilings, and guardrails. ¹ This paper addresses the protection of those elements by describing available alternatives and providing an overview of fusion-bonded epoxy coating materials, case histories, trends in the industry, and specific concerns about the use of fusion-bonded-epoxy coated rebar (FBECR).

Mechanism of Reinforcing Steel Corrosion in Concrete

The traditional view of the reinforced concrete structure is that the concrete is protective to the reinforcing steel bars through the combined effects of the chemical reactions between the steel and the cement hydration products and the environmental barrier provided by the concrete cover. If these conditions are maintained within the concrete mass, the steel bars do not corrode and the structure should have the expected trouble-free life span.

Poor quality reinforced concrete structure contributes to a faster deterioration of the steel reinforcing bars. Low degree of compaction, excess water in the concrete mix, and the hydration process are considered the main factors to create voids within the concrete and make the concrete structure porous.

Porosity of concrete allows penetration and ingress of aggressive elements (e.g., chloride, oxygen, carbon dioxide, and other materials that vary from one location to another) to the embedded steel rebar and to initiate corrosion.

The primary factors controlling the initiation of the steel corrosion and its mechanism in concrete are summarized in the following points:

• The rate of steel depassivation
• The initiation of the macrocells due to the differential aeration and chloride absorption
• The low resistivity attributed by the concrete pore water
• The presence of oxygen to accelerate the corrosion process

The corrosion of steel in concrete is an electrochemical process, which results in the formation of a corrosion cell. The following corrosion mechanism is the most likely for steel rebar embedded in the concrete when significant variations exist in the surface characteristics of the steel. The steel surface initiates cathodes and anodes electrically connected through the body of the steel bar. The “half cell reaction” takes place, by inducing an electromotive force known as standard redox potential when the metal is connected to a hydrogen electrode – see Equation 1.

Equation 1

For iron: Fe --à Fe⁺² + 2 e - (Anode)

The electrons liberated at the anode migrate to the cathode and react in various ways dependant upon the pH value and the availability of oxygen. See Equation 2, Equation 3, and Equation 4.

Equation 2

2e + 2H + ½ O₂ ------à H₂O

Equation 3

2e + H₂O + ½ O2 -----à H₂O

Equation 4

2e + 2H ---------------- à H₂

The anodic and cathodic reactions are autocatalytic and result in the transformation of metallic iron (Fe) to rust. The rust formation is accompanied by a significant increase in the volume, suggested as large as seven times that of the original Fe volume. The volume increase causes concrete cracking and spalling.

Effect of Chloride Ions

When the steel is placed in a highly alkaline solution (pH >11.5), even in the presence of oxygen, corrosion will not be initiated. In fact, slightly rusted bars will be dissipated when placed in strong alkali. That is the reason why, during construction, slightly rusted steel bars do not create a concern.

The chloride ions ingress does not lower the pH in the concrete. However, it destroys the passive layer on the steel bars. The depassivated steel bars do not corrode in the presence of the chloride ions only. The corrosion occurs after the presence of the carbon dioxide lowers the pH below 11, thus contributing to corrosion initiation.

Sources of chloride are either in the concrete mix, mainly from the sand, aggregates, or the water used, or as chloride ingress from the environment, such as in the marine atmospheric environment.

Effect of Carbonation

Carbonation is the alkalinity loss in the concrete mass. The product of the reaction between carbon dioxide in the normal outside air and the alkaline products, mainly the calcium hydroxides, is calcium carbonate. In case of high water/concrete ratio, carbonation continues to the depth where the reinforcing steel bar is embedded.

When carbon dioxide penetrates through the concrete cover in the presence of water in the pores, it drives the pH to lower values which depassivates the steel

Other hydration products in the cement can go through the same reaction with carbon dioxide causing a significant quality loss of the cement and faster deterioration of the concrete mix.

Effect of other Elements

Sulfide can be found in the cement as a contaminant (more than 0.2%). The sulfide ion has been found more destructive to the steel rebar embedded in the concrete if it goes higher than the regulated percentage shown. Regardless of the sulfide ion source, it has been the cause of several cases of hydrogen embrittlement – particularly in pre-stressed rebar.

Mechanism of FBE coated steel corrosion in concrete

In the wake of the premature failure of FBE coated steel rebar in the Florida Keys bridge substructures, many research works by users and academia investigated the performance of FBE coated steel in various environments and service conditions. Most of the laboratory test results confirmed that the FBE material, applied under controlled conditions, passed successfully all qualification and service simulated tests. However, in a few cases, field samples showed poor adhesion to the extent of delamination and disbonding of the FBE rebar. Often times, that delamination was used as the definition for a ‘coating failure,’ rather than corrosion or concrete distress. [ii] In order to understand what went wrong to cause premature failure of the FBE rebar, numerous data were collected from various fields for investigation and assessment. [iii]

Corrosion control of the FBE coating is a function of the coating’s ability to provide a barrier against water, oxygen, chloride, and other aggressive elements [iv] that prevents permeation through the coating film to attack the metal substrate. There are critical properties required for corrosion protection FBE coatings that include adhesion and wetting ability to the rebar.

Reduction in adhesive strength will increase the delamination process rate. [v]^, [vi]. An investigation into delamination of FBE coatings in a simulated pore solution environment suggested the following delamination mechanism: [vii]

·         Delay time before initiation of observable delamination processes may be a function of water penetration through the coating to the interfacial or interphasial coating/substrate region.

·         Delamination of FBE coatings from steel substrates is predominantly caused by hydroxyl ions.

·         Rate of FBE delamination is controlled by transport processes from a pore in the coating and along the delaminated coating/substrate interface to the disbondment front.

·         The locality of failure of coating adhesion is in the interfacial or interphasial coating/substrate region.

·         The rate of FBE delamination in near-passive conditions is controlled by hydroxyl ion migration from the bulk external solution to the coating/substrate disbondment front.

·         The rate of FBE delamination in the condition of underfilm corrosion is controlled by hydrated cation movement to the cathode site.

Design of FBE coating powder for steel rebar coating

New technologies are under continual development to optimize the properties of the FBE coating to improve coating utility. The stoiciometric ratio must be controlled by the equilibrium between the curing group and the epoxy group. For example, increasing the level of curing agent may reduce the cross-link density and increase flexibility, while decreasing chemical resistance.

Impact resistance or hardness is a function of the cross-link density. Higher densities can be achieved using low molecular weight curing agents that show tightly cross-linked structures. Adding non-reactive diluents can interfere with this structure, providing the end product with more flexibility but less toughness.

Mechanical adhesion is the gripping force that results from the roughness of the substrate, (i.e. peaks and valleys). Changing from a round to angular surface profile and increasing the depth of the valleys can improve this type of adhesion. Polar adhesion is the hydrogen bonding which occurs between the substrate and epoxy coating.

Chemical bonds are formed through electron sharing by groups on the substrate and epoxy resin. These bonds are by far the strongest and contribute most to adhesion. Groups such as nitrogen and oxygen can bond with iron and silica.

Corrosion Protection Strategies

There are several corrosion protection strategies available. The five most prominent approaches take advantage of properties of the corrosion cell to reduce corrosion damage to structure steel: surface sealers, concrete barrier, chemical stabilization, electrochemical, and steel coating. ¹ A final alternative is to replace the steel with materials such as stainless steel or fiber-reinforced plastic. [viii]

Surface sealers: Sealers include membranes and materials such as silanes, siloxanes, and methacrylates that function by providing an impervious layer on or in the concrete between external corrodents and the steel. [ix] Corrosion caused by chloride already in the concrete from contaminated aggregate or concrete additives can still occur. Damage to the sealer allows water and chloride penetration. Practical considerations are important –sealers are site applied and subject to the vagaries of weather and construction practices. Surface preparation of the existing concrete can be an important factor. Some materials may remain tacky after application. [x] A key element for successful implementation is for the sealer to prevent the ingress of water but allow the passage of water vapor to prevent blistering and peeling.

Concrete barrier: Increasing concrete depth has proven effective in slowing the ingress of chloride to the steel. The biggest problem with this process, besides increased cost, is an increase in cracking propensity.

In principle, concrete barriers make part or all of the concrete less permeable to water and the associated ions. The typical procedure is to use overlays composed of latex-modified concrete, low-slump concrete, asphalt, or polymer concrete. Problems encountered include increased cracking, increased permeability with age, scaling, and water entrapment. Use of silica fume can significantly reduce permeability and increase electrical resistance of the concrete [xi] but is expensive and increases the risk of higher corrosion rates in the presence of cracks. [xii] Fly ash and natural pozzolans should perform similarly, [xiii]^, [xiv] but the results have varied from increased corrosion rates when used with admixtures to reduced corrosion rates with externally applied chlorides. [xv]

Chemical stabilization: Chemical protection relies on changing the concrete environment to reduce corrosion. Calcium nitrite, the most commonly used inhibitor, does not reduce the permeability of the concrete, nor does it prevent corrosion. Rather, it competes with chloride to react with the steel and reduce the corrosion rate. Two drawbacks are that it acts as a set accelerator for concrete, and normally needs a retarder. Second, the amount required is difficult to predict because exposure varies in different parts of the structure. 4

Organic based inhibitors, such as amine- and ester-based admixtures work by slowing chloride permeation and forming a protective film on the steel surface. ⁷

Electrochemical: Cathodic protection works by imposing an electric potential to oppose the corrosion cell. It requires an anode current distribution system and a power supply. The major drawback is that it is a technically sophisticated, expensive system that requires trained-engineer site visits. It also requires high maintenance expenditures and an external power supply, often in remote areas. ¹ The long-term effects of cathodic protection treatment are not well defined. ⁸

coated rebar: Galvanizing provides protection through a zinc barrier between the steel and the environment and by acting as a sacrificial anode for the steel.[xvi] Zinc does corrode, but the volume of the corrosion products is often less than that of iron products. Therefore, corrosion takes longer to cause cracking and spalling of the concrete. Observations from a seawater exposure evaluation showed clear evidence of progressive corrosion of the zinc layer under natural exposure conditions. [xvii]

Epoxy-coated reinforcement is used extensively in construction to protect steel from corrosion. [xviii] Epoxy coating works by preventing chloride and moisture from reaching the surface of the steel. Its greatest advantage lies in its applicability to existing designs without changes in load capacity or section size, the only change is in the modification of development length.¹ FHWA’s 2003 National Bridge Inventory showed more than 54,000 US bridges contained FBECR in either the top mat or in both the top and bottom mats. [xix] Well over one-hundred-thousand structures utilizing FBECR are now in place with only a handful of problem applications, which will be discussed later.

Epoxy Coated Rebar—Manufacturing Process

The application of fusion-bonded epoxy to reinforcing steel is straightforward and uncomplicated: clean the steel, heat it to the proper temperature, apply the powdered-epoxy coating material, allow the coating to cure, and inspect. However, the details are important and must be understood and implemented to assure a quality coating that will extend the working life of a structure in a corrosive environment. [xx] These same steps apply whether the steel is fabricated before or after coating. However, the equipment configuration and the powder coating gel and application characteristics need to be designed to meet the coating process.

FBECR Performance

There is an overwhelming preponderance of experience that shows FBECR does what it was originally designed to do: reduce the level of corrosion of concrete encased steel to significantly increase the life of the structure. [xxi] Example surveys follow.

Minnesota bridge deck: Bridge number 19015, in Minnesota, has carried a heavy volume of traffic since constructed in 1973, endured extreme seasonal temperatures as well as freeze/thaw cycles, and received several annual doses of deicing salt for twenty years at the time of the study. These are conditions that routinely caused bridge-deck deterioration within ten to twelve years prior to the use of FBECR. Despite the fact that it was constructed using coated bar only in the top mat, and the technology was new at the time of construction (modified pipe-coating application equipment was used), it received a rating of 8, on a scale of one to eight, on its 19th annual inspection—the same rating it received on it’s first. No corrosion-related maintenance or repair work was required during those years. [xxii]

West Virginia bridges: This evaluation surveyed twelve of the earliest bridge decks utilizing epoxy-coated rebar. A number of bridges constructed about the same time with black bar acted as a control. The study included a visual survey, an acoustic chain drag to determine delamination, and chloride testing when weather allowed.

Based on experience with black-bar bridges of the same vintage, the expectation was that there would be a number of spalls, significant delamination of the concrete and cracking extending through the deck. What was found was mathematically non-existent spalling and a uniform absence of delamination. The use of FBECR did not eliminate concrete cracking, but it greatly reduced the corrosion associated with black bars and cracks. Chloride levels ranged from 2.1 to 5.3 pounds per cubic yard (1.2 to 3.2 kg/m³).

The researchers concluded, “ . . . from the data gathered in this investigation that the use of epoxy coated reinforcement does result in dramatic reduction of delamination in bridge decks and by inference an increase in useful life expected of the deck.” [xxiii]

Washington state bridges: In late 1992, the Washington State DOT surveyed four bridges constructed with epoxy coated rebar in the late seventies-early eighties—including the Hood Canal floating concrete pontoon bridge in seawater. Weather conditions permitted chloride sampling of only two of the bridges where the levels were in the ten to twelve pounds per cubic yard (4.5 to 5.5 kg/ m³) range.

The conclusion: “ . . .the system (ECR) is doing a good job of corrosion protection so far in the structures we tested.” [xxiv]

North Carolina—substructures in a marine environment: Visual inspections, coring, and laboratory studies were performed on three bridges constructed in 1985. No extraordinary cracks or deterioration attributable to corrosion was observed. Despite chloride levels ranging from one to over nineteen pounds per cubic yard (0.6 to 11.3 kg/ m³), the coated reinforcement bars were not significantly affected—very slight rusting was detected only in areas with coating damage such as pinholes and holidays. “It is concluded that the epoxy coating in the selected bridges is providing adequate corrosion protection for the reinforcement steel.” [xxv]

*

In spite of the one-hundred-thousand plus structures constructed with FBECR and only a handful of reported problems, the effectiveness of the technology as a long-term corrosion-protection system is currently a subject of debate. Those cases, though small in number, have often, and frequently inaccurately, been publicized.

Part of the concern stems from the early expectations of the technology as a panacea for all corrosion related problems in concrete. [xxvi] Early studies showed that even non-specification-coated rebar would provide a great improvement in reduced corrosion rates. [xxvii] The implication of these studies was that any coated rebar would protect the structure against corrosion and early quality control regimes reflected that misunderstanding by providing little attention to the details important for good-quality coating application to rebar. There was also little concern about handling damage to the coating.

More recent studies and experiments have utilized new expectations, often erroneously referring to a debondment of the coating as an indication of coating failure. [xxviii] However, Sohanghpurwala and Scannell, in a study of core samples representing 3715 bridge decks (almost 4 million m²), found that “although progressive corrosion must be accompanied by complete adhesion loss, coating adhesion alone was not found to be a good predictor of corrosion condition.” [xxix]

Following are examples reporting poor performance of coated rebar.

Florida keys: The often-cited Florida Keys bridges’ substructures provided a wake up call for the FBECR industry. In the late 1980’s, out of five bridges, the substructures of four showed signs of deterioration after only six to ten years. The bridges are located in a subtropical marine environment and “are continuously subjected to salt spray in the splash zone, combined with wetting and drying cycles, high temperatures, and chlorides and moisture, which produce a very corrosive environment.” ¹⁹ Importantly, the discovery of this “problem” removed the halo effect from epoxy-coated rebars: there were circumstances where FBECR, as supplied at the time, did not solve all the problems. This resulted in many studies sponsored by public and private sources and a much greater understanding of the corrosion phenomenon, the importance of quality control in the coating process, and the idea that a next generation of FBE may be required.

Follow-up work and studies of other Florida bridges have highlighted the importance of adhesion and the effects of damage to the coating during the installation process. [xxx]^, [xxxi]

Oregon test pile: In 1980, the Oregon Department of Transportation (DOT) constructed several reinforcement concrete beams and lashed them to an existing bridge substructure in the tidal zone of Yaquina Bay along the Oregon coast. Nine years later, one of the piles was autopsied with a surprise finding: even with chloride levels in excess of twenty pounds per cubic yard (11.9 kg/ m³), two of the longitudinal bars had light corrosion and disbondment, but two of the bars did not show signs of distress and the coating was well adhered. 9 Subsequent testing of the bars showed that the good performing bars had a coating thickness of 9 to 12 mils (225 to 300 microns) compared to 4 to 6 mils (100 to 150 microns) for the distressed bars. [xxxii]

Tests of the bars with thin coating also showed inadequate blast cleaning and evidence that the bars had been salt contaminated prior to coating. This provided further evidence of the importance of understanding and implementing the details for coating application.

Canadian strategic highways research program (C-SHRP): The second phase of this program designed to evaluate the effectiveness of FBECR for long-term (fifty year) performance was completed in 1992. It evaluated bars from coaters and job sites, and investigated field sites. The conclusion was that state-of-the-art coated rebars “will not be effective in providing long-term . . . corrosion protection to reinforcement in salt contaminated concrete.” The study postulated that the failure mechanism involved progressive loss of coating adhesion. ²⁵

To summarize, the problem seen in the Florida Key bridges resulted in two key changes. The first was it raised a concern about all structures utilizing FBECR and caused many surveys of existing structures. The results of those surveys have been very reassuring, FBECR is performing to reduce the corrosion that damages concrete structures. Evaluation of hundreds of structures in many different environments point to that same conclusion. Second, it resulted in many research studies designed to understand the few problems that were unearthed during these surveys. That understanding has caused significant improvements and changes in industry standards, procedures, and expectations.

Industry Trends

Industry standards: Early ASTM standards were written around the procedures developed by the original National Bureau of Standards (NBS, now National Institute of Science and Technology (NIST)) during the evaluation of non-metallic coating materials for reinforcement corrosion prevention. [xxxiii] Very few changes in the specifications, none substantive, were made until observations of the Florida Keys bridges demonstrated a need for review.

As a result of the observations of the Oregon test pile (which showed a direct correlation between corrosion areas and coating thickness and other studies), both AASHTO and ASTM specifications increased the requirement for median coating thickness by about two mils (fifty microns). The allowable holiday (microscopic holes in the coating not visible to the unaided eye, but detectable with an electric probe) count in the application plant was reduced. Good handling practices of FBECR were defined to minimize damage to repair of exposed steel improve the performance of the FBECR. [xxxiv]

Quality: As a result of the findings of the importance of the application process to the performance of FBECR, the industry quickly responded by developing improved quality control processes and standards. One example is the 1992 introduction of the Concrete Reinforcing Steel Institute (CRSI) Voluntary Certification Program. Examples of key measurements under the program include surface profile, bar cleanliness (removal of mill scale and visible contaminants), inorganic salt detection, application temperature, and coating thickness. Adhesion tests such as hot-water immersion and cathodic disbondment were added. Follow up studies showed a significant improvement in the quality of the coated rebar as applicators learned how to meet the certification requirements. [xxxv]

FBECR in both mats: In a nine-year laboratory test sponsored by the FHWA in heavily salt contaminated concrete, slabs with FBECR in both the top and bottom mat with 0.5% intentional damage showed macrocell current density only slightly increased from zero. The results were almost the same as for stainless steel bars. ^{19, [xxxvi]}

Coat after fabrication rebar: The US Office of Naval Research funded a study to evaluate methods of protecting reinforcing steel in waterfront concrete structures. [xxxvii] The 76-month study exposed concrete slabs to a subtropical-marine intertidal environment. The results suggested that in a splash zone environment, rebar should not be bent after coating. A result was the development of FBECR designed for application to reinforcing steel after fabrication. [xxxviii] Coating after fabrication significantly reduces the amount of coating damage sustained during the rebar bending process. This greatly reduced the macrocell current density when top and bottom mats were electrically connected in the FHWA sponsored study. ¹⁹ Finally, since the coating does not have to be as flexible, pipecoating-like technology can be used for the FBE to promote adhesion retention. ^{20, [xxxix]} This technology is gaining popularity in oceanfront and splash-zone concrete construction. [xl]

Specific Concerns

Pitting corrosion: In a severe corrosion environment, where coating damage penetrates to the steel substrate, there will be corrosion. There is no exception based on size or location of the damage. [xli] This is different from pit corrosion, which is an extremely localized attack occurring when steel passivity is destroyed only locally, forming a small anodic area. For uncoated steel, the larger surrounding cathodic areas drive the anodic reaction resulting in a pit. In a sense, the pit cathodically protects the surrounding metal.

In the case of coated rebar, however, the coating restricts the availability of surrounding cathodic areas and restricts the corrosion activity, alleviating its severity. For coated steel, an occasionally expressed concern is about macrocell pit corrosion. ¹⁸

Macrocell driven pit corrosion implies a large cathode area driving a small anode such as in a holding tank with coated sides, but an uncoated bottom. The very nature of damage to the epoxy coating on the rebar makes that scenario unlikely. Even in the case of uncoated bottom mats and coated top mats, there are few, if any, reports of pit corrosion resulting in significant loss of cross section. That is likely due to the relatively limited interconnectivity between bars because of the insulating characteristics of the coating. ¹⁶

Bonding to concrete: In general, uncoated bars provide better bond strength than coated bars. There are three components to bond strength: adhesion, friction, and mechanical bearing of the concrete on the steel deformations. Both adhesion and friction relate to roughness of the steel.(or coating) Because FBECR is smooth and concrete does not adhere well to its surface, bond strength develops primarily through mechanical bearing.

Different studies have given different results, but most give values of 65 to 90% as the relative level of bond strength for epoxy-coated bar compared to black bar.

There are several other factors that significantly affect the bond strength of rebar – coated or uncoated: cover, casting position, concrete slump, and degree of consolidation. There is a nearly linear increase in bond strength with increasing concrete cover. Casting position affects bond strength because increasing the amount of concrete below the bar increases settlement and bleeding which lowers bond strength. Ultimate bond strength decreases with increasing slump. Lack of vibration reduces bond strength.

In summary, while there are several other significant variables, coating on rebar does reduce relative bond strength. That means that increased development length is required for splice and anchorage lengths – there is no requirement for increased cross-sectional area of the steel. [xlii]^{, [xliii], 33}

What does this mean in practical terms? Using a typical bridge design for a three-thousand ft (914 m) bridge, fifty ft (15 m) wide, the added splice length would be approximately twelve hundred ft (366 m) of additional rebar for an additional cost of approximately six-hundred dollars on an eleven million dollar project – an approximate 0.005% increase in cost.

Because of concern raised about the effect of loss of adhesion of epoxy coating to steel on bond strength, [xliv] the FHWA sponsored a study comparing pullout strength among uncoated bars, coated bars, and debonded coated bars. Their findings showed that there were measurable differences, but they were not large enough to constitute a structural safety problem. The conclusion: “A 20 to 30 percent degree of disbondment between the epoxy coating and its steel substrate for bars used as the main flexural reinforcement of a one-way slab does not compromise the slab’s flexural capacity.” [xlv]

Conclusion

The increasing scrutiny following the Florida Keys bridges phenomenon shows continued successful performance of epoxy-coated rebar to protect structures from corrosion induced deterioration. New information generated by those evaluations and the many research studies are changing and improving the coating and construction industry. Significantly improved attention to detail has resulted in far superior coated reinforcement compared to only a few years ago.
References