The need for supplementary protection of reinforced concrete against the corrosive effects of chloride salts is far more common than that for carbonation- induced corrosion and several approaches have been employed which rely on: (a) surface treatment of the concrete to improve its resistance to chloride penetration, or (b) measures aimed at improving the tolerance of the reinforcement to levels of chloride in excess of normal `threshold’ values. In the former category, a wide range of different surface treatments has been used with a view to controlling chloride ingress and these belong to a number of generic groups, including membranes and barrier coatings (principally film-forming organic polymers and polymer-modified cementitious coatings), pore blocking impregnants (both organic and inorganic materials that penetrate the pores of concrete and block them), and pore lining impregnants (silicones, silanes, siloxanes and similar substances that penetrate into the pores and line them with a hydrophobic surface film to inhibit water ingress). The nature and properties of these materials have been reviewed elsewhere (Concrete Society, 1997; Basheer et al., 1997; Bertolini et al., 2004; Bamforth, 2004) and their performance has been found to vary substantially, both within and between generic types of material. A European standard, addressing issues on testing for effectiveness and durability of surface protection systems, has also been published recently (BS EN 1504-2, 2004) as one of a gradually evolving ten-part series of standards dealing with repair and protection of concrete being prepared under the auspices of CEN Technical Committee TC104. Further information may be obtained from the website of the Concrete Repair Association at www.cra.associationhouse.org.uk.
As regards measures aimed at improving the tolerance of the reinforcement to levels of chloride in excess of normal `threshold’ values, a long list of techniques has been applied with varying degrees of success. This is unsurprising given the many factors that can affect chloride threshold values for plain carbon steel reinforcement. Salient features of the following approaches to improving chloride tolerance will be considered briefly here:
· corrosion resistant alloy steel reinforcement
· coated plain carbon steel reinforcement
· corrosion inhibitive admixtures
· cathodic prevention.
Corrosion resistant alloy steels
The case for judicious use of corrosion resistant alloy steels as reinforcement for concrete exposed to severe chloride contamination has been strengthened in recent years in spite of the relatively high initial cost of the more resistant alloys. This has happened partly as a result of independent long-term studies of the performance of different varieties of corrosion resisting steel, a notable example being the work reported by Treadaway et al. (1989), which demonstrated the excellent durability of certain types of stainless steel over a ten year exposure period in concrete containing high levels of chloride. It is important to note, however, that the term `stainless steels’ is used collectively to describe a large family of over 60 iron alloys with a very wide range of properties. The alloys concerned all contain a minimum of 12% chromium (with carefully controlled levels of several other significant alloying elements such as nickel, molybdenum, carbon, etc.) and their microstructures are categorised as ferritic, austenitic, martensitic or duplex (austenitic-ferritic). Compositional limits and microstructures are usually designated according to BS EN 10088-3 (1995) in Europe (or a similar AISI classification in the USA), typical compositions of the austenitic and duplex types most widely used for reinforcing bars being given in BS 6744 (2001) and in a guidance document entitled `Stainless Reinforcing Steels’ available from the website of UK CARES at www.ukcares.com. Stainless steels owe their corrosion resistance to the formation of stable passive films and their varying susceptibility to localised forms of corrosion, such as pitting, crevice attack and stress-corrosion cracking, is highly sensitive to features of alloy composition, heat treatment and environment.
Reviews of the nature and performance of stainless steels used in concrete have been published recently (European Federation of Corrosion, 1997; NuÈrnberger, 2005) and a feature of significant interest has been the observation that the alloys concerned tend to behave as relatively inefficient cathodes when galvanically coupled to carbon steel in concrete, providing they are free from oxide scale of the kind that may be formed during welding or heat treatment (Sùrensen et al., 1990; Bertolini et al., 2004). This has led to proposals for stainless steel reinforcement to be used selectively in the most vulnerable parts of concrete structures with plain carbon steel reinforcement used in other less severely exposed regions, as galvanic interactions between the two materials are believed to be of less concern than was once feared. It is of course possible to eliminate such galvanic interactions entirely by means of plastic sleeves or paints to prevent contact between the dissimilar metals but this would add to the cost of using stainless steel reinforcement (Bamforth, 2004).
Coated plain carbon steels
Many different metallic and non-metallic coatings have been applied to rein- forcing steel with a view to improving its resistance to chlorides. The most widely used coating of the former type has been zinc, applied by hot dip galvanising, but there has been long-standing controversy over the effectiveness of this technique as a means of achieving required service lives for structures exposed to high levels of chloride contamination (Arup, 1979; Bautista and Gonzalez, 1996). Several factors are known to influence the performance of galvanized coatings in chloride-contaminated concrete, in particular the propor- tion of chloride, the alkalinity of the cement and the metallurgical structure of the coating (Gonzalez and Andrade, 1982; Page et al., 1989). This can result in very high variability in the rates of consumption of galvanised coatings being observed under different circumstances and, while it is clear that galvanised coatings are capable of delaying the onset of significant corrosion of reinforcing steel, it can be difficult to translate this into reliable estimates of the number of additional years of maintenance-free service.
According to some authorities, it appears that only a slight increase in life will be obtained in severe chloride environments (ACI Committee 222, 1996). Others have proposed that a threshold level of 1% chloride by weight of cement may be adopted for design purposes when galvanised steel reinforcement is used (Bamforth, 2004). It has been suggested that the coating itself should not be used as the primary or sole means of corrosion protection, rather it should be used in conjunction with an adequate cover of dense impermeable concrete suited to the type of structure and exposure conditions (Yeomans, 2002). It is also advised that galvanised bars should not be used in metal-to-metal contact with uncoated steel to prevent the formation of galvanic couples (ACI, 1996). If galvanised bars are to be welded, it is to be expected that local loss of coating will occur and it is recommended that the affected area should be cleaned and treated with a suitable zinc-rich paint (Bertolini et al., 2004).
Of numerous non-metallic coatings for steel reinforcement that have been proposed, the only type to be widely used has been fusion bonded epoxy (FBE), applied as powder to cleaned and heated bars to form a layer that is usually ~ 200 um in thickness. FBE coated reinforcement has been used in North America since the mid 1970s in highway bridge decks and other structures exposed to deicing salt and, in the late 1970s, it was also used for the substructure elements of marine bridges in the sub-tropical state of Florida. The coating is intended to isolate the steel from contact with moisture and chloride ions, but the chief difficulty in using FBE coated bars effectively has been preventing damage to the coating during transportation, handling, fixing, placing and compaction (ACI, 1996; Yeomans, 1994). Since extensive corrosion of FBE coated rein- forcement in the substructures of marine bridges in Florida was reported (Sagues et al., 1990; Sagues and Powers, 1997) there has been controversy over the effectiveness of the approach, a brief but informative history of which has been provided by Manning (1995). In consequence, further investigations were initiated in 1993 by the American Federal Highway Administration (FHWA), the various findings of which are summarised in a FHWA report (Virmani and Clemena, 1998).
European experience with FBE reinforcement has been more limited than that in North America and serious doubts have been expressed over the North American practice of using epoxy coated bars in the more severely exposed regions of structures (e.g., the top mat of a bridge deck) coupled to uncoated bars elsewhere. The reason for this concern is that the uncoated bars are expected to provide cathodes of large area in comparison with the anodes formed at breaks in the epoxy coating (Bamforth, 2004).
Corrosion inhibitive admixtures
Corrosion inhibitors are substances which, when added (in small quantities) to an aggressive environment, reduce the corrosion rate of a particular metal in that environment. They are sometimes regarded, in a sense, as `retarding catalysts’ (Fontana, 1986) or as substances that `form a protective coating in situ by reaction of the solution with the corroding surface’ (Jones, 1992). The term, as applied to steel in concrete, will be used throughout this chapter to describe `chemical substances that prevent or retard corrosion by action at the steel/ concrete interface’ (European Federation of Corrosion, 2001), and this implies that adequate (preferably small) concentrations of inhibitor need to be available at the steel/concrete interface in order for corrosion inhibition to be induced. Substances that prevent or retard corrosion of steel in concrete in other ways, e.g. by modifying bulk transport properties of the concrete and thereby limiting the availability of aggressive agents at the steel, are not normally considered as corrosion inhibitors (Bertolini et al., 2004; BuÈchler, 2005).
The classification of a substance as a `corrosion inhibitor’ does not define its mode of action precisely and different inhibitors can work in different ways, some of them retarding the anodic process (anodic inhibitors), others retarding the cathodic process (cathodic inhibitors) and some combining both of these functions (mixed inhibitors). Good examples of corrosion inhibitors (of the anodic type) for steel in aqueous media are alkalis, such as NaOH, KOH and Ca(OH)2. As discussed in Section 5.2.2, these are the substances produced during cement hydration that normally passivate steel, retarding its anodic dissolution and so reducing its corrosion rate to an imperceptible level in an environment such as uncontaminated, dense concrete.
Over the past 50 years, there have been many attempts to enhance the pro- tective character of concrete pore electrolytes by introducing corrosion inhibi- tive admixtures at the time of manufacture of concrete. Clearly, however, any such chemical admixture for concrete has to fulfil a number of criteria in order that adverse effects on workability, cement hydration, development of mechanical properties or other characteristics of the material may be avoided and this effectively rules out several compounds that are often used as corrosion inhibitors for steel in other circumstances.
Of the corrosion inhibitors that have been employed as admixtures in fresh concrete, by far the most widely applied has been calcium nitrite. This has been used since the late 1970s in the USA and has also found applications in other countries for providing supplementary protection against chloride-induced corrosion (Berke and Weil, 1994). It is an anodic inhibitor, which is thought to reinforce the passive film on steel by oxidation of Fe2+ ions produced anodically at defects in the film (Gaidis and Rosenberg, 1979):
In the presence of chloride ions, it has been suggested that the ratio of concentrations of [NO2ÿ]/[Clÿ] must exceed a critical value, ~0.5±1.0, for full passivity to be maintained (Gaidis and Rosenberg, 1987; El-Jazairi and Berke, 1990; Andrade et al., 1986). In circumstances where inadequate dosages of nitrite are used or where its concentration is reduced locally, e.g. by leaching from cracked or permeable concrete, there may be a risk of intensified pitting in the presence of chloride ions. To guard against this, it is important that calcium nitrite-based admixtures should be used only in conjunction with appropriate quality control measures. This is to ensure that the initial concentration of inhibitor is adequate and that other measures aimed at promoting durability are in place, including adequate cover, low water/cement ratio concrete and controlled maximum surface crack widths (Berke and Weil, 1994).
Despite the relatively long record of performance of calcium nitrite-based admixtures, concerns regarding their long-term effectiveness, toxicity and environmental impact have tended to limit their fields of application (Bertolini et al., 2004; BuÈchler, 2005). Efforts to identify alternative corrosion inhibitive admixtures have continued and, during the 1990s, several new types have been introduced, including a number of proprietary formulations based on amines, alkanolamines and their salts with various acids (Maeder, 1994; Elsener et al., 1999) or emulsified mixtures of esters, alcohols and amines (Nmai et al., 1992). Limited information about the detailed compositions of these organic inhibitors has been published and none has yet established a record of long-term per- formance comparable with that of calcium nitrite. A state-of-the-art review on corrosion inhibitors for steel in concrete has been published for the European Federation of Corrosion (2001).
The term `cathodic prevention’ was proposed by Pedeferri to describe the application of the well known technique of cathodic protection as a preventative (rather than as a curative) measure to reinforced or prestressed concrete structures prior to the onset of chloride-induced corrosion (Pedeferri, 1996). Cathodic protection, in its more usual applications as a remedial measure for existing structures that have already suffered corrosion, will be considered in Section 5.9. Whether used on new or old structures, however, the main features of the technique are similar and simply involve applying a direct current between the reinforcing bars (acting as the cathode of a cell) and an external anode of an appropriate type, the magnitude of the current density being adjusted to a level needed to induce passivation rather than pitting. For cathodic prevention, the required current density is generally very small, < 2mA/m2, as it is necessary only to shift the potential by a small amount to achieve a condition of `imperfect passivity’ in which the initiation of pitting is suppressed even though high levels of chloride may eventually penetrate through the concrete cover (see Fig. 5.12).
There are several practical advantages to designing and installing cathodic protection systems for new structures rather than applying them as retrofit measures to existing ones after corrosion has been initiated, although this clearly raises the initial cost and must be evaluated bearing in mind the prospect that new and improved anode systems might become available during the early years of a new structure’s working life. For these and other reasons, the deliberate application of cathodic prevention to new reinforced concrete structures in highly aggressive chloride-laden environments has been limited to date, although it has been used on major structures in several countries (Broomfield, 1997). It is interesting, however, to observe that the reinforcement in many offshore structures is connected to cathodic protection systems that are ostensibly being used to protect the exposed steelwork on the structures, so the reinforcement concerned actually receives unintended cathodic prevention current densities, estimated to be in the range 0.5±1.0 mA/m2 (ACI, 1996).