Experimental assessment of the effect of sodium chloride deicing and anti-icing solutions on exposed, coated, and reinforcing steel in highway bridges

In North Carolina, winter roadway maintenance strategies vary across the state, with some areas using brine for anti-icing pre-treatment, some using granular salts for deicing, and some using a mix of granular salts with traction enhancing materials, such as sand. An adverse effect of all of these treatment methods is the initiation and progression of corrosion-based deterioration in both reinforced concrete and steel components of these structures. In this study, chloride initiated corrosion rates associated with each treatment approach are quantitatively evaluated for different steel and reinforced concrete bridge components. Corrosion rates on uncoated steel specimens and undercutting of conventional coating strategies under cyclic exposure to solutions of various sodium chloride concentrations are assessed using a modified laboratory procedure. Transport of chlorides to steel superstructure elements below the bridge deck and spatial susceptibility to corrosive loss associated with the different anti-icing and deicing treatments are evaluated through a field study using installation of a large number of sacrificial steel coupons on nine representative bridges in the Ashville and Greensboro regions. Lastly, differences in chloride ingress rates between the winter maintenance strategies within reinforced concrete specimens are investigated through accelerated simulation of North Carolina weather and deicing applications within an environmental chamber. The results of laboratory testing of steel specimens demonstrate quantitative differences between the performance of different steel alloys under cyclic exposure as well as coating methods to blistering and undercutting. Field sampling of corrosive losses on bridges after one winter season suggest that deicing treatments result in more accelerated corrosion of superstructure elements below the bridge deck than anti-icing treatments and provide plausible indications of additional factors that may affect corrosion rates that are supported by statistical regression of the field data. Lastly, controlled laboratory testing with reinforced concrete specimens strongly suggests that anti-icing brine treatments result in higher ingress of chlorides than deicing treatments per application rate, which is attributed to differences in absorption rates driven by the high chloride concentration used in brines.