Corrosion of steel in concrete at various moisture and chloride levels

Abstract: About 7000 concrete bridges in Sweden were built before 1965. The annual cost for maintenance of these bridges is about 0.6% of their total value. The cooling water tunnels at Swedish nuclear power plants have been exposed to seawater for about 30-40 years. The maintenance cost for these tunnels has been evaluated to be about one billion Swedish kronor. It is important to evaluate the risk of corrosion for these types of structures and due to the relatively their old age, corrosion of steel in concrete might be an increasing problem in the future. Generally, corrosion of steel in concrete is induced by either carbonation or by chlorides. Carbonation means that carbon dioxide in air reacts with calcium within the concrete. This means that the pH of the concrete is decreasing and the steel start to corrode. Chloride induced corrosion means that chlorides are transported through the concrete to the steel and the corrosion rate can then increase significantly. It has been proposed in the literature that a certain chloride threshold level exist where an initiation of a significantly high corrosion rate occur. For decades, researchers have been trying to determine this chloride threshold level and the results scatter to a large extent. This is probably due to many different experimental setups and different properties of the used steel and concrete. One important factor to consider is the moisture condition of the concrete. It is generally accepted that the corrosion rate of the steel is low in dry concrete due to high resistivity. For very wet concrete, the corrosion rate is also low due to slow transportation rate of oxygen to the steel surface. At an intermediate moisture condition, the corrosion rate is high due to relatively low resistivity and high transportation rate of oxygen. In the present study, samples made of steel cast in chloride containing mortar were exposed to different moisture conditions. The moisture condition was either static at a certain relative humidity or dynamic where the relative humidity was cycling between 75% and 100%. The lowest chloride concentration which caused initiation of corrosion was 1% Cl by mass of cement and was measured for samples exposed to 97% RH. At higher or lower moisture conditions than 97% RH, the corrosion rate was lower. For samples exposed to dynamic moisture conditions, the lowest chloride concentration which initiates corrosion was measured to be 0.6% Cl by mass of cement. Based on these results it was suggested that the chloride threshold level is lower than 1% Cl by mass of cement for samples exposed to static moisture conditions and even lower chloride concentration can initiate corrosion in dynamic moisture conditions. The present study also assessed the corrosion properties of steel in a cooling water tunnel. The corrosion potential was measured during one year and it was found that the steel in concrete was in an electrical connection with water pumps made of stainless steel and in contact with sacrificial anodes. The connection with the pumps increased the corrosion potential and the connection with the sacrificial anodes decreased the corrosion potential. To assess if there is an increased risk of galvanic corrosion due to the connection with the pumps, a complementary study was made to measure potentials where the corrosion increase significantly. The results suggested that it is not likely that the steel in the tunnel suffer of galvanic corrosion, due to that the measured corrosion potentials in the tunnel was lower than the measured potentials in the complementary study.