Modeling the corrosive loss of port infrastructure in the Duluth-Superior Harbor and the North Shore of Lake Superior.

Abstract

Corrosion of steel infrastructure in the Duluth-Superior Harbor (DSH) is a concern for those who own and maintain structures that are critical for maritime transportation. Long-term corrosion rates of steel structures in the DSH were 2 to 4 times faster than is typical for other freshwater habitats, ranging from 0.06 to 0.14 mm/yr. The highest rates of steel corrosion were found at moderately to severely corroded sites and were within the lower range for corrosion of steel in seawater. It is important to know the potential roles that water quality and microorganisms play in the corrosion process to better understand why steel structures in the DSH are corroding faster than expected and to develop methods to prevent or mitigate this problem. Dissolved oxygen increased and chloride and sulfate concentrations decreased in the late 20th century in the DSH after the Western Lake Superior Sanitary District became operational. In 2010, only alkalinity, chloride, and conductivity were positively related to long-term steel corrosion rates in the DSH and interestingly, dissolved oxygen was inversely related to it. The Larson-Skold Index, which measures the corrosivity of water towards steel, decreased or remained constant at three sites between 1972 and 1996. The Larson-Skold Index was usually below the threshold for concern in 2010-2011, indicating that changes in water chemistry alone may not be responsible for the severe corrosion of steel in this harbor. Gallionella, a genus of iron-oxidizing bacteria, was more abundant within corrosion tubercles on severely corroded steel structures in the DSH where more than 109 Gallionella 16S rRNA copies/dry g of tubercle were observed, representing 2% to 34% of the total prokaryotic cells recovered from these tubercles. Gallionella bacteria were less abundant (~107 copies/dry g) on steel structures at sites along the north shore of Lake Superior. DNA from Gallionella was not detected in the water surrounding corroding steel structures in the DSH or in harbors on the north shore of Lake Superior, indicating that these bacteria were enriched on corroding steel. Sulfate-reducing bacteria (SRB) were also more abundant within tubercles at severely corroded sites but at least 2 orders of magnitude less abundant than Gallionella bacteria, accounting for less than 1% of prokaryotic cells within these corrosion tubercles. Considered alone, neither the abundance of Gallionella or SRB was related to long-term steel corrosion rates in the DSH. A multiple linear regression model was developed using water quality parameters and bacterial abundances to predict steel corrosion rates. The model used alkalinity, sulfate concentration, and log10 transformed SRB abundance to predict the long-term steel corrosion rate. It overestimated the long-term corrosion rate of a steel structure in one harbor along the north shore of Lake Superior and slightly underestimated the corrosion rate at a second harbor. It is clear that the corrosion of steel structures is accelerated in the DSH compared to other freshwater environments and water chemistry is not likely the sole cause of this corrosion. Rather, a combination of chemical and microbiological factors appears to influence the corrosion of steel structures in this harbor.