Past Bank Erosion as a Guide for Bank Erosion Prediction at Pipeline Crossings

Abstract

This paper presents a methodology which uses past bank erosion behaviour as a predictor of future performance. The methodology employed in the bank erosion study consists of the following main steps: identifying a reach to examine, classifying the watercourse, estimating key hydrotechnical properties, obtaining historical air photographs of the reach, georeferencing or orthorectifying the airphotos, mapping the position of the channel edge, obtaining the historical records of nearby gauges to estimate the return period of floods that have occurred between successive pairs of historical air photographs, and finally combining the results to provide correlations between the rates of bank erosion and the rarity of the floods that have occurred. More than 70 bank erosion studies have been completed in the past two years at a variety of watercourses. This paper provides three case histories that illustrate the methodology and then proceeds to provide some tentative relationships that could be used to focus future bank erosion studies on those sites most active, and used to provide a preliminary estimate of the amount of bank erosion that could be expected in both design settings and existing pipeline integrity evaluations. In this study wandering rivers are more laterally active than other channel pattern types. Although the smallest floods do not cause large-scale changes to the banks, significant bank erosion can be caused by either moderate (20-year) or extreme (100-year) events with a rough trend to larger bank erosion in larger floods. No significant correlation between the time elapsed between successive air photos and the magnitude of erosion was found, suggesting that bank erosion is an event-driven process rather than time dependent.