Geostatistical analysis to identify characteristics involved in sewer pipes and urban tree interactions

Abstract Tree roots cause significant and costly damage to sewer systems. However, the contributions made by tree and sewer characteristics to sewer system damage are not yet well understood. Previous research has proved inconclusive and there is a lack of agreement on the importance of the variables involved, which means that the results cannot be extrapolated globally. This study presents a methodology to improve the understanding of tree root-sewer conflicts through a geostatistical analysis of a tree inventory and CCTV video fault reports. The methodology is applied to a 11 km 2 case study in Bogota, Colombia. The analysis used tree and sewer pipe georeferenced data, principal component analysis, and linear, Logit, and Poisson regression models. The proposed methodology identifies the pipe characteristics and the tree species most responsible for the root intrusion events into sewers. As expected, larger trees are prone to cause more pipe damage; therefore tree-pipe distance is a relevant parameter control for reducing potential deterioration. Cumulative precipitation and pipe length showed no effect on observed root intrusion events, and pipes made from brick had higher odds to present root intrusion in contrast with concrete pipes. The presented methodology uses readily available information and software, and can be modified depending on the requirements of each particular case study. This approach will therefore allow for more efficient use of costly, site-specific pipe investigations, and is especially useful for cases where there is a general lack of knowledge regarding the characteristics that favour the negative interactions between pipes and tree roots.