Abstract

Underground infrastructure serves an essential need for the society. Huge number of facilities is dedicated to facilitate the well-being’s needs. Sewer infrastructure, one of the facilities, plays a major role in maintaining healthier environment. Its main duty is to transfer sewage material to treatment plants or any designated disposal area. Therefore, providing well performing sewer systems is essential to avoid any breakdown. Nevertheless, sewer pipelines’ condition in North America is deteriorating. In fact, studies have shown that 30% of municipal infrastructure in Canada is in either fair or very poor condition. As a result, there is a significant requirement for inspection and rehabilitation. Many municipalities utilize Closed Circuit Television (CCTV) inspection technique in inspecting sewer pipelines. However, this technique suffers from significant subjective and imprecise conclusions. Hence, studying, analyzing and applying different sewer inspection technologies and designing a condition assessment model are necessary to reduce subjectivity and errors and produce accurate and reliable results.
This research aims to develop an automated tool to quantify: deformation, settled deposits, infiltration and surface damage sewer defects. The automated approach is dependent upon using image processing techniques and several models to analyze output data from 2D laser profiler, sonar and electroscan. Other than using ASTM F1216 formula, the research suggests applying the roundness factor in quantifying the deformation defect.
The research develops a condition assessment model, based on the aforementioned defects, to arrive to an aggregated index suggesting the condition of sewer pipelines. Multi Attribute Utility Theory (MAUT) approach is used for each defect. The research also suggests a methodology to evaluate the surface damage defect of sewer pipelines for reinforced concrete, vitrified clay and ductile iron sewer pipeline materials. An interface, using MATLAB, was developed to implement the designed quantification algorithms and the MAUT model on real case studies.
After implementing and validating the two deformation quantification methods, the Mean Absolute Error (MAE) utilizing the ASTM F1216 was 4.27%, while the MAE using the roundness factor was 4.83%. The maximum difference percentage was found to be 40.06%; however, the minimum difference percentage was 0.59%. The average difference percentage for all the cases was calculated as 16.67%. Later, the MAUT model was validated with actual case studies. Three rounding types (rounding to nearest number, rounding up and down) were tested to change the aggregated index, containing decimals, to a whole number. Mean Absolute Error (MAE) was utilized to compare the rounding types. In all case studies, rounding up type produced the lowest MAE values. When rounding up the computed index in case study 1, the MAE for Concordia Sewer Protocol (CSP), Water Research Centre (WRc) and New Zealand were 0.33, 0.33 and 0.42, respectively.
This research shall encourage subject matters to utilize technologies, other than or beside CCTV, to conclude sound results. The developed automated user interface shall reduce inaccuracy and subjectivity through the application of robust image processing algorithms. After extending this research in including several sewer’s components and defects, the condition assessment model shall aid asset managers to allocate their maintenance and rehabilitation budgets.