Transient response analysis for fault detection and pipeline wall condition assessment in field water transmission and distribution pipelines and networks.

Abstract

Condition assessment of water distribution pipeline assets has been the focus of water authorities for many years. Transient response analysis, including Inverse Transient Analysis (ITA), provides a new potential method for performing specific nondestructive tests that gives much broader information regarding the condition of pipelines than existing technologies. The basic concept involves inducing a transient in a pipeline and measuring its pressure response. The pressure response is theoretically a function of the condition of the pipeline wall (which is the fundamental characteristic related to the propagation of a transient wavefront) and reflections and damping from any fault that may be present. If an accurate transient model of the pipeline under examination can be developed then it may then be possible to isolate particular parameters in it (relating to the wall thickness of the pipeline or faults such as blockages, air pockets and leaks) and fit these to give optimal matches between the model predicted and measured response of the pipeline. This process is often referred to as inverse analysis (and hence the derivation of the name Inverse Transient Analysis). While a significant amount of numerical and laboratory investigation has been carried out focussing on the use of ITA for leak detection, few field studies have been undertaken. The goal of this research is to determine whether transient response analysis and Inverse Transient Analysis (ITA) can be applied in field situations to provide useful information regarding the condition of pipeline walls and the presence of specific faults such as blockages, air pockets and leaks. Numerous field tests are conducted on large scale transmission pipelines, small scale distribution pipelines and a distribution network in order to obtain a view of the nature of the measured transient responses at each scale and to identify any common characteristics. The capacity of existing transient models to replicate the measured responses is then assessed and they are found to be generally incapable of replicating the field data. Given the physical complexity of field pipelines, and a number of complex phenomena that have been traditionally neglected, this result is not unexpected. The research proposes the development of transient models that can be calibrated to measured responses. These models incorporate mechanisms for including mechanical dispersion and damping and follow precedents developed in other fields of engineering in which damping of transient phenomena is significant.