Field application of transient analysis methods for pipeline condition assessment
Thesis DisciplineCivil Engineering
Degree GrantorUniversity of Canterbury
Degree NameMaster of Engineering
Reliable water supply is an integral component of any functioning society, improving quality of life through health and sanitation and promoting economic growth. Condition assessment of pipelines is central to the management of water systems and reliable water supply. As demands on water systems increase through population growth, suitable methods of condition assessment are required for strategic planning of ongoing maintenance and renewals of aging pipeline infrastructure. A developing pipeline condition assessment approach is fluid transient response analysis. The approach analyses a systems response to specifically generated transients and identifies characteristic behaviour that can determine the state of that system using direct or inverse analysis procedures. For the analysis procedures to be effective the interaction between transients and pipeline features needs to be well understood. Effective analysis is also dependent upon the ability to identify the response caused by a system on a specifically generated transient signal.
The interaction of fluid transients with extended blockages and changes in pipe wall thickness have been investigated by examining numerical and experimental results in time and frequency domains. These features can be caused through deterioration of pipeline condition and it is found that they can significantly alter the fundamental period of a pipelines transient response, maximum and minimum transient pressure heads and the evolution of the transient signal. The ability of existing models to represent extended pipeline faults under transient conditions is evaluated through direct comparisons of numerical transient responses with new experimental results from the laboratory and through inverse analyses. Direct comparisons show that the transient behaviour is modelled with a good level of accuracy over the first few periods of oscillation and that the level of accuracy decays with time. Inverse analyses show that despite reducing model accuracy with time extended period analysis can provide improved results due to the significant changes an extended fault can have on the periodic response. A periodic Fourier analysis of the data demonstrates that the damping rates of the experimental response are higher than the numerical predictions and that the transient signals exhibits larger damping rates in the higher frequency components of the response.
Research has been focussed on procedures using valve closures as the primary method of transient signal generation, which create single step waves of large magnitude. These valve closure signals can be difficult to identify in live water distribution networks due to hydraulic noise, limiting the application of the technology. To improve field application of transient based condition assessment methods a new signal generation system has been developed. The system, referred to as Pipe SONAR (PCT/EP2015/059540) uses a piezoelectric actuator capable of generating customised, small amplitude pressure signals that vary in frequency and magnitude. The signal generation system is suitable for use on live water distribution networks and has been tested in an international field testing program on 31 sites in New Zealand and China. The field testing program includes tests on steel, cast iron, ductile cast iron, reinforced concrete, asbestos cement, medium/high density polyethylene and polyvinyl chloride pipes with pipe diameters ranging from 100 mm to 600 mm. The Pipe SONAR technology has been assessed through application to two condition assessment techniques; the first is identifying valve condition via a method based on measuring signal transmission through a valve and assesses the ability of the valve to achieve a water tight seal; the second is determining pipeline wave speed for determining pipe wall condition. The technology is found to be practical and functional for application to field use.