Dr Aaron Zecchin

Qualifications

Academic Qualifications

2010 Completed PhD (Civil Engineering) University of Adelaide, Adelaide, South Australia.

2003 Completed BSc (Maths and Computer Science) at the University of Adelaide, South Australia, Australia.

2002 Completed BE (Civil) with first class honours at the University of Adelaide, South Australia, Australia.

Research Interests

Methods for the identification and location of faults within any system or dynamic process are fundamental for reliability, efficiency and safety of operation. The proposed research deals with such methods in their application to general closed conduit fluid line networks (e.g. water distribution systems (WDSs), fluid transmission lines, petroleum lines and oil-filled electrical cables), and is loosely termed anomaly detection within fluid networks. In this instance, an anomaly is defined as a localised (or slightly distributed) physical characteristic that is not a part of the nominal makeup of the fluid line. The main anomalies concerned with this research are (i) leaks, a perforation or opening in the side of the line that causes unwanted release of fluid into the surrounding environment, or (ii) blockages, a localised (or slightly distributed) build up of material (i.e. calcium deposits in WDSs or wax deposits within petroleum lines) within the fluid line that causes reduction in the effective cross sectional area, and hence hydraulic performance, of the line.

Due to the critical nature of leaks (e.g. material loss and potential intrusion points) and, to a lesser extent, blockages, their detection and identification are important. Much research effort has been applied to this problem (almost exclusively to leak detection) over the last 30 years from a variety of fields (e.g. civil, chemical, control, electrical, and petroleum engineering). The methods vary greatly in their approaches, biased by the emphasis in each discipline. There are still many limitations, for example, noise characterization and filtering, modelling the dissipative effect of unsteady friction, dealing with unknown system parameter values (e.g. resistance coefficients and wave speeds), and application of methods to generalised networks. The aim of the proposed research is to develop solutions to some of these key issues.

 Other interests:

• Single dimensional time and Laplace-domain transient models for pipeline fluid flow
• Frequency-domain detection and estimation methods
• Behavioural analysis of population based meta-heuristics and their application to water resource problems

Entry last updated: Wednesday, 20 Oct 2010