On the signum function and its effect on acoustic correlation for leak location in buried plastic water pipes

Autor(es)

Almeida, F. C. L. ; Brennan, M. J. ; Paschoalini, A. T. ; Joseph, P. F. ; Gao, Y. ; Vestroni, F ; Romeo, F ; Gattulli, V

• Tipo de documento: Artigo Científico
• Fonte: X INTERNATIONAL CONFERENCE ON STRUCTURAL DYNAMICS (EURODYN 2017); v. 199, p. 6-pg., 2017-01-01.

Resumo

Water loss through leakage in water pipeline distribution systems is a serious issue due to the increasing scarcity of water, such as the recent problem faced by Sao Paulo State in Brazil. Although older metallic pipes are now being replaced by plastic pipes they still suffer from leakage. Unfortunately, correlation techniques, which are used to locate leaks by correlating signals from two vibration sensors attached to the pipe, are less effective in plastic pipes because of higher leak noise attenuation. Hence, the gain setting of the sensors used to collect leak signals have to be carefully selected to enhance the effectiveness of correlation techniques when applied to plastic pipes. However, this is not simple in practical situations, so that the acquired data can become saturated (clipped) or be very small. This paper describes the effects of clipping on the estimation of time delay by severely distorting the signals by using the signum function. It transpires, that although this adds some noise to the original signals, and potentially reduces the bandwidth of which there are measurable leak noise signals, it does not have a profound effect on time delay estimation and hence the accuracy of the leak location. Leak noise signals measured in controlled conditions on a bespoke test-rig constructed by South Staffs Water plc, are used to demonstrate how this process works. (C) 2017 The Authors. Published by Elsevier Ltd. (AU)

• Processo FAPESP: 13/50412-3 - Desenvolvimento de um correlacionador de sinais nacional otimizado para localização e detecção de vazamentos em dutos de água subterrâneos da SABESP, juntamente com dispositivos efetivos para o treinamento de equipes de detecção de vazamento
• Beneficiário: Michael John Brennan
• Modalidade de apoio: Auxílio à Pesquisa - Parceria para Inovação Tecnológica - PITE