Methodology for Determining Optimum Vibrating Wire Piezometer Pairing Relationships in VWP Monitoring Clusters

Abstract

Current trends in the mining sector and specifically tailings storage facilities have seen a significant increase in monitoring frequency, instrumentation installed on site and field tests conducted. Monitoring methodologies are also shifting away from analog and towards digital electronic systems. These instruments are also being integrated with online dashboards. Owing to all these factors, the instrument that is now most commonly being installed to meet these requirements are VWP’s (Vibrating Wire Piezometers). However interpreting VWP results and deriving the phreatic surface from these are not as straight forward as initially assumed, it requires engineering judgement and a methodology to determine and verify optimum pairing of VWP clusters. Obtaining the correct phreatic surface from VWP’s is critical as this will have a direct impact on the trigger levels and TARP’s of the online dashboard. Inaccuracies in calculating the phreatic surface can lead to the triggering of incorrect levels, which may result in flawed assessments of stability. The primary approach relied on phreatic surface and hydraulic gradients from CPTu testing being compared to the phreatic surface and hydraulic gradient determined from various combinations of VWP’s in a cluster at the time of CPTu testing. In cases where no historical VWP data is available at the time of CPTu testing, a methodology was also investigated using standpipe piezometers only. Piezometric head was converted to pressure and linear regression used to determine the phreatic surface. Results from the primary approach showed that certain pairs of VWP’s yield phreatic surfaces and hydraulic gradients that match the CPTu findings. Standpipe interpretation provided a good starting point and correlates with primary identified pairs. This methodology provides a verification tool to provide confidence when selecting VWP combinations for dashboard reporting.

Full Paper