Bed-mounted electro magnetic meters: Assessment of the (missing) technical parameters

Abstract

Flow measurements in Urban Drainage Systems (UDS) are essential for pollution control and system management. Since the accuracy of, today the most popular, Acoustic Doppler Velocimeters is impeded by several factors, this research is focused on the alternative, or a supplemental, Electro-Magnetic Velocity (EMV) meters. EMV meters are more robust and can provide accurate low flow measurements, even when covered with porous sediment. However, the downside of EMV is the small control volume (CV) where the flow velocities are integrated in a non-linear manner to obtain a single, one-dimensional measured velocity. For a better understanding of the sensor output and measured mean flow velocity with quantified uncertainty, it is necessary to determine the size of the CV and to understand the non-linear integration principle within the CV. Valuable technical parameters, needed for describing these EMV properties, are typically not provided by the manufacturers. Fundamentally, they could be defined with the magnetic field and “virtual” current distributions. To allow for a more practical interpretation of the EMV operating principle, a simplified model of an EMV sensor is proposed here. The suggested model describes the EMV operating principle with only two technical parameters, one-dimensional weighting function w and the reach of the CV, the τmax. Furthermore, a methodology is proposed for defining these two parameters, using two lab flume experiments. The first one is focused on the investigation of the EMV output, when the EMV is covered by the porous sediment with different depths. The second experiment involves the determination of the longitudinal velocity distribution within the lab flume and the CV of the EMV meter. A backward analysis is suggested to formulate a minimization problem, from which the unknown technical parameters are assessed. The proposed procedure was applied on the examined Flat DC-2 EMV meter. Derived one-dimensional weighting function w exponentially drops with the distance from the electrodes, while the reach of the CV was found to be τmax = 8.7 cm. These parameters, and the simplified model, were validated against the EMV outputs acquired in the lab flume, without sediment presence.