LAUSR

Abstract The accurate measurement of multiphase flows is a major challenge in the process industry. In this paper, we present an experimental 8-port microwave tomography system for imaging of multiphase flows in metal pipes, primarily intended for oil–gas–water flows. A special sensor design is proposed which accounts for the requirements of the process industry and allows for broadband measurements in the frequency range from 0.7 GHz to 5.5 GHz. The electromagnetic behaviour of the sensor can be accurately modelled by a 2D model based on the finite element method (FEM) resulting in a moderate computational effort for image reconstruction. The hardware for data acquisition and the algorithm for image reconstruction are reported, focussing on the sensor design and the modelling of sensor's electromagnetic behaviour. The permittivity distributions in case of different static dielectric phantoms modelling oil/water-in-gas and gas/water-in-oil flow distributions were successfully reconstructed at frequencies between 1.25 GHz and 2.5 GHz using the one-step Gauss–Newton method.