Computerized Tomography Application in Oil Industry Using KT-2 Signature

Oil industry situations incorporate analysis of the performance of three-phase pumps used in pumping of oil under various conditions. The very popular method of nondestructive evaluation of solid materials, X-ray tomography, has been found suitable also for such flow applications under certain restrictions. Convolution back-projection (CBP) is the most commonly used algorithm that utilizes the projection data (obtained from a CT scanner) to reconstruct the phase distributions in multiphase flows. The practical implementation of this algorithm utilizes the filter functions that are user dependent, and it incorporates the finite Fourier cutoff frequency. It leads to certain errors in reconstruction. Present study concerns with one of the inherent errors arising in CBP implementation leading to characterization of the flow cross-sections. Kanpur Theorem-2 (KT-2) utilizes different Fourier cutoff frequencies (equivalent to different numbers of data-rays) in CBP reconstructions with a fixed filter function. This KT-2 signature provides interesting information about the distribution of different phase fractions in a three-phase bubble column reactor. The variations in these signatures for different flow conditions provide the information about the flow patterns. The performance of three-phase pumps can be characterized by this technique, thus providing useful input for maintenance activities.