PROCESS TOMOGRAPHY FOR MODEL FREE ADAPTIVE CONTROL (MFAC) VIA FLOW REGIME IDENTIFICATION IN MULTIPHASE FLOWS

Abstract

Multiphase flows are frequently found with oil/gas/water/sand in the oil & gas industries, in processes handling dry particulates such as fluidized bed and particulate flow and slurries and sedimentation, to quote a few industrial applications. All these processes have different flow regimes with characteristic distribution of the different materials in the flow. With increasing sensor data from processes and associated possibilities for data fusion, process tomography offers non-intrusive real time sensing methods for identifying these flow regimes, which in certain cases can lead to hazards to personnel and plants in the process industries. In this paper, two scenarios of control using process tomography are presented from the oil and gas industries and powder technology. Twin plane Electrical Capacitance Tomography (ECT) with a plethora of other sensors for measuring pressure, flow rate in two phase flow involving water and air is one application. The other is particulate flow in a fluidized bed. In the case of water and air, ECT was used in a multiphase flow loop with different combinations of air and water mass flow rates enabling the generation of different flow regimes. In the case of particulate flow, different scenarios of flow conditions were generated using particulates and observing the flow regimes based on the distribution and flow rate of the particulates. For the identification of flow regimes, capacitance values from the 12-electrode ECT module at a rate of 100 frames in 200ms, were logged in for data analytics. In the final stages of using the processed data, in the case of two-phase flow 5 outputs consisting of the identified flow regimes, viz. plug, slug, annular, stratified and wavy flows were used as inputs for control and decision making. Similarly, in particulate flow ECT data were used to estimate the air velocity for fluidization as well as in identifying slug and plug flow, which are valuable to the process engineer in determining the range of air velocities for safe operation of the system used for particulate transport, often involving fluidization bed and pipelines. Tomography is used to discern features and tomometry is used in implementing the algorithms.