Abstract
Previous studies have yielded promising results as to the use of impedance measurements in two phase flow metering in estimating the flow rate of individual flows in oil fields (i.e. water and oil flows). However, most of these approaches are not accurate in the case of high water-cut or within the water-cut range 40% to 60%, where the mixed liquid is neither a good conductor, nor a good insulator. With the increased out-aging of the oil fields, these two ranges are usually dominant which require the usage of new methods. This paper presents a new non-radioactive probe which combines acoustic measurements with the impedance measurements to accurately provide in real-time the flow rate of each individual flow within the whole water-cut range. A dedicated pattern recognition algorithm using a neural network and relying on the physical properties of the fluid dynamic has been implemented. Experimental measurements performed on a laboratory scale two phase flow loop show that the proposed flow meter can in fact become a competitive two phase flow meter, since less than 5% relative error could be achieved for the whole water-cut range.
Original language | English |
---|---|
Pages (from-to) | 8-19 |
Number of pages | 12 |
Journal | Flow Measurement and Instrumentation |
Volume | 21 |
Issue number | 1 |
DOIs | |
Publication status | Published - Mar 2010 |
Keywords
- Acoustic arrays
- Acoustic signal processing
- Artificial intelligence
- Computer architecture
- Neural network
ASJC Scopus subject areas
- Modelling and Simulation
- Instrumentation
- Computer Science Applications
- Electrical and Electronic Engineering