Application of soft computing techniques to multiphase flow measurement: A review

https://doi.org/10.1016/j.flowmeasinst.2018.02.017Get rights and content

Highlights

  • This paper reviews soft computing techniques for multiphase flow metering.

  • The sensors used and their working principles are included.

  • The merits and limitations of the soft computing techniques are described.

  • Trends in future developments in the area of research are discussed.

Abstract

After extensive research and development over the past three decades, a range of techniques have been proposed and developed for online continuous measurement of multiphase flow. In recent years, with the rapid development of computer hardware and machine learning, soft computing techniques have been applied in many engineering disciplines, including indirect measurement of multiphase flow. This paper presents a comprehensive review of the soft computing techniques for multiphase flow metering with a particular focus on the measurement of individual phase flowrates and phase fractions. The paper describes the sensors used and the working principle, modelling and example applications of various soft computing techniques in addition to their merits and limitations. Trends and future developments of soft computing techniques in the field of multiphase flow measurement are also discussed.

Introduction

Multiphase flow is defined as a simultaneous flow of materials with two or more different phases (i.e. gas, liquid or solid) or unseparated components (e.g. water and oil) [1]. Multiphase flow (including two-phase flow which is a common example of multiphase flow) is widely seen in many industrial processes. Oil/gas/water mixtures are perhaps the most common gas-liquid and liquid-liquid two-phase or three-phase flows during the processes of production, transportation and custody transfer in the oil and gas industry. Meanwhile, air entrainment is unavoidable when marine fuel is transferred from a bunker barge to a receiving ship, particularly during the start and stop processes of the bunkering. For the fiscal purpose, accurate mass flow metering of marine fuel is essential in bunkering centres. Pneumatically conveyed pulverized fuel (coal, biomass or mixture of both) in power plants forms gas-solid two-phase or three-phase flow. Individual flowrates and fractions of biomass and pulverized coal provide the information of co-firing ratio, which is useful to improve the combustion efficiency and reduce emissions of NOx and COx. In some circumstances hydraulic transport of solids, such as sand, iron concentrates and phosphate matrix, in the type of slurry flow is employed in the mining, chemical, pharmaceutical and food industries. In such industrial processes accurate measurement of multiphase flow is highly desirable to realize flow quantification, operation monitoring, process optimization, and product quality control. It must be noted that multiphase flow is not restricted in industrial processes and covers many other application areas such as regional particle deposition and airflow in human tracheobronchial airways [2], [3] in the medical area. However, this review focuses primarily on the measurement of multiphase flow in the process and related industries.

Individual flowrates (volumetric flowrate or mass flowrate) and phase fractions are most important parameters to characterize a multiphase flow. Over the past three decades substantial progress has been made to develop new techniques that may offer solutions to the industrial measurement challenges. Thorn et al. [4], [5] and Falcone et al. [6] have reviewed the developments of three-phase flowmeters, particularly for the petroleum industry. Possible techniques for the measurement of gas-solid flow in pneumatic conveying pipelines and circulating fluidized beds have been discussed in detail by Yan [7], Zheng and Liu [8], and Sun and Yan [9]. Albion et al. [10] have reviewed the intrusive and non-intrusive measurement techniques for monitoring slurry transportation in horizontal pipelines. Among these techniques, on-line multiphase flowmeters are the devices to measure the mixed flow without any separators and sampling lines. They can be classified into direct and indirect measurement groups according to measurement strategies deployed. The direct measurement of a phase flowrate is often realized using a Venturi flowmeter, Coriolis flowmeter and cross-correlation techniques etc., whilst a phase fraction is usually determined from radiation absorption, electrical impedance and microwave techniques etc. An indirect measurement method determines the individual phases through the analysis of the time variant signals acquired from a set of sensors. In general, the relationship between the sensor outputs and the flowrate or fraction of each phase cannot be deduced theoretically. In this case, empirical models are commonly developed from experimental data using statistical methods. With the recent development of artificial intelligence and machine learning, soft computing techniques provide alternative approaches to traditional statistical methods and extend the capabilities of empirical models.

This review focuses on the indirect methods incorporating soft computing techniques to measure the individual phase flowrates and fractions of multiphase flow. Section 2 outlines the principal constituents of soft computing techniques and provides a brief description of some techniques which are already applied in multiphase flow measurement, i.e. artificial neural network (ANN), support vector machine (SVM), genetic algorithm (GA), genetic programming (GP) and adaptive neuro-fuzzy system (ANFIS). Section 3 presents the example applications of soft computing techniques in two-phase or three-phase flow measurement. Section 4 summarizes the findings of the review and discusses the trends and future developments of soft computing techniques in the field of multiphase flow measurement. Section 5 concludes this review and likely future development.

Section snippets

Soft computing techniques

Soft computing is a collection of methodologies that aim to exploit the tolerance for imprecision and uncertainty to achieve tractability, robustness and low solution cost [11]. It is sometimes referred as computational intelligence, covering a range of computational techniques in computer science, artificial intelligence and machine learning. Sometimes, the term ‘soft computing’ is used interchangeably with soft sensors or virtual sensors. Soft sensor is a common name for a piece of software

Applications of soft computing techniques

The applications of soft computing techniques to multiphase flow measurement are mainly concentrated on the estimation of phase flowrates and phase fractions and the identification of flow regime. The estimation of phase flowrates and phase fractions is equivalent to solve a problem of function approximation while the identification of flow regime is a classification problem. As the review focuses on the measurement of phase flowrates and phase fractions, the research purely on flow regime

Sensor fusion

Traditional sensors incorporating soft computing techniques provide an effective solution to the measurement of phase flowrates and phase fractions. Table 2 summarizes the sensors used, soft computing methods, experimental conditions and measurement errors of the indirect measurement systems discussed in this review. It should be noted that ‘GVF variation’ in Table 2 represents the absolute error of the measurement from the reference value while the rest of the results is the averaged relative

Conclusions

This review has attempted to present the applications of soft computing techniques to multiphase flow measurement and define the state-of-the-art in the development of multiphase flowmeters incorporating such techniques. This review covers the research which have been conducted within the past 15 years and focuses on the measurement of phase flowrates and phase fractions using conventional sensors incorporating soft computing techniques.

Multiphase flow measurement is a complex and difficult

Acknowledgements

The authors would like to acknowledge the National Natural Science Foundation of China (No. 61573140), the Fundamental Research Funds for the Central Universities (No. 2016MS23) and China Postdoctoral Science Foundation (No. 2015M581045) for providing financial support for this research.

References (76)

  • H. Li et al.

    A new void fraction measurement method for gas-liquid two-phase flow in small channels

    Sensors

    (2016)
  • G. Zheng et al.

    Gas-liquid two phase flow measurement method based on combination instrument of turbine flowmeter and conductance sensor

    Int. J. Multiph. Flow

    (2008)
  • M. Meribout et al.

    Integration of impedance measurements with acoustic measurements for accurate two phase flow metering in case of high water-cut

    Flow Meas. Instrum.

    (2010)
  • X. Wang et al.

    Concentration measurement of three-phase flow based on multi-sensor data fusion using adaptive fuzzy inference system

    Flow Meas. Instrum.

    (2014)
  • R. Liu et al.

    A neural network to correct mass flow errors caused by two-phase flow in a digital Coriolis mass flowmeter

    Flow Meas. Instrum.

    (2001)
  • M. Henry et al.

    Two-phase flow metering of heavy oil using a Coriolis mass flow meter: a case study

    Flow Meas. Instrum.

    (2006)
  • M. Henry et al.

    Coriolis mass flow metering for three-phase flow: a case study

    Flow Meas. Instrum.

    (2013)
  • L. Wang et al.

    Input variable selection for data-driven models of Coriolis flowmeters for two-phase flow measurement

    Meas. Sci. Technol.

    (2017)
  • L. Wang et al.

    Mass flow measurement of gas-liquid two-phase CO2 in CCS transportation pipelines using Coriolis flowmeters

    Int. J. Greenh. Gas Control

    (2018)
  • T. Wang et al.

    Coriolis flowmeters: a review of developments over the past 20 years, and an assessment of the state of the art and likely future directions

    Flow Meas. Instrum.

    (2014)
  • S. Lahiri et al.

    Development of an artificial neural network correlation for prediction of hold-up of slurry transport in pipelines

    Chem. Eng. Sci.

    (2008)
  • R. Baker

    Flow Measurement Handbook-industrial Designs, Operating Principles, Performance, and Applications

    (2016)
  • R. Thorn et al.

    Recent developments in three-phase flow measurement

    Meas. Sci. Technol.

    (1997)
  • R. Thorn et al.

    Three-phase flow measurement in the petroleum industry

    Meas. Sci. Technol.

    (2013)
  • G. Falcone et al.

    Multiphase flow metering: current trends and future developments

    J. Pet. Technol.

    (2002)
  • Y. Yan

    Mass flow measurement of bulk solids in pneumatic pipelines

    Meas. Sci. Technol.

    (1996)
  • J. Sun et al.

    Non-instrusive measurement and hydrodynamics characterization of gas-solid fluidized beds: a review

    Meas. Sci. Technol.

    (2016)
  • K. Albion et al.

    Multiphase flow measurement techniques for slurry transport

    Int. J. Chem. React. Eng.

    (2011)
  • L. Zadeh

    Soft computing and fuzzy logic

    IEEE Softw.

    (1994)
  • P. Zarchan et al.

    Fundamentals of Kalman Filtering: A Practical Approach

    (2015)
  • S. Das, A. Kumar, B. Das, A. Burnwal, On soft computing techniques in various areas, in: Proceedings of the National...
  • M. Meireles et al.

    A comprehensive review for industrial applicability of artificial neural networks

    IEEE Trans. Ind. Electron.

    (2003)
  • M. Hgan et al.

    Neural Network Design

    (2011)
  • Kevin Gurney

    An Introduction to Neural Networks

    (2004)
  • C. Cortes et al.

    Support-vector networks

    Mach. Learn.

    (1995)
  • H. Drucker et al.

    Support vector regression machines

    Neural Inf. Process. Syst.

    (1997)
  • M. Jordan et al.

    Support Vector Machines, Information Science and Statistics

    (2008)
  • A. Ben-Hur et al.

    Support vector clustering

    J. Mach. Learn. Res.

    (2001)
  • Cited by (0)

    View full text