SPE Projects, Facilities & Construction
Volume 1, Number 2, June 2006, pp. 1-7

SPE-95749-PA

Unified Modeling of Gas/Oil/Water-Pipe Flow—Basic Approaches and Preliminary Validation

View full textPDF ( 322 KB )

DOI  More information 10.2118/95749-PA http://dx.doi.org/10.2118/95749-PA

Citation

  • Zhang, H.-Q. and Sarica, C. 2006. Unified Modeling of Gas/Oil/Water-Pipe Flow—Basic Approaches and Preliminary Validation. SPE Proj Fac & Const1 (2): 1-7. SPE-95749-PA.

Discipline Categories

  • 4.5.1 Piping Design
  • 5.1 Design and Optimization
  • 4.5 Pipelines, Flowlines and Risers

Summary

A unified model is proposed for the prediction of flow behavior during production and transportation of gas, oil, and water through wellbores and pipelines. In gas/oil/water three-phase-pipe flow, the phase distributions and hydrodynamics are described on the basis of two criteria: gas/liquid flow pattern and oil/water mixing status. The three-phase flow is treated as gas/liquid two-phase flow if the two liquids are fully mixed or as a three-layer stratified flow at low flow rates in horizontal or slightly inclined pipes. Most three-phase flows fall between these two extremes: partially mixed with slippage between the two liquid phases. Closure relationships describing the distribution between the liquid phases, namely mixing and inversion, are proposed. The model predictions are compared with experimental data of gas/oil/water pipe flows. Significant improvements are observed over the predictions by the two-phase unified model of Zhang et al. (2003a), which assumes a fully mixed liquid phase.

Introduction

Three-phase gas/oil/water flow is a common occurrence in the petroleum industry during the production and transportation of produced fluids.  Three-phase flow behavior, such as liquid holdups and pressure gradient, may be very different from those of two-phase flow. These can have significant impact on design and many flow-assurance issues including hydrate formation, emulsion, wax deposition, and corrosion. Therefore, a reliable and general model needs to be developed for prediction of gas/oil/water-pipe flow behaviors under different flow conditions.

Experimental observations have shown that the flow structures of three-phase-pipe flow are much more complicated than that of two-phase-pipe flow. Açikgöz et al. (1992) classified flow patterns of horizontal three-phase flow into 10 categories. Pan et al. (1995) identified seven flow patterns for horizontal air/oil/water flow. For vertical air/oil/water flow, Woods et al. (1998) identified eight flow patterns. Multiphase-flow hydrodynamic modeling is based on flow-pattern definitions. More flow patterns imply more discontinuities and greater complexity in the hydrodynamic models.  A successful model should unify the predictions of both flow-pattern transitions and hydrodynamic behavior and minimize these discontinuities at the same time.

A unified gas/liquid two-phase flow model has been developed by Zhang et al. (2003a) for predictions of flow-pattern transitions, pressure gradient, liquid holdup, and slug characteristics for all inclination angles from –90° to 90° from horizontal. The model is based on the dynamics of slug flow, which shares transition boundaries with all the other flow patterns. The equations of slug flow are used, not only to calculate the slug characteristics, but also to predict transitions from slug flow to other flow patterns.

Similar methodology also can be used for gas/oil/water three-phase flow. In three-phase pipe flow, the gas-vs.-liquid-phase distribution and structures may be of primary importance compared with the distribution between liquid phases because of the differences among the physical properties of the three phases. Therefore, gas/liquid two-phase flow patterns may be adapted to describe gas/oil/water three-phase flow, and additional closure relationships could be used to describe the distribution between the liquid phases, namely mixing and inversion.

View full textPDF ( 322 KB )

History

  • Original manuscript received: 24 October 2005
  • Revised manuscript received: 30 January 2006
  • Manuscript approved: 3 February 2006
  • Version of record: 20 June 2006