document

Abstract

Predicting long-term reservoir performance with realistic wellbore models is fraught with uncertainty owing to the complexity of two-phase flow. That is because even a calibrated two-phase flow model departs from its expected performance trend when changes in flow conditions occur. These inevitable changes include gas/liquid ratio, wellhead pressure, and flowline pressure with time, among others. Influx of water further exacerbates the prediction problem.

This study explores the possibility of using simplified approaches to computing bottomhole pressure (BHP) from wellhead pressure (WHP), measured rates, gravity of producing fluids, and tubular dimensions. BHP computations on three independent data sets comprising 167 gas-condensate well tests suggest that the no-slip homogeneous model applies quite well. Statistical results show that the homogeneous model compares quite favorably with mechanistic two-phase flow models. However, the main advantage of the simplified model is that its recalibration with field data is not required because the gas/oil ratio increases with time, thereby making the model more robust.

Most field datasets suggest random error in BHP calculations; uncertainty in rate measurements appears to be the most probable cause. High-GLR systems can tolerate large errors in rate measurements, but low-GLR wells demand greater accuracy.

Introduction

Two-phase flow modeling for gas-condensate wells has not received as much attention as that for oil wells. Recent SPE books^1,2 on this topic make very little mention of this flow condition, presumably because modeling is supposed to conform to that offered for oil wells. This study probes this premise, among other issues.

The popular Gray correlation³ appears to do a good job in most gas-condensate wells. However, applicability of this correlation outside the bounds of its specified parameters remains unclear. Take the upper limits of condensate/gas (CGR) ratio of 50 STB/MMscf or flow-string diameters of 3.5 in., for instance. Questions arise whether one should use a different model when one of these criteria, among others, as set by Gray is not met.

Boundaries of applicability often get violated beyond a correlation's original intent; Gray's correlation is no exception in this regard. Practicality demands that a user specifies one computational approach for flow in pipes when long-term integrated reservoir/wellbore/flowline performance is sought over a field's producing life. Declining CGR and increasing water production with time have the potential for complicating any modeling effort. What also remains unclear is how to treat the multicomponent fluid mixture that enters the wellbore/flowline system, after undergoing compositional calculations in the reservoir.

Besides the two-component gas/liquid Gray correlation,³ other approaches have emerged for modeling gas-condensate flow. The semimechanistic model of Govier and Fogarasi⁴ represents the multicomponent approach with flash calculations. In contrast, the wet-gas concept offered by Peffer et al.⁵ suggests extreme pseudoization with single-component gas. Nonetheless, the simplified approach of Peffer et al . with good accuracy is appealing. A minor drawback of both methods is the neglect of accelerational term, which may be significant in wells producing fluids at high gas/liquid ratio (GLR).

This paper advocates the use of two-component homogeneous model to circumvent issues with any rigorous two-phase flow modeling, such as delineating flow-pattern boundaries, estimating slip between phases, and doing flash calculations. We show that Gray's correlation is essentially a homogeneous model, and that of Ansari et al.⁷ also simplify to homogeneous model when mist flow occurs in gas-condensate wells. The steady-state version of the transient simulator OLGA⁸ also lends support to the notion of homogeneous modeling.

Computational Approach and Results

We used data from the literature^4,5 and those from a few West African fields with medium to high-CGR production. In all, we examined 167 independent tests. The methods used involved those of Gray, Aziz et al ., homogeneous, Ansari et al., steady-state OLGA, and the wet-gas approach, advanced by Peffer et al. In two datasets, we used the wet-gas concept but including the accelerational term. In other words, the accelerational term, besides friction and hydrostatic, is implicit in all methods reported in this study.