Summary
Measurement of gas and condensate relative permeabilities typically is performed through steady-state linear coreflood experiments using model fluids. This study addresses experimental measurement of relative permeabilities for a rich-gas/condensate reservoir using a live, single-phase reservoir fluid. Using a live, single-phase reservoir fluid eliminates the difficulties in designing a relatively simple model fluid that replicates the complicated thermodynamic and transport properties of a near-critical fluid. Two-phase-flow tests were performed across a range of pressures and flow rates to simulate reservoir conditions from initial production through depletion. A single-phase multirate experiment was also performed to assess inertial, or non-Darcy, effects. Correlations were developed to represent both the gas and condensate relative permeabilities as a function of capillary number. A nearly 20-fold increase in gas relative permeability was observed from the low- to high-capillary-number flow regime. Compositional simulations were performed to assess the impact of the experimental results for vertical- and horizontal-well geometries.
Introduction
Well-deliverability estimates for gas/condensate systems require accurate prediction of both gas and condensate effective permeability. This is particularly important within the near-wellbore region where the pressures often fall below dewpoint causing retrograde condensation. Within this region, pressure gradients in both flowing phases are large and the interfacial tension between the gas and condensate is low. This results in relative permeabilities that are rate sensitive. Under these conditions, both capillary number and non-Darcy effects must be considered in modeling of gas/condensate flows. The relative permeabilities increase with increasing capillary number and are reduced by inertial, or non-Darcy, flow effects.
Gas and condensate relative permeabilities are typically determined by steady-state linear coreflood experiments. Numerous experimental studies have been performed demonstrating an improvement in both gas and condensate relative permeability at high velocities and at low interfacial tension (Henderson et al. 1998; Henderson et al. 1997; Ali et al. 1997). These studies used model fluids to represent the reservoir fluid, which generally represented leaner gas/condensate systems. Chen et al. (1995) performed similar experiments using a recombined gas/condensate system from a North Sea field. Proper recombination with surface gas and condensate samples, however, assumes that the correct condensate/gas ratio is known. Using single-phase downhole samples obtained at pressures above the dewpoint eliminates this uncertainty.
Fevang and Whitson (1996) have shown that krg for a steady state process is a function of the krg/kro ratio, where the krg/kro ratio is a function of pressure. The dependency of krg on both the capillary number (Nc) and the krg/kro ratio for a pseudosteady-state process has been demonstrated experimentally by Whitson et al. (1999) and Mott et al. (1999). These studies used either model fluids or recombined reservoir fluids with krg/kro ratios primarily within the range of 1 to 90. The lower krg/kro ratios represent richer fluids, while the higher krg/kro ratios represent leaner fluids. The fluids studied in this paper, however, are significantly richer, with krg/kro ratios in the range of 0.05 to 0.15 on the basis of fluid compositions at initial reservoir conditions.
Non-Darcy or inertial effects reduce relative permeabilities. This has been demonstrated through linear coreflood experiments by several investigators (Lombard et al. 2000; Henderson et al. 2000; Mott et al. 2000). Multirate non-Darcy single-phase experiments were performed as part of this study because of the anticipated high flow rates from this reservoir.
The objectives of this study were (1) to experimentally measure gas and condensate relative permeabilities for a rich gas/condensate system using a live, single-phase reservoir fluid; (2) assess the magnitude of inertial effects through the measurement of the non-Darcy coefficient; and (3) evaluate the impact of the capillary-number-dependent relative permeabilities and non-Darcy effects on the performance of vertical and horizontal wells.