Oil & Gas - Maximising Recovery Programme (formerly SHARP) IOR Views

Introduction

Miscible Water-Alternate-Gas (WAG) injection is a commonly used EOR technique that combines the benefits of microscopic sweep efficiency obtainable from miscible gas injection with the better economics and frontal stability obtained from waterflooding.  Typically water and gas are injected alternately, in what is termed slug injection, although they may also be injected simultaneously.

Although WAG recovery schemes have been implemented in oil fields since the 1950s (see Caudle and Dyes, 1958; Christensen et al 2001) there are very few detailed experimental investigations of this process reported in the literature.  Moreover none of these laboratory studies provide sufficient experimental details to allow the testing of the ability of numerical simulators to predict the behaviour of these displacements.  This is despite the fact that almost all field trials have recovered less oil from WAG than initially predicted by simulation (see Christiansen et al, 2001). It is possible that this difference between theory and practice may be due to the combined influences of unknown or uncertain reservoir heterogeneity and gravity; however, it may also be due to an inadequate physical description of the WAG displacement process in reservoir simulators.

We have recently performed a series of well-characterised experiments investigating secondary WAG displacement behaviour during simultaneous and slug injection.  These experiments were performed using analogue fluids in glass-bead packs so we were able to visually observe the progress of the water and solvent fronts as well as record oil recovery and effluent profiles. Their aim was to:

• Investigate the impact of first contact miscible WAG injection on oil recovery.
• Clarify the physical processes during displacements.
• Provide benchmark data-sets to validate reservoir simulations.

A full description of the experimental details and our attempts to predict the experimental results using conventional simulation have been reported in Al-Shuraiqi, 2005 and Al-Shuraiqi et al, 2005.  In this article we summarise our principal findings and conclusions.

Experimental Overview

The experiments were performed in a homogeneous, 23cm×10cm×0.6cm, glass bead-pack mounted horizontally to minimise the influence of gravity. The thickness of the pack was selected to have essentially two-dimensional flow so that direct comparison with 2D numerical simulations could be made.  In all displacements ISOPAR V was used to represent oil and paraffin was used as the miscible solvent.  This resulted in a solvent-oil mobility ratio of about 7 and a water-oil mobility ratio close to 10.  Oil was injected into a fully water saturated pack until an initial immobile water saturation of approximately 8% was achieved.

Seven simultaneous and four slug injection WAG experiments were performed in all. Three simultaneous displacements were performed at a rate of 5ml/min to investigate the effect of WAG ratio on recovery, using WAG ratios of 1:4, 1:1 and 4:1. A further four simultaneous displacement experiments were performed at a WAG ratio of 1:1 using constant rates of 1, 2, 4, and 6ml/min to investigate the effect of rate on recovery.  The slug injection experiments were performed at a WAG ratio of 1:1 using slug sizes of 5%, 10%, 30% and 50% pore volumes.  This is in addition to a series of miscible and immiscible displacements used to characterise the pack properties.

Figure 1: Comparison between Actual Water-Solvent Fractional Flow Curves and those Calculated from a Water-Oil Displacement

Figure 2:  Experimentally Observed Rate Effects on Oil Recovery for Water-Flooding, Simultaneous WAG 1:1 Injection and Continuous Solvent Injection

Summary of Key Findings

• It is usually assumed in first contact miscible WAG simulations that the water-solvent fractional flow curve can be calculated from the water-oil relative permeabilities.  We found that the actual water-solvent fractional flow was significantly different from that derived from the water-oil relative permeabilities (Figure 1).
• The oil recovery from simultaneous WAG displacements showed significant rate dependence (Figure 2).  This was found to be due to the rate dependence of the water-oil relative permeability curves.
• The optimum WAG ratio for simultaneous WAG injection, obtained from Stalkup’s construction (2:3), was found to be very similar to that obtained experimentally (~1:1).  However, the experiment demonstrated that solvent viscous fingering was not significantly suppressed by the water at this WAG ratio (Figure 3). As a result solvent breakthrough predicted by the simulation was much later than in the experiment.
• The simulator could not predict the recovery and effluent profiles observed for the simultaneous WAG injection experiments at WAG ratios of 1:4 and 4:1. This is despite the fact that the simulator has been used successfully in the past to predict a variety of miscible experiments (e.g. Christie et al, 1989; Davies et al, 1991, Muggeridge et al , 2002)
• It was very difficult to achieve true simultaneous WAG injection.  This was due to a combination of inlet effects and the fact that solvent and water tended to establish separate pathways through the porous medium (Figure 3).  This was not due to small scale heterogeneity or wettability effects in the bead pack as the same effect was observed when the experiments were repeated, and in this case the solvent and water pathways were located in different parts of the pack.  The simulator could not model this behaviour.
• The recovery obtained from slug injection tended towards that obtained from simultaneous injection as the slug size decreased.  This was because the water slugs were completely broken down by solvent fingering in all cases except for the very largest slugs (30% and 50%).  The agreement between experimental and simulated recovery and effluent profiles was good although the match between the simulated and actual saturation distributions was not (Figure 4).

Figure 3: Fluids Injected on Left and Produced on Right - Water (red) and Solvent (blue) Fronts at 0.4 PVI of Water and Solvent Observed Experimentally During Simultaneous WAG Injection at a Ratio of 1:1 – (this ratio is close to the optimum WAG ratio at which, according to previous studies, viscous fingering is suppressed)

Figure 4: Fluids Injected on Left and Produced on Right - Comparison Between the Experimental (left) and Simulation (right) Saturation Fronts for the 5% PV and 50% PV Slugs at WAG Ratio 1:1 (for the experiments water is red and solvent is blue - there is significantly more solvent fingering observed in the experiments than is predicted by numerical simulation)

Conclusions

We have produced a good quality set of experimental data suitable for benchmarking/validating the numerical simulation of simultaneous and slug WAG displacements.  Initial comparisons with conventional finite difference simulation suggest that there are a number of aspects of WAG displacement that are not properly modelled by simulation.  This is despite the fact that excellent agreement is obtained between simulation and experiment for first contact miscible solvent injection and immiscible waterflooding. It is of particular concern that there is significantly more solvent fingering in both slug and simultaneous WAG experiments, than is predicted by the simulations.  However, further investigation is required before the implications of these findings for field scale WAG displacements are fully understood.

Our current efforts are focussed on providing a similar set of data for use in validating the simulation of multi-contact miscible displacements.

References

1. Al-Shuraiqi, H., Muggeridge A.H. and Grattoni C., “Numerical and Experimental Investigation into the Mechanisms of First Contact Miscible Simultaneous and Slug WAG Injection”, presented at the 13Th  EAGE European IOR Conference, Budapest, May 2005.
2. Al-Shuraiqi H., “Mechanisms of Oil Recovery via First Contact Miscible WAG Injection”, PhD Thesis, Imperial College 2005.
3. Caudle, B.H. and Dyes, A.B.: “Improving Miscible Displacement by Gas-Water Injection,” Transactions of the American Institute of Mining, Metallurgical, and Petroleum Engineering, 213 (1958), 281-284.
4. Christensen, J.R., Stenby, E.H. and Skauge, A.: “Review of WAG Field Experience,” SPE Reserv. Eval. Eng. 4(2), 97-106, 2001.
5. Christie M.A., Jones A.D.W. and Muggeridge A.H., "Comparison between Laboratory Experiments and Detailed Simulations of Unstable Miscible Displacement Influenced by Gravity", in “North Sea Oil and Gas Reservoirs - II”, (Graham & Trotman) 1994, 245-250 (Proc. of the North Sea Oil and Gas Reservoirs Conference, 1989).
6. Davies G.W., Muggeridge A.H. and Jones A.D.W., “Miscible Displacements in a Heterogeneous Rock: Detailed Measurements and Accurate Predictive Simulation”, SPE 22615, 66th Annual Technical Conference and Exhibition of the SPE, (1991)
7. Muggeridge, A.H., Jackson, H.D., Al-Mahrooqi, S., Al-Marjabi, M. and Grattoni, C.A.: “Quantifying Bypassed Oil in the Vicinity of Discontinuous Shales”, SPE 77487, 77th Annual Technical Conference and Exhibition of the SPE, (2002).
8. Stalkup, F.I.: “Miscible Flooding Fundamentals,” Society of Petroleum Engineers Monograph Series, 1983.