SPE 94073

Oil Recovery by Near-Miscible SWAG Injection M. Sohrabi, A. Danesh, and D. H. Tehrani, Petroleum Engineering Institute, Heriot-Watt U.

Copyright 2005, Society of Petroleum Engineers This paper was prepared for presentation at the SPE Europec/EAGE Annual Conference held in Madrid, Spain, 13-16 June 2005. This paper was selected for presentation by an SPE Program Committee following review of information contained in an abstract submitted by the author(s). Contents of the paper, as presented, have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material, as presented, does not necessarily reflect any position of the SPE, their officers, or members. Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes without the written consent of the Society of Petroleum Engineers is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O. Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.

Abstract We present micromodel visualization of pore-scale displacement mechanisms and fluids distribution during recovery of residual oil by near-miscible Simultaneous-Water-And-Gas (SWAG) injection under conditions of very low gas-oil interfacial tension (IFT) and negligible gravity forces. A 3-phase fluid system with gas-oil IFT of 0.08 mN/m, representing a near-miscible fluid system, together with high-pressure (5100 psia) micromodel were used for the flow experiments.

We demonstrate that in near-miscible SWAG injection, the oil recovery continues significantly and almost all of the oil that has been contacted by the gas will be recovered, after the main gas breakthrough. This oil is recovered by a microscopic mechanism, which is strongly linked to the low IFT between the oil and gas and to the perfect spreading of the oil over water, both of which occur as the critical point is approached. We have performed experiments with different SWAG (gas/water volumetric) ratios and we show that the ultimate oil recovery during SWAG injection is almost independent of SWAG ratio, within the tested range of 20 to 50%. This makes SWAG injection an attractive IOR process especially for reservoirs where continuous and high rate gas injection is not possible (e.g. due to supply constraint). In oil reservoirs that cannot export gas (remoteness from the market) WAG (for abundant gas volume) or SWAG (for low or high volume of gas) can be implemented. Introduction Oil recovery from mature waterfloods can be enhanced by combined injection of water and gas in an alternating (WAG) or simultaneous (SWAG) scheme. However, less experience has been gained on SWAG compared to WAG and hence the process is less known.

SWAG appears to provide better control of gas mobility than WAG, resulting in improved sweep efficiency in addition to more steady gas production and GOR response. In some oil reservoirs (especially offshore) a relatively small amount of produced gas and/or a rapidly falling gas rate makes it uneconomical to develop a gas export solution. It is also not economically viable to supply gas to these reservoirs for a continuous gas injection scenario e.g. remoteness from source of gas. Re-injection of the produced associated gas together with water to provide reservoir pressure support, better sweep and hence increased recovery, appears to be the best overall solution. In these cases, SWAG offers a solution whereby a changing mixture of injection fluids could be used and a total void replacement could be achieved by combining produced gas and water supplemented by sea water to the required injection volume. From environmental point of view, in cases where export of gas is not economical, re-injection of the produced gas in a SWAG scheme can significantly reduce or in certain cases eliminate the need for flaring. SWAG injection can cut both capital and operating costs and improve gas handling and oil recovery. It can also eliminate the need for separate water and gas injection lines to the injection wells. The improved gas handling and oil recovery have been reported for SWAG injection at Siri field1,2, the Joffre Viking3 CO2 miscible flood and SWAG emulation at the Rangely4 CO2 miscible flood. Pilot tests performed on Kuparuk River Field in Alaska5-6 have also demonstrated the feasibility of SWAG injection. In this paper we report pore scale events observed in micromodel experiments during near-miscible SWAG processes with hydrocarbon gases. The main purpose of the study was to gain a better understanding of the pore-scale near-miscible displacement mechanisms underlying the SWAG process. We present visualisation of fluid distribution and pore-scale mechanisms for recovery of residual oil by near-miscible SWAG injection under conditions of water-wet and negligible gravity forces in high-pressure glass micromodels. The results reveal the pore-scale mechanisms of SWAG process at near miscibility. In particular, we show that behind the main gas front, the recovery of the oil continues and almost all of the oil that has been contacted by the gas will be produced. The ultimate oil recovery during SWAG injection is almost independent of SWAG (water/gas) ratio within the test range of 20 to 50%.

2 SPE 94073

EXPERIMENTAL FACILITIES For the displacement experiments a high-pressure glass micromodel rig has been used. The rig has been designed to operate at pressures of up to 6000 psia with high quality images of fluids distributions and pore-scale mechanisms operating during displacement experiments being recorded. The experiments reported here were carried out at 5100 psia pressures and 100 F temperature and the fluid system used represented a near-miscible fluid system. Fig. 1 shows the glass micromodel (and its pore pattern) used in this study. The pore pattern was etched on the glass surface using acid etching technique. The depth of the pores was between 35-45 m with the width of the pores being between 35 to 400 m. Details of the experimental set up have been described elsewhere7,8. Fluid System. The experiments reported here used a simple, well-characterised, three-component, three-phase fluid system that exhibited a near-miscible (near-critical point) gas/oil system at 37.8 C (100 F) and 5100 psia. The aqueous phase was distilled water with 0.1% (wt/wt) blue dye. The hydrocarbon phases were mixtures of n-decane (n-C10) and methan (C1). To avoid mass transfer during the experiments, the fluids were pre-equilibrated at the conditions of the experiments, (100 F and 5100 psia). At 100 F, the critical pressure of the C1-C10 system is around 5300 psia so at 5100 psia (pressure of the experiments) the gas-oil system is very close to its critical point and hence, nearly miscible. At the prevailing conditions of the experiments the viscosity of the hydrocarbon vapour (gas) and hydrocarbon liquid (oil) is estimated to be 0.0378 and 0.1085 mPa.s respectively. Equilibrium values of three-phase interfacial tensions (IFT) of the fluid system, at the above temperature and pressure, were measured and are given below:

Interfacial Tension/mNm-1Gas/Oil 0.08 +0.01Gas/Water 41 +2Oil/Water 41 +2

EXPERIMENTAL RESULTS Using the above fluid system and micromodel a number of SWAG experiments were carried out. All of the displacements were performed quasi-statically at 5100 psia and 100 F with the micromodel mounted horizontally to minimize gravity effects. Here we report the results of two near-miscible SWAG injection experiments. The first experiment was carried out with a gas to water ratio of 50% and in the second one the gas to water ratio was reduced to only 20%. In either experiments, the micromdel was first flooded with water to simulate a conventional waterflooding of an oil reservoir. Then water and gas, which was nearly miscible with the oil, were injected in a SWAG mode to recover some of the residual oil. Details of the experimental procedure and discussion for these two experiments are given in the following sections.

SWAG with 50% gas to water ratio. Initially the micromodel was fully saturated with clear distilled water and pressurised to 5100 psia and subsequently displaced with blue live water, equilibrated with gas and oil at 5100 psia and 100 F. To simulate primary drainage (initial migration of oil into the water bearing porous media), equilibrated oil was injected from one end of the horizontal micromodel. The injection of the oil was carried out at a low velocity 1.2 m/d (Nc 1E-7) and continued until oil reached the other end of the micromodel. Fig. 2 depicts the initial water saturation established after oil injection with the blue colour representing water and the bright colour representing oil. Fig. 3 is a magnified image of a section of the micromodel at the end of primary drainage of water (oil injection through water saturated micromodel). It shows the relative position of the wetting phase (blue water) and non-wetting phase (white oil), in a water-wet micromodel. The direction and shape of the water-oil interfaces are good indication of water-wet conditions. As well as occupying some of the smaller or dead-end pores water is also present as a wetting layer over the pore surface, even in oil-filled pores, as expected. Water flooding began after the initial oil injection (and establishment of connate water). Water was injected into the micromodel at a low rate of 0.01 cm3/h, corresponding to a capillary number of 1E-7 and a linear velocity of 1.2 m/d. Due to a very low rate of water injection, high water/oil IFT (41 mN/m) and the micromodel being strongly water-wet, water was observed to enter the model by layer flow (as opposed to piston wise displacement) with no sharp waterfront. The water layers, surrounding the oil, were seen to thicken progressively. Also the oil was displaced mostly from the sides and corners of the pores, with residual oil remaining in the middle of the pores in shape of narrow oil filaments. Fig. 4 compares a section of the micromodel before and after water flooding. Comparison of Fig. 4(a) and Fig. 4(b) highlights the fact that during water flooding the dominant mechanism has been layer flow as opposed to piston-type. The slow thickening of water films at the sides and corners of the oil filled pores was a direct consequence of a capillary dominated flow regime. A similar flow mechanism was previously observed and reported by the authors7. The waterflood continued until no more oil production or changes in fluids distribution was observed. At the end of this waterflood, simultaneous injection of the near-miscible gas and water commenced with the same total rate as the previous waterflood. SWAG injection was carried out from the same end of the micromodel as that in the waterflood. As soon as the invading gas came in contact with the residual oil, the gas front was covered with a small amount of oil during SWAG injection. As the gas advanced further, a small bank of oil formed and moved ahead of the gas front. The oil bank, moving ahead of the gas front, supplied oil to pores where residual oil was present. This resulted in thickening of those oil filaments, which came in contact with the invading gas or the oil front ahead of it. As a result of this, the oil ganglia left behind after the initial waterflood could be

SPE 94073 3

divided in two types. The first type is oil in form of narrow filaments surrounded with thick water layers. This is the oil that has not been contacted by the in coming gas (or it associated oil bank). The second type is fat oil ganglia connected with the gas channels. These have been formed by local oil transfer brought about by the gas invasion. Fig. 5 depicts the fluid distribution during the near-miscible SWAG injection after 1 hour of simultaneous water and gas injection, with blue, bright and yellow colour representing water, oil and gas, respectively. As it can be seen, despite a very low value of gas-oil IFT, gas has made only a single channel through the network of pores with a significant amount of oil being bypassed by the gas. There are two reasons for this. First, in water flooded porous media the water saturation is high which tends to block the movement of gas and shield the residual oil from being contacted by it. This effect (water shielding) is even more severe in low gas-oil IFT conditions. The second reason is the topology of the porous medium, which causes some of the oil that has been contacted by the gas or its associated oil bank, to become apparently trapped in dead-end or semi dead-end pores. Dead end-pores are those that are physically dead-end and semi dead-end refers to those pores, which are not physically dead-end but the flow of oil is restricted by water at one end (water/oil IFT is by far larger than gas/oil IFT in near-miscible gas injections). During SWAG injection, we observed that the saturation of the oil, which was connected to the gas channel, kept reducing and led to a complete recovery of all the oil that was contacted by gas. Fig. 6 shows the distribution of the fluids within the micromodel after continuation of the near-miscible gas flood for 3.5 hours. The gas and water saturations have significantly increased and the oil saturation has decreased. In fact, as it can be seen in Fig. 6, almost all of the residual oil that had been contacted by the initial gas channel has now been recovered. To demonstrate this point more closely Fig. 7 has been prepared. This shows the magnified pictures of a section of the micromodel at the beginning of the SWAG injection and 3.5 hours after the injection has started. By comparing the location of the gas-oil and water-oil interfaces in these two images, it can be seen that, as the SWAG injection continued, the gas-oil and water-oil interfaces approached each other thereby invading the oil-occupied pores, albeit, the gas-oil movement has been much more pronounced than the advancement of the water-oil interface. As SWAG injection continued more oil was recovered and eventually all of the oil, which had been contacted by gas was produced.

SWAG with 20% gas to water ratio. In this experiment exactly the same procedure that was explained for the previous test was followed with the only difference being that a lower gas to water ratio of 20% was used during SWAG injection. Fig. 8 and Fig. 9 present the results of the SWAG injection with 20% gas when the injection had just begun and 3.5 hours

after the start of the injection respectively. Examination of these Figures and comparison with the results of the 50% SWAG ratio case reveals that in both cases, recovery of oil after the gas breakthrough continued and has been a significant part of the overall oil recovery. It is interesting to note that in the SWAG test where only 20% of the total injection volume (gas + water) was gas, qualitatively, almost the same amount of oil was recovered as was in SWAG with 50% gas. In both cases, as a result of continued injection of water and gas, almost all of the residual oil that had been contacted by gas had been produced. In other words, injection of gas in SWAG mode, even in small quantities, helps improving the oil recovery. This is, of course true only under our prevailing experimental conditions and the range of SWAG ratios tested and might not be suitable to generalise to actual field conditions. DISCUSSION In near-miscible SWAG injection experiments that are reported here, a significant part of the residual oil (to waterflood) was recovered. Careful examination of the micromodel images and videos recorded during these near-miscible SWAG injections reveals that crossflow takes place between the oil in the bypassed pores and the gas path, after the main gas front has passed. As a result of this crossflow, oil is transferred from the bypassed pores to the pores swept by gas where it can be recovered quickly. A combination of a very low gas/oil interfacial tension (IFT), perfect wetting (or spreading) conditions of the oil (which is sometimes referred to as critical-point wetting), and also simultaneous flow of the near-miscible gas, oil and water in the main flow stream facilitates this pumping action during near miscible SWAG injection. Close to the critical point any third, non-critical phase will be covered by a layer of one of the near-miscible phases. This phenomenon is known as critical-point wetting. According to Cahn9, a transition from partial to complete wetting is expected to occur when the gas-oil critical point is approached, e.g. upon an increase in temperature, pressure or oil becoming richer in dissolved gas. Consequently, in near-miscible gas floods the oil perfectly wets the solid surface (wetting layer) or water substrate (spreading layer in water-wet systems) and therefore the bypassed oil remains connected to the oil in the pores already swept by the gas. When gas invades a porous medium containing oil and water, even at very low or zero gas/oil IFT, there is always going to be some oil left behind in the pores. This is mainly because some pores show less resistance to the gas flow than the others. As a result, when the main gas front has passed, there is still some oil (bypassed at the pore level) occupying pores that are connected to the established gas path. In cases where the gas is near-miscible, due to a very low gas-oil IFT the capillary forces are no longer dominant, therefore the threshold capillary pressure resisting the entry of the gas into the bypassed oil is very small hence, the gas pressure can overcome the very small capillary barrier and moves the interface into the oil phase. This provides an effective driving force to transfer the bypassed oil into the main flow stream.

4 SPE 94073

Once the oil has been transferred into the main gas path, its flow is further enhanced by coupling with the flow of the gas which is in the middle part of the pores and the water which flows as layers on the walls of the pores. Since the IFT between the oil and the gas is very low (near-miscible) the gas and oil can flow together10,11 in the middle of the pores with water flowing alongside them in the corners and tighter parts of the pores cross section. The results show that the amount of gas is not a crucial issue as long as there is enough gas to connect up the isolated oil ganglia remained after the initial waterflood and also as long as enough gas is circulated to maintained a continuous gas path in the porous media. This process best suits the cases where only little gas is available or some gas is available seasonally. There may be some oil trapped in pores that are isolated and cannot be easily contacted by gas. Therefore, we concluded that practically all the oil that could be contacted by gas was recovered, in our experiments. Conclusions 1. Even in a homogeneous porous medium and low IFT

floods, there is always some trapped oil (bypassing at the pore level) due to topological effects, water-shielding and dead-end pores.

2. Near-miscible SWAG injection can produce a significant

amount of the residual oil from waterflooded reservoirs. 3. In near-miscible SWAG injection the recovery of oil

after breakthrough of the main gas front is significant and can lead to total recovery of the oil that is contacted by gas.

4. In SWAG injection with high gas-water ratio, extra

amount of injected gas does not significantly help the oil recovery.

5. Oil that is bypassed can be recovered by crossflow from

the bypassed pore to the fast track gas path. Acknowledgements The above study has been sponsored by: The UK Department of Trade and Industry, BP Exploration Operating Company Ltd, BHP Billiton Ltd, Norsk Hydro a.s., PDVSA Intevep, Shell UK Exploration and Production, Statoil and Total Exploration and Production UK which is gratefully acknowledged. References 1. Berg, L.I., Stensen, J.A., Crapez, B. and Quale, A.:

SWAG Behavior Based on Siri Field Data paper SPE 75126, SPE/DOE Improved Oil Recovery Symposium, Tulsa, Oklahoma, 13-17 April 2002.

2. Quale, E.A., Crapez, B., Stensen, J.A., Berge, L.I.:

SWAG njection on the Siri Field An Optimized Injection System for Less Cost paper SPE 65165, SPE European Petroleum Conference, Oct. 2000.

3. Stephenson D.J., Graham A.G., Luhning R.W.: Mobility Control Experience in the Joffre Viking Miscible Carbon Dioxide Flood, SPE Reservoir Engineering, Aug. 1993, p183-188.

4. Attanucci V., Asbsen K.S., Hejl K.A., Wright C.A.:

WAG Process Optimization in the Rangely Carbon Dioxide Miscible Flood, paper SPE 2662, SPE 68th Annual Technical Conkrence, Oct. 1993.

5. Stoisits, R F., Krist G. J., Ma, T. D., Rugen, J.A., Kolpak

M. M., and Payne, R L.: Simultaniuous Water and Gas injection Pilot the Kuparuk River Field, Surface Line Impact, Paper SPE 30645, SPE Annual Tech. Conf., Oct. 1995.

6. Ma, T. D., Rugen, J. A. and Youngren, G. K.:

Simultaneuous Water and Gas Injection Pilot at the Kupamk River Field, Reservoir Impact, Paper SPE 30726, SPE Annual Technical Conference and Exhibition, Oct. 1995.

7. Sohrabi, M., Tehrani, D.H., Danesh, A., and Henderson,

G.D.: Visulisation of Oil Recovery by Water Alternating Gas (WAG) Injection Using High Pressure Micromodels Water-Wet System paper SPE 63000 presented at the 2000 SPE Annual Technical Conference and Exhibition, Dallas, Oct. 1-4.

8. Sohrabi, M., Tehrani, D.H., Danesh, A., and Henderson,

G.D.: Visualisation of Oil Recovery by Water Alternating Gas (WAG) Injection Using High Pressure Micromodels - Oil-Wet & Mixed-Wet Systems paper SPE 71494 presented at the 2001 SPE Annual Technical Conference and Exhibition, New Orleans, Louisiana, 30 September3 October.

9. Cahn, J.W., Critical point Wetting, J. Chem. Phys.

(1977) 66, 3667. 10. Williams, J.K. and Dawe, R.A., photographic

observations of unusual flow phenomena in porous media at interfacial tensions below 0.1 mNm-1, Journal of Colloid and Interfaces Science, 124 No.2, 691-696, August 1988.

11. Jamiolahmady, M., Danesh, A., Tehrani, D.H and

Duncane, D.B., A mechanistic model of gas-condensate flow in pores, Transport in porous media 41: 17-46, 2000.

SPE 94073 5

Fig.1- Horizontal micromodel 100% saturated with blue water. The left picture shows the whole micromodel in small scale. The right picture is the same but divided in two halves, in a larger scale, for clarity.

6 SPE 94073

Fig. 2- Initial oil and water distribution before water flooding (Virgin Conditions).

SPE 94073 7

Fig.3- A magnified section of the water-wet micromodel at the end of primary drainage of water demonstrating shapes of water (blue)/oil (white) interfaces.

(a) (b)

Fig.4- A magnified picture of the micromodel before (a) and after (b) water flood.

8 SPE 94073

Fig.5- Fluid distribution within the micromodel after 1 hour near-miscible SWAG injection with a gas to water ratio of 50%.

SPE 94073 9

Fig.6- Fluid distribution within the micromodel after 3.5 hours near-miscible SWAG injection with a gas to water ratio of 50%. Almost all of the residual oil that had come in contact with the gas has been recovered.

10 SPE 94073

Fig.7- A section of the micromodel during gas injection. A) at the beginning of SWAG injection B) 3.5 hours after the SWAG injection started. Note that the gas channels have widened and oil ganglia have shrunk as a result of the gas/oil and water/oil interfaces approaching each other.

SPE 94073 11

Fig.8- Fluid distribution within the micromodel after 1 hour near-miscible SWAG injection with a gas to water ratio of 20%.

12 SPE 94073

Fig.9- Fluid distribution within the micromodel after 3.5 hours near-miscible SWAG injection with a gas to water ratio of 20%.