Efficiency of Horizontal and Vertical Well Patterns on the Performance of Micellar-Polymer Flooding

1Department of Natural Resources and Geoenvironmental Engineering， Hanyang University， Seoul 133-791， South Korea

*Corresponding author.

Supported by Korea Gas Corporation.

Received 12 January 2012; accepted 15 February 2012

Abstract

There is increasing interest in micellar-polymer flooding because of the need to increase oil production from depleted and waterflooded reservoirs. Using horizontal wells for injection and production in a micellar-polymer flood process， higher sweep efficiency is expected compared with the use of conventional patterns by vertical wells. However， the use of horizontal wells is very sensitive to the well pattern designed to operate the process. This paper presents an analysis of how the overall performance of a micellar-polymer flood process in anisotropic reservoirs is influenced by the well pattern using horizontal injector and producer in different configurations.

A three-dimensional numerical simulator for fluid flow and mass transport is used to analyze the effectiveness of well combinations in micellar-polymer applications. The potential for a horizontal well application was assessed through different situations in combinations of injection and production wells and degree of reservoir anisotropy. Results from the study have demonstrated that significant amount of oil can be recovered additionally and injectivity was remarkably improved by utilizing a combination of horizontal wells. The improvement of injectivity through a horizontal injection well was higher when it was combined with horizontal producer parallel to the injector. The overall performances in anisotropic reservoirs strongly depend on the type of wells considered and the orientation of the horizontal wells with respect to the permeability directions. Combination of horizontal wells placed parallel to the low permeability direction yields the best performance. In high permeability ratio reservoirs， the presence of horizontal injectors is more significant in defining the efficiency of the micellar-polymer flood than the horizontal producers.

Key words: Micellar-polymer flood; Horizontal well; Anisotropy; Injectivity

INTRODUCTION

Micellar-polymer flooding is an enhanced oil recovery(EOR) mechanism that can be often utilized after the natural drive or waterflood mechanisms become ineffective. During micellar-polymer flooding projects， water with surfactant is injected to achieve ultra low interfacial tension (IFT) which causes to decrease the residual oil saturation trapped by capillary forces. Lowering the mobility ratio， polymeric additives are used to improve a sweep mechanism that drives the reservoir oil toward the production wells[1， 2].

The efficiency of this EOR process is dependent on a number of parameters that are specific to the field under study. Most of the micellar-polymer flooding projects use vertical wells as injectors and producers， however， horizontal wells promise greater injectivity and productivity characteristics. The higher injectivities allows surfactant-polymer slug and polymer solution to be injected at much higher rates or lower pressures in horizontal wells than in vertical wells， which leads to allowing oil to be recovered quicker or with less energy. The use of horizontal wells has been increasing very rapidly throughout the oil industry as advances in drilling techniques continue. Horizontal wells have been used primarily in problem reservoirs or to solve specific production problems including low permeability fractured formations， low-permeability gas reservoirs， unconventional gas sources such as coal-bed methane or shale， gas or water coning， and thin formations. Because the horizontal well technology is recently being applied to the production of crude oil by waterflooding or enhanced oil recovery methods， little information is available on the horizontal-well applications for chemical floods[3-5].

Although there is relatively little published information on the use of horizontal injection or production wells other than for thermal recovery[6]， the need for patterns of both horizontal injection and/or production wells to increase the rate of flooding in EOR processes has been increased. In this study， a comparison of the efficiencies of both horizontal and vertical wells in micellar-polymer flooding operations is performed. The influences of reservoir anisotropy coupled with various design parameters of well patterns are also studied. The study of these effects may assist the project design engineers in choosing the most optimal operating conditions that will maximize the efficiency of the process. With the goal of identifying these conditions， the effectiveness of the horizontal and vertical wells in micellar-polymer flooding projects is examined by simulating the process numerically.

1. MATHEMATICAL THEORY

Simulation of micellar-polymer flood processes includes modeling phase behavior， oil-water IFT and as a function of surfactant concentration， surfactant and polymer concentration-dependent viscosities， salinity and permeability dependence of adsorption， shear-thinning rheology of the fluids， and permeability reduction[7]. The numerical study was performed with the UTCHEM software， which is a general reservoir simulator. Among the most advanced chemical EOR simulators， UTCHEM has proved to be particularly useful for modeling multicomponent and multiphase transport processes[8]. UTCHEM has been extensively verified by comparing to analytical solutions and experimental measurements for its ability to predict the flow of fluids through the reservoir. Thus， UTCHEM will be used in this study for simulating multi-dimensional micellar-polymer flood processes for enhanced recovery of remaining oil in the reservoir.

where κ is component index including water (κ=1)， oil (κ=2)， surfactant (κ=3)， and polymer (κ=4); l is phase index including aquatic (l=1)， oleic (l=2)， and microemulsion (l=3) phases; φ is porosity;Clu is overall

concentration of component κ (volume fraction);ρk is

density of component κ [ML?3]; np is number of phases; Cκl is concentration of component κ in phase l (volume fraction); ul is Darcy velocity of phase l [LT?1]; Sl is saturation of phase l; Rκ is total source/sink term for component κ (volume of component κ per unit volume of porous media per unit time); Dκl is dispersion tensor. The overall concentration (Clu ) denotes the volume of

component κ summed over all phases.

2. NUMERICAL MODELING

In this study， investigations on the effects of reservoir anisotropy on both horizontal and vertical wells performances in micellar-polymer flooding operations have been conducted. Numerical simulation runs were conducted in a three-dimensional oil reservoir that includes the reservoir thickness， i.e.， gravity and capillary forces are simultaneously considered. To simulate the performance of the micellar-polymer flood processes， a hypothetical study site of one-quarter of an injectionwell-centered five-spot is considered. The modeled system used in this study is a box-shaped reservoir with a horizontal area of 253×253 ft2 and a vertical thickness of 25. Vertically， the simulation domain consists of five layers; and each layer is discretized into 23×23 grid blocks. The outer boundary is represented as a noflow. The reservoir investigated in this study was assumed to have already been waterflooded and is a potential candidate for micellar-polymer flood. The model assumes that the reservoir is originally saturated with oil and connate water. Initial saturation of water was assumed to be uniform spatially in the reservoir at 0.50. The uniform permeability of 250 md is assumed for both horizontal and vertical directions.

Micellar-polymer flooding process considered in this study involves the injection of a surfactant-polymer slug followed by a polymer drive and chase water injection. Fluids are pumped into the injection well at constant pressure of 650 psi over a simulation period of 1，000 days. The reservoir fluids are recovered from the production well operating at a sand face pressure of 500 psia， the same pressure as the initial reservoir pressure.

In order to clarify the effects of various parameters during the flow through the reservoir， comparisons were made among results from simulations under different sce-narios of micellar-polymer flood. Different combination of injection-production well pattern and reservoir anisotropy were considered for this study. Reservoir oil， formation water， petrophysical properties of the reservoir， and injec- tion/production procedure were identical for all calculations. Input parameters for the simulations are those that define the physical properties of reservoir， fluid properties， and chemical properties， as given in Table 1.

3. RESULTS AND DISCUSSION

The model evaluated the flow of brine associated with surfactant and/or polymer and oil through a reservoir during the process. To understand the effects of various parameters on the oil recovery， simulation was performed with the injection sequence of micellar-polymer flooding followed by waterflooding. Volumetric fraction of surfactant in the injecting fluid is 0.03 during 0 to 180 days. Polymer concentration is 0.05% during 0 to 180 days， 0.025% during 180 to 360 days， and 0 % for remaining period in the chase water.

Several cases were studied in which the sensitivity of oil recovery and injection rate to the well configuration(well type and length) was determined. Performance of micellar-polymer flooding with vertical wells was determined by comparing the oil recovery and injection rate from a base case with the oil recovered from the various micellar-polymer floods over production period. 3.1 Well Types

Extensive simulations were undertaken to investigate the feasibility and compare applicability of micellar-polymer flood through vertical and horizontal wells. The objective of this parametric study is to investigate the effect of horizontal well orientation on the overall performance of micellar-polymer flooding projects using different injector/ producer combinations. The ratio of horizontal well length to reservoir length was 0.52， which corresponds to well length of 132 ft. During this part of the investigation， the following nine injection and production well combinations are considered:

(1)First combination (VzIVzP): vertical injector and vertical producer (base case)

(2)second combination (VzIHxP): vertical injector and horizontal producer along the x-direction

(3)third combination (VzIHyP): vertical injector and horizontal producer along the y-direction

(4)forth combination (HxIVzP): horizontal injector along the x-direction and vertical producer

(5)fifth combination (HxIHxP): horizontal injector along the x-direction and horizontal producer along the x-direction

(6)sixth combination (HxIHyP): horizontal injector along the x-direction and horizontal producer along the y-direction

(7)seventh combination (HyIVzP): horizontal injector along the y-direction and vertical producer

(8)fifth combination (HyIHxP): horizontal injector along the y-direction and horizontal producer along the x-direction

(9)sixth combination (HyIHyP): horizontal injector along the y-direction and horizontal producer along the y-direction

Figure 1 presents a schematic representation of these combinations.

Results of the calculations for isotropic reservoirs are shown and compared in Figure 2 for various combinations of vertical and horizontal floods. As presented in the figure， the oil recovery and injection rate are highly influenced by the well patterns. It can be seen that the predicted values from the reservoir simulation illustrate higher oil production and lower water production from horizontal micellar-polymer flooding.

The fifth and ninth combinations yield the highest additional production to the base case. This is expected since these combinations adopted only horizontal wells for injection and production and the producer and the injector have the same orientation. With horizontal wells aligned to the same orientation， areal sweep patterns are a lot closer to line-drive geometry the production will be maximized during the early injection period. The other combinations including horizontal wells also show markedly better performances than the base case as presented in Figure 2. The improvement of oil recovery can be attributed to the higher injection rate and larger area open to flow and resulting improved sweep efficiency in horizontal wells than that of vertical wells.

Figure 2(b) compares the results of injection rate for different injector-producer combinations， against the performance of the first combination which is considered as the base case. This comparison indicate that horizontal well floods result in the much higher rate compared to an equivalent five-spot flood at the same pressure. At early time， the presence of horizontal injectors seems to be more beneficial than the presence of horizontal producers. Conversely， the presence of horizontal producers is more beneficial than the presence of horizontal injectors after water breakthrough.

The combination of horizontal injector and producer shows the highest injection rate. The injection rate in HyIHyP at the same operating pressure would be about 6.1 times higher than that for the base case， which represents a significant improvement in injectivity and high effectiveness to sweep the reservoir oil over values attained by the model of vertical wells. The result implies that the same volume of fluid can be injected at much lower pressure， in turn. The higher injectivities allowed by horizontal injection wells can help to alleviate substantially less injectivity of a vertical injection well. The higher injectivity associated with horizontal wells can also help to mitigate the effects of chemical and thermal degradation of injecting fluids.

Figure 3 depicts oil saturation of the middle layer for five different well patterns including VzIVzP， VzIHxP， HxIVzP， HxIHyP， and HxIHxP after 410 days of injection. The sweep efficiency advantage of horizontal well flooding patterns would be observed best for horizontal injector and horizontal producer aligned to the same orientation. Analyzing the results presented in Figure 3， the swept region did not cover the entire area of the layer with vertical injector or producer. In cases of horizontal injector and producer， the flood front covered almost the entire region. The pore volumes injected for the well patterns are 0.43， 0.79， 0.49， 1.36， and 1.08 for VzIVzP， VzIHxP， HxIVzP， HxIHyP， and HxIHxP， and cumulative oil recoveries are 0.32， 0.51， 0.33， 0.59， and 0.60， respectively. The highest sweep efficiency was obtained for a pattern HxIHxP in which injected fluid and the produced fluid are flowing by two parallel horizontal wells. This type of well pattern is called inverted line drive pattern and has the advantage of using the entire length of the horizontal section for sweep. As compared to the sweep patterns that developed between vertical wells， areal sweep patterns are closer to line-drive geometry in horizontal wells.

3.2 Reservoir Anisotropy

This parametric study is to examine the performances of horizontal and vertical well combinations in micellarpolymer flooding a reservoir that exhibits anisotropic permeability characteristics. In the runs considered， the ratio of permeability of the reservoir along the x-direction to the permeability of the reservoir along the y-direction is assigned to be 0.25， 0.5， 1.0， 2.0， and 4.0， keeping the geometric mean constant. Nine different well combinations are also considered during this parametric study.

Figure 4 shows the change in cumulative oil recovery when the first combination is replaced by the second through the ninth combinations summarized above. When the horizontal wells are placed parallel to the low permeability direction or orthogonal to the high permeability direction， the horizontal combination yields the best performance. The overall performance of the combinations considered in this parametric study strongly depends on the type of wells considered and the orientation of the horizontal wells with respect to the permeability directions.

For permeability ratio less than 1.0 or higher permeability to the y-direction， combinations with horizontal well along with x-direction performs better. Due to the rapid and better sweep efficiency， the cumulative oil recovery of HxI and/or HxP remained remarkably higher. The results in Figure 4(a) indicate that HxIHyP， HyIHxP， and HxIHxP provide increment of 44%， 54%， and 73% more oil than VzIVzP does. Comparing the performances of HxIHyP and HyIHxP， the reservoir performance is more strongly influenced by injection well up to 230 days， then by production well. Comparing Figure 2 and Figs. 4 (a) and (b)， one can observe a higher oil recovery and a less difference in oil recovery among cases for smaller ratio of permeability.

Results in Figs. 4(a) and (b) also show that the advantages of combinations of horizontal injection and production wells for micellar-polymer flooding are greatest for patterns aligning orthogonal to high permeability direction. Therefore， equivalent (or better) water injection and oil production rates can be achieved in anisotropic reservoirs with far fewer horizontal wells along proper direction than with vertical wells. The smaller the permeability ratio， the larger difference in oil recovery was observed. For example， differences in oil recovery between VzIVzP and HxIHxP for kx/ky = 1.0， 0.5， and 0.25 are 57%， 64%， and 73%， respectively. Although individual horizontal wells cost much more， the total drilling costs could be less than for vertical-well patterns because fewer wells are drilled at the wider spacing.

CONCLUSIONS

Based on the studies carried out in this work in order to evaluate the oil recovery efficiency of a micellar-polymer flooding process with various combinations of horizontal and vertical wells， the following conclusions are drawn. Because of the improved injectivity and the potential for increased recovery by better sweep efficiency， the use of horizontal wells during micellar-polymer flooding could offer remarkably significant benefits as compared to the results obtained in a conventional pattern processed by vertical wells. Regardless whether the formation exhibits isotropic or anisotropic permeability characteristics， it is essential to place the horizontal producers and injectors parallel to each other to obtain better performance efficiency than the vertical well combination. A very favorable injectivity and sweep occur when two opposed horizontal wells parallel in the pattern are used for injection and production. Compared to five spot patterns with vertical wells， the combination of parallel horizontal wells can increase oil recovery by 40% and injectivity by as much as a factor of two.

In anisotropic reservoirs， the oil recovery will be maximized when the horizontal producers and injectors are orthogonally placed to the high permeability direction. In high permeability ratio reservoirs， the presence of horizontal injectors parallel to the low permeability direction becomes more dominant in defining the efficiency of the micellar-polymer flood than the horizontal producers.

REFERENCES

[1] Thomas， S.， Farouq Ali S. M. (2001). Micellar Flooding and ASP-Chemical Methods for Enhanced Oil Recovery， J. Canadian Petrol. Tech.， 40(2)， 46-52.

[2] Hirasaki， G. J.， Miller， C. A.， Puerto， M. (2008). Recent Advances in Surfactant EOR， IPTC 115386， SPE Annual Technical Conference and Exhibition. September 21-24， Denver， CO.

[3] Taber， J. J.， Seright， R.S. (1992). Horizontal Injection and Production Wells for EOR or Waterflooding. SPE 23952 SPE Permian Basin Oil and Gas Conference. March 18-20， Midland， TX.

[4] Dakhlia， H.， Wu， W.， Lim， M. T.， Delshad， M.， Pope， G. A.， Sepehrnoori， K. (1995). Simulation of Surfactant Flooding Using Horizontal Wells. CIM 95-82， Petroleum Society of CIM 46th Annual Technical Meeting. June 7-9， Banff， Alberta， Canada.

[5] Westermark， R. V.， Schmeling， J.， Dauben， D. L.， Robinowitz， S.， Weyland， H. V. (2006). Application of Horizontal Waterflooding to Improve Oil Recovery from Old Oil Fields. SPE 99668 SPE/DOE Symposium on Improved Oil Recovery. April 22-26 Tulsa， OK.

[6] Shedid， S. A. Abbas， A. A. (2000). Comparison of Chemical Steam Floods Through Vertical and Horizontal Wells. CIM 65482 SPE/Petroleum Society of CIM International Conference on Horizontal Well Technology. November 6-8， Calgary， Alberta， Canada.

[7] Fadili， A.， Kristensen， M.R.， Moreno， J. (2009). Smart Integrated Chemical EOR Simulation. IPTC 13762 International Petroleum Technology Conference. December 7-9， Doha， Qatar.

[8] Center for Petroleum and Geosystems Engineering (2000). UTCHEM-9.0 a Three-Dimensional Chemical Flood Simulator. Austin: U. of Texas at Austin.

[9] Delshad， M.， Pope， G. A.， Sepehrnoori， K. (1996). A Compositional Simulator for Modeling Surfactant Enhanced Aquifer Remediation， 1. Formulation. J. Contaminant Hydrol.， 23(4)， 303-327.