RETRACTED ARTICLE: A simulation approach: Miscible carbon dioxide injection in a carbonate reservoir

Pet.Sci. E L T C Ehsan Heidaryan 2 Jamshid Moghadasi 1 Marylena Garcia Quijada 0 0 Texas A&M University, College Station , Texas , USA 1 Petroleum Engineering Department, Petroleum University of Technology , Ahwaz , Iran 2 Faculty of Energy, Kermanshah University of Technology , Kermanshah , Iran The purpose of this study is to optimize the existing carbon dioxide (CO2) flood in deep dolomite formations by improving oil sweep efficiency of miscible CO2 floods and enhancing the was built to optimize the injection pattern. The model is a quarter of an inverted nine-spot and covCers20 conformance control. A full compositional simulation model using a detailed geologic characterization acres of field formation. Geologic description was used to construct the simulation grids. The sIimulation was performed to validate the model. Several sensitivity runs were made to improve TtheCO2 sweep layers represent actual flow units and resemble the large variation of reservoir properties. History match was determined approximately. Simulation results also indicate that a water-alRternating-gas (WAG) efficiency and increase the oil recovery. Finally, the optimum CO2 injection rate for dolomite formations incremental oil recovery of 18%. The additional recovery increases to 34% Aapolymer if is injected as ratio of 1:1 along with an ultimate CO2 slug of 100% hydrocarbon pore volume (HCPV) will allow an a conformance control agent during the course of the WAG process at a ratio of 1:1. According to the incremental oil recovery of 26%. D results, a pattern reconfiguration change from the nine spot to staggered line drive would represent an Miscible injection; carbon dioxide; dolomite formEation; water-alternating-gas 1 Introduction as immiscible and/or miscible. Recovery Amechanisms in Carbon dioxide (CO2) flooding processes can be classified oil swelling, and solution-gas drive. In Rgeneral, CO2 is very miscible flooding processes involve reduction in oil viscosity, the oil and reduces oil viscosity T (Martin and Taber, 1992) . soluble in crude oil at reservoir pressures; therefore, it swells multiple-contact miscibility Eprocess, which starts with denseMiscibility between CO2 and crude is achieved through a into the oil, making Rlighter it and often driving methane out phase CO2 and hydrocarbon liquid. The CO2 first condenses ahead of the oil bank. The lighter components of the oil then vaporize into the CO2-rich phase, making it denser, more like the oil, and thus more easily soluble in the oil. Mass transfer continues between the CO2 and the oil until the two mixtures become indistinguishable in terms of fluid properties (Jarrel et al, 2002). Because of this mechanism, good recovery may occur at pressures high enough to achieve miscibility. In general, the high pressures are required to compress CO2 to a density at which it becomes a good solvent for the lighter hydrocarbons in the crude oil. This pressure is known as minimum miscibility pressure (MMP), that is the minimum pressure at which miscibility between CO2 and crude can occur (Martin and Taber, 1992) . This pressure is usually determined experimentally. The water-alternating-gas (WAG) process consists of the injection of water and gas as alternate slugs by cycles or simultaneously (SWAG), with the objective of improving the sweep efficiency of water flooding and miscible or immiscible gas-flood projects by reducing the impact of viscous fingering. A schematic of the WAG process is shown in the literature (Caudle and Dyes, 1959). During a WAG process, the combination of higher microscopic displacement efficiency of gas with better macroscopic sweep efficiency of water helps significantly increase the incremental production over a plain water flood. A wide variety of gases have been employed for a wide range of reservoir characteristics in the miscible mode; however, CO2 and hydrocarbon gases represent approximately 90% of the injected gases (Jarrel et al, 2002). 2 Reservoir performance Cores analyses have revealed that the "main pay" interval presents three rock types: a pelletal dolomite packstone with interparticle and intercrystal porosity (pelletal packstone); a fossiliferous dolomite wackestone with moldic porosity (moldic wackestone); and a fossiliferous dolomite packstone with moldic and interparticle porosity (moldic packstone). The pelletal packstone rocks occur both as homogeneous units and in burrows and irregular patches in the wackestones. They have excellent reservoir rock properties; permeability can be as high as 152 mD and interparticle porosity is up to 24.3%. In the wackestone rocks, molds and vugs are the dominant pore types, ranging in size up to 6 mm. Within the rock, molds are not in contact with other; core observation has revealed that the molds are isolated in the rock by a relatively tight matrix. This isolated moldic porosity negatively affects the reservoir properties of the rock. Permeability is less than 1 mD, even though moldic porosity may range as high as 10% (Mathis and Sears, 1984) . Summary of reservoir data are given in Table 1. In preparation for the CO2 flood, the random waterflood pattern was converted into a nine-spot pattern. Oil production response was observed soon after injection began and the oil cut rose from below 14% to 31%. In Fig. 1 CO2 response can be clearly seen on a plot of oil cut versus cumulative production. The WAG flood in the WAG area was started with a constant 1:1 gas/water ratio (1% hydrocarbon pore volume (HCPV) of CO2 and 1% HCPV of water). The original injection schedule involved injecting alternating 6-month slugs of CO2 and water until a 40% HCPV slug of CO2 had been injected. 60 50 40 % t,u 30 c li O20 10 0 0 E Carbon dioxide injection L C I T 75 100 125 150 175 Cumulative oil production, MMB 200 25 50 Fig. 1 Oil Rcutversus cumulative oil production during continuous CO2 flood Reservoir characteristics Producing area, acres Formation Average depth, ft Gas-oil contact, ft Average permeability, mD Average porosity Average net oil pay thickness, ft Oil gravity, degrees API Reservoir temperature, °F Primary production mechanism Secondary production mechanism Original reservoir pressure, psia A1805 Tertiary production mechanism CO2 miscible Bubble point pressure, psia R 1805 Average pressure at start of secondary recovery, psia ±800 / ±1100 Target reservoir pressure for CO2, Tpsia Initial formation volume factor E(FVF), bbl/STB R Solution gas-oil ratio at original pressure, bbl/STB Solution gas-oil ratio at start of secondary recovery original pressure, scf/bbl Oil viscosity at 60 °F and 1100 psia, cP Minimum miscibility pressure, psia Values 25505 Dolomite 5200 1325 5 A 3 Parameters modeling 0.12 DThereservoir oil is saturated black oil with a stock tank 137 gravity of 33° API and an initial gas-oil ratio (GOR) of 660 33 Esacref/S1T,8B0.5Inpistiiaalartesaerrevfoeirrepnrceessduerpethanodf b5u,b0b0l0e fptoainntdp1r0es5su°rFe. 105 TThe CO2 minimum miscibility pressure (1,300 psia) was Solution-gas drive determined experimentally. Reservoir fluid compositions are WaCterflood available in Table 2. The tuning of the equation of states (EOS) in this work followed the methodology suggested by Khan et al (1992) to characterize CO2 oil mixtures. The P-R EOS (Peng and Robinson, 1976) was chosen as the EOS model because it has been found adequate for low-temperature CO2/oil mixtures (Khan et al, 1992). The viscosity model considered to match the oil viscosity of the reservoir fluid 2200 was the Lohrenz-Bray-Clark (LBC) model (Lohrenz et al, 1.312 1964) which is a predictive model for gas or liquid viscosity. Detailed PVT experiments are listed in Table 3; and the 420 related preliminary matches by the basic EOS. The maximum oil relative permeability is 65% at 15% 1060 connate water saturation (Swc=15%). At 60% water saturation 1.18 i(nSjwecc=t6e0d%,t)h,ethweatoeirl rreellaattiivvee ppeerrmmeeaabbiliiltiytyi niscrzeearsoe.s,Arseawcahtienrgias 1300 maximum value of 50% at 60% water saturation. WAG 1 1 No treatment Polymer History 20 0 180 160 140 120 d /B 100 T ,So 80 Q 60 40 20 0 180 160 140 /d120 TB100 ,S 80 o Q 60 40 20 0 0 0 5 10 15 25 30 35 40 20 Years 5 10 15 20 25 40 T Years Fig. 10 Comparison of staggered-line-drive, line-drive, E and nine-spot well patterns R both the staggered-line-drive pattern and the line-drive pattern create an immediate peak above 100 STB/d in the production rate, which represents approximately a 66% of increase in production as a result of the pattern reconfiguration. 6 Conclusions (Edited by Sun Yanhua) Line drive T 9 spot C A 45 50 Recovery from a WAG process is a function of the injection rate as well as WAG ratio and the CO2 slug size . WAG injection is effective in increasing the sweep efficiency of the injected CO2 in reservoirs. Simulation shows that tertiary CO2 flood would have a maximum recovery of 18% at a 1:1 WAG ratio and a CO2 slug size of 100% HCPV . CO2 at a WAG 1:1 ratio. The injection of visEcous water and The optimum injection rate for the pattern is 300 RB/d of an incremental oil recovery of 32% and L20% respectively. polymer results in a positive production response that yields inverted nine-spot pattern to a staCggered-line-drive improves Modeling suggests that converting the pattern from the the production oil rate by 66%I.The CO2 pattern modeling T Caudle B H and Dyes A B. Improving miscible displacement by gasJarrel P M , Fox ACE , Stein M H , et al. Practical aspects of CO2 flooding . water injection . Transactions of AIME . 1959 . 213 . 281 -284 Monograph Series . Society of Petroleum Engineers. 2002 Dphase CO2/oil mixtures . Paper SPE/DOE 24130 presented at SPE / Khan S A , Pope G A and Sepehrnoori K. Fluid characterization of threeEOklahoma DOE Enhanced Oil Recovery Symposium , 22 - 24 April 1992, Tulsa, Lohrenz J , Bray B G and Clark C R. Calculating viscosities of reservoir fluids from their compositions . Journal of Petroleum Technology. Martin D F and Taber J J. Carbon dioxide flooding . Journal of Petroleum Technology . 1992 . 44 ( 4 ): 396 -400 Mathis R L and Sears S O. Effect of CO2 flooding on dolomite reservoir rock . Paper SPE 13132 presented at SPE Annual Technical Conference and Exhibition , 16 -19 September 1984 , Houston, Texas Peng D Y and Robinson D B . A new two-constant equation of state. Industrial & Engineering Chemistry Fundamental. 1976 . 15 ( 1 ): 59 - 64