Summary Twenty-eight tertiary gas-injection corefloods were conducted to define the pressure/recovery curves for five crude oils. These results have been compared with slim-tube recovery tests using the same five reservoir fluids. Field-scale CO2 floods have been or are currently under way in three of the studied reservoirs (S, H, and M). The other two reservoirs (F and A) have been considered for CO2 and hydrocarbon-gas injection, respectively. The results for the four CO2 injection systems consistently showed that oil recovery decreases dramatically below the minimum miscibility pressure (MMP) exactly like slim-tube tests, while the recovery in the hydrocarbon-gas injection system decreased almost linearly with decreasing pressure both above and below the MMP. To use short cores, it is expedient to use an external transition-zone generator. A method has been developed that successfully circumvents the oil resaturation problem caused by most transition-zone generators. Introduction There is considerable economic and operational incentive to operate a gas-injection project at the lowest possible pressure. Incentives include purchasing smaller gas volumes and decreased gas-compression costs. Also, the possibility of operating a gas-injection project below the MMP means that reservoirs that might otherwise be excluded could be considered as candidates for gas injection. The effect of pressure on solvent (CO2 or hydrocarbon-gas) floods has historically been almost exclusively defined by slim-tube tests. But the slim tube is not representative of reservoir porous media. It is an oil-filled, homogeneous sand or bead pack containing no water. Most slim-tube recovery data exhibit a sharp decline as pressure decreases below the MMP. The sharp decline is used to help identify the MMP. Does this rapid decrease in recovery efficiency translate to corefloods and ultimately to the reservoir? The pressure effect in reservoir rock is normally not measured for several reasons. A suite of corefloods conducted with a specific solvent over a range of pressures represents considerable expense. Additionally, conducting corefloods below the MMP presents experimental difficulties. Yet, the only way to specify recovery efficiency below the MMP correctly is with reservoir-condition corefloods.

This paper describes a unified testing program to identify the best polymer gel systems for a particular field application. Beaker tests are used to rapidly screen gel systems.Coreflood tests are then employed for the final system selection.Results are presented from a full-scale laboratory evaluation of 15 commercially available polyacrylamide polymers with a Cr(III)-redox crosslinker for profile modification in a thick, multizone reservoir. Beaker tests showed that low levels of polymer hydrolysis were required to produce consistent gels at reservoir conditions.In addition, the pH of the gel solution in the buffered field brine had a large effect on the gel properties. Allied C~lloids Alcoflood®935L and American CyanamidCyanagel 150 were selected for further evaluation in coreflood tests.Either polymer gel maintained a permeability reduction of at least 97 percent after 50 days of static aging at reservoir conditions. The stability of an Alcoflood 9351 gel was evaluated during continuous injection and with exposure to near wellbore pressure gradients.

No abstract

l%m pap-sf was prepared for presentation at the SPEIDOE Tenth Symposwm on improved Oil Raccrieq held m Tulsa OK, 21.24 April 1996 This pawr was selected Iof presentabon by an SPE Pmgmm Comm!ftae follmwng rewaw of u4umama contained m an abstract submtied by the author(s) contents of the papr have not been ww..ved by the Socmty cd Petroleum Engmems and are subject to cowectton by the a@vX@) The matOnal, as prasented, does not necessarily reflect 'anypowbon ot the socii of Ps@c4Wm Engineers, m off!cers, or members Papers presented at SPE meedngs are subject~r awsw by EdiWlal COmmtttees Of the SOcmty of Patmfeum Engmeara Perrmsstoñ @41YIS re@rl@d to an abstract of not more than S00 words illustrations may I@ be COpId Tha atsb.ct should contain conspicuous acknov.iedgmnt of where and by '+4mm the papa IS presented Write Llbfanan, SPE, P O Sox SSSSSS, Rwhardson, lX 7SC@3-3SSd, U S A fax 01214-952.9435Abstract Tertiary miscible core flood experiments (Lab A) were initiated to determine residual od saturations (SOJ for a Permian aged limestone reservoir. Cores were collected ffom several wells located in different areas of the reservoir and S-values ranged from 5.3 to 6.6 V0PV. The gas floods were conducted above the MMP with a 50% methane gas. In a separate project, additional floods were conducted on a dilTerent set of cores horn the same reservoir (Lab B). These experiments were conducted above the W with a 65% methane gas and the S-values ranged from 10.4 to 1i .9%PV. These difkrences in displacement et%ciency will significantly impact the economics of a field scale project, The experimental procedures were compared and the differences were not significant. NMR T2 measurements were conducted on both sets of cores to examine the porous microstructure. NMR T2 relaxation curves (from which T2=,T~1,T 22and pore size distributions were derived) were compiled for each core. The TIE relaxation times associated with Lab A samples were much smaller than those associated with Lab B (i.e. Lab A cores had a much smaller mean pore size). The pore size distributions associated with Lab A cores were multimodal while those associated with Lab B cores were unimodal. Therefore, the cores had very different porous microstructure. Most importantly, SOmwas strongly correlated with T21, T22and~' (r =0.98, 0.94 and 0.97, respectively). As pore size increased, SOm increasedThe larger pores were apparently associated with dominate flow paths which resulted in bypassing and less efficient displacement. Secondary gas floods were also conducted and a sunilar relationship was found,

No abstract