Microbial treatments are potentially cost-effective for increasing oil production, even in today's economic conditions. Field applications that use microorganisms can range from single-well to full-scale waterflooding, but both processes require some knowledge of the reservoir conditions. Coordinated laboratory and field testing is needed to develop microbial oil recovery technology further.

Summary. The use of microorganisms to enhance oil recovery has become a technically feasible technology for production from stripper wells (those that produce less than 10 B/D [1.6 m/d]). As a result of microbial growth and the production Of CO and/or chemicals, oil recovery can be effectively increased in certain reservoirs with temperatures and salinities hospitable to microorganisms. Research at the Natl. Inst. for Petroleum and Energy Research has led to development of laboratory facilities for evaluating microbial systems for microbial enhanced oil recovery (MEOR) applications. Results from microbial core studies using Berea sandstone have shown that bacteria can vary greatly in their ability to recover residual oil after waterflooding, giving from 7.5 to 71 % recovery efficiency. The type of core encapsulation did not affect oil recovery. Variations in rock permeability from 134 to 1,920 md indicated that some microorganisms may exhibit a better recovery efficiency in lower-permeability cores. The results also show that gas production by microorganisms is not the only factor affecting oil recovery; some bacteria that produce surfactants but no gas could provide equally efficient additional oil recoveries. Likewise, the type of surfactant also makes a difference; two species that produce surfactants can vary in recovery efficiency. Another contributing mechanism to microbial oil mobilization may be by increasing areal sweep efficiency from microbial growth. A comparison of oil recovery efficiencies between microorganisms using a medium-to-light-gravity oil (Delaware-Childers) and two heavy oils (Wilmington, CA, and Chaffee, CA) indicates that MEOR may be effective for heavy oils as well as light oils. Correlations are presented between the recovery efficiency of a particular microbial species and its ability to mobilize crude oil in etched-glass micromodels. Introduction The most popular technique for MEOR involves the direct injection of microorganisms into the reservoir. The in-situ MEOR processes require growth and metabolism of injected microorganisms, which provide chemicals that can aid in releasing oil from reservoir rock. Several mechanisms for microbial oil recovery have been proposed.Production of gases-CO, H, N, and CH -could increase pressure in the reservoir and reduce oil viscosity. Controlled microbial activity should produce gases in pore spaces, including parts of the reservoir that would normally be bypassed in gas flooding processes.Microbial acid production, primarily of low-molecular-weight fatty acids, can cause rock dissolution.Biosurfactant production can cause decreases in surface and interfacial tensions.Microorganisms have been shown to cause wettability alterations in glass micromodels, reservoir flow cells, and Berea sandstone.Microbial growth and polymer production in high-permeability regions of the reservoir can allow additional injected fluids to bypass those regions and flow through the desired high-oil-saturation areas of the reservoir. Oil displacement tests in Berea cores of MEOR processes were used to evaluate species of Bacillus, Clostridium, and other genera for their abilities to release crude oil from rock and survive in porous media. Such experimental parameters as core encapsulation, crude oil type, rock permeability, and microbial characteristics were varied and the effects on crude oil recovery were investigated. The results from these studies were compared with visual observations in micromodels to determine the mechanisms of oil recovery by microorganisms. Experimental Apparatus and Procedures Coreflood Apparatus. The experimental equipment is shown in Fig. 1. The fluid separators are piston devices used to inject microbial solutions and other fluids into the cores and were designed to prevent corrosive fluids from contacting the pumps. The frontal advance rate for the cores was 1 ft/D [0.3 m/d]. Micromodel. Glass micromodels used in this study are shown in Fig. 2. Construction of the micromodels has previously been described by Chatzis. The flow rate was adjusted to 0.01 ml/min, which corresponded to approximately 8 ft/D [2.4 m/d]. The micromodels were brine-saturated, oil-saturated with Delaware- Childers crude, and waterflooded to residual oil saturation (ROS) before microbial injection. Core Preparation. Blocks of Berea sandstone were obtained and cut in cylindrical cores 10 in. [25 cm] in length and 1 in. [2.5 cm] in diameter. The cores were fired at 800 deg. F [427 deg. C] for 24 hours to stabilize the clays. Cores were either encapsulated in epoxy with inlet and outlet valves at the ends or were encased in rubber sleeves and placed inside stainless-steel Hassler coreholders. The cores were then evacuated and flushed with brine. Darcy's law was used to determine absolute permeability of each core. Crude oil was injected into the cores until no additional water was produced (about 24 hours; pressure drop was less than 1 psi [ less than 6.9 kPa]), then brine was flushed through the core until no more crude oil was produced. The cores thus simulated a waterflooded ROS condition designated by S. Crude Oil. Crude oil samples were obtained from the Bartlesville sand in the Delaware-Childers field in northeastern Oklahoma or from the Wilmington or Chaffee fields in California. Delaware- Childers oil has a gravity of 31 deg. API [0. 87 g/cm ]; Wilmington, 17 deg. API [0.95 g/cm ]; and Chaffee, 14 deg. API [0.97 g/cm ]. Chemicals and Media. The molasses used in these experiments was Mr. Blackstrap 87; its composition was 5 % crude protein; 0. 5 % crude fat; 38 % total sugars; and 56.5 % fiber. It was dissolved in water to make a 4.0 wt% solution and was filtered through cheesecloth to remove the suspended fiber particles. Microorganisms. Microorganisms N12, N17, N18, and Tu6 are Gram-negative, facultatively anaerobic rods that produce CO and acids when fermenting sucrose. Bacillus A and B, Bacillus licheniformis (Bacillus C; ATCC #27811), and Bacillus licheniformis (Bacillus D; ATCC #39307) are all facultatively anaerobic rods that produce acids and varying amounts of surfactant when fermenting sucrose. Clostridium A and B are anaerobic species that produce CO, and acids when fermenting sucrose. System PS1 is a mixture of strictly aerobic microorganisms. SPERE P. 489^

Simulation and experimental results on the transport of microbes and nutrients in one-dimensional coreflooding experiments are presented, and the development of a three-dimensional, three-phase, multiple-component numerical model to describe the microbial transport phenomena in porous media is described. The governing equations in the mathematical model include net flux of microbes by convection and dispersion, decay and growth rates of microbes, Chemotaxis and nutrient consumption, and deposition of microbes on rock grain surfaces. Porosity and permeability reductions due to cell clogging have been considered and the production of gas by microbial metabolism has been incorporated. Governing equations for microbial and nutrient transport are coupled with continuity and flow equations under conditions appropriate for a black oil reservoir. The computer simulator has been used to determine the effects of various transport parameters on microbial transport phenomena. The model can accurately describe the observed transport of microbes, nutrients, and metabolites in coreflooding experiments. Input parameters are determined by matching laboratory experimental results. The model can be used to predict the propagation of microbes and nutrients in a model reservoir and to optimize injection strategies. Optimization of injection strategy results in increased oil recovery due to improvements in sweep efficiency.

Summary To determine the effect of water-soluble polyacrylamide polymer adsorption and flow behavior on oil recovery, relative permeability and mobility were determined from flow experiments at various polymer concentrations. A selective reduction of the relative permeability to water with respect to the relative permeability to oil was observed for both Berea and reservoir sandstone cores. Adsorbed polymer layer increases water wettability. Relative permeability reduction could be attributed to both wettability change and pore-size restriction due to the adsorbed polymer layer. An empirical model was proposed to correlate the relative permeability reduction and the amount of polymer adsorption. Depletion-layer effect results in a reduced polymer viscosity in porous media with respect to bulk solutions. Modification of the existing shear-rate model allows for accurate prediction of this effect. The integration of the new models in UTCHEM provides a more accurate tool for engineering design of polymer applications. Introduction Water-soluble polyacrylamide polymers have been used to reduce water production in oil wells and for mobility control in injection wells for decades.1 One of the attractive properties of polyacrylamides is their ability to reduce the relative permeability to water more than the relative permeability to oil in porous media. From the published field tests on well treatments by polymer adsorption in the 1970's and 1980's,1,2 only a few jobs were considered to be economically successful. Results could not be interpreted due to the lack of detailed information. At present, the importance of laboratory research and simulation study are emphasized for successful field design. The selective permeability reduction by polymer adsorption was traditionally termed as "permeability reduction" or "residual resistance factor," which is equivalent to the endpoint relative permeability. Previous laboratory results1 indicated that the maximum reduction of the endpoint relative permeability to water caused by polymer adsorption can be as high as a factor of 10, while the reduction of the endpoint relative permeability to oil is less than 2. If crosslinking or swelling agents are applied, the maximum reduction of the relative permeability to water could be more than two orders of magnitude. The mechanisms of this selective reduction have been explored by several researchers.3–5 An understanding of these mechanisms could be obtained from the selective permeability reduction by gels.6 Measurement of the residual resistance alone may provide a qualitative estimation. To model the effect of polymer adsorption, however, a measurement of the relative permeability is necessary, especially when residual saturation and the shape of the relative permeability curves change after polymer adsorption. Modification of the relative permeability by polymer adsorption has been intensively studied recently.3-5,7-10 Ali et al.4 and Barrufet and Ali,5 derived the relative permeability from drainage capillary pressure measured by an ultracentrifuge and showed that the reduction of the relative permeability caused by starch-based polymers is dependent on saturation. The reduction was interpreted as a change in lubrication along the pore walls. Direct measurement of relative permeability after polymer adsorption was also seen in Refs. 3 and 7 through 10. In water-wet porous media, it was found that the residual oil saturation remained almost the same after polymer treatment. At residual oil saturation, the quantity of adsorbed polymer per gram of rock was also found to be almost the same as at 100% water saturation, but the endpoint relative permeability reduction to water was increased in the presence of residual oil. Based on a capillary bundle model, a correlation of the relative permeability curves with polymer-layer thickness was proposed by Zaitoun and Kohler.3 However, the relationship of the polymer-layer thickness with the quantity of adsorbed polymer is still unknown, and further modification of the capillary bundle model may also be needed to model complex pore matrices. On the other hand, as polymer propagates through porous media, polymer solution will be diluted in the propagation front due to dispersion and adsorption, and the dilution could extend to the entire slug if the slug size is too small. So far, few researchers have related the variations of the relative permeability curves as a function of polymer concentration or the quantity of adsorbed polymer. Therefore, one of the objectives in the present study is to measure and correlate relative permeability curves as a function of polymer adsorption. Polymer solution mobility was also studied as a function of polymer concentration. Effective viscosity at low shear rate in porous media is lower than that in bulk solution at the same shear rate. The dependence of the depletion-layer effect on polymer concentration as well as porous media will be examined in this paper. Finally, the numerical models of relative permeability and mobility as a function of polymer concentration developed in this study will be incorporated in UTCHEM. Several cases will be studied to compare incremental oil recovery predicted by these new models with that predicted by previous descriptions of polymer behavior in porous media. A simplified layered reservoir model will be used for comparative simulation runs. Polymer flooding and near-wellbore polymer treatments will also be simulated. Results from these simulations should provide guidelines for future field strategies. Experiment Porous Media. Both strongly water-wet and mildly oil-wet cores were chosen to study the influence of wettability on polymer adsorption, two-phase relative permeability, and polymer solution mobility. The mildly oil-wet medium is a Warden reservoir sandstone core from Santa Fe field, Stephens County, Oklahoma. Two strongly water-wet media are Berea sandstone cores with different permeabilities. Table 1 summarizes the petrophysical properties of these sandstone samples. Fluids. Synthetic brines were prepared to represent reservoir brine (produced water) composition and makeup water (injection water) composition used in the Warden reservoir. Produced water has a total dissolved solid (TDS) of 31,300 ppm which contains 29 g/L NaCl, 0.94 g/L CaCl2, 0.77 g/L MgCl2, 0.11 g/L KCl, and 1.1 g/L NaHCO3, and injection water has a TDS of 1,490 ppm which contains 0.343 g/L CaCl2, 0.252 g/L MgCl2, 0.176 g/L Na2SO4, and 0.72 g/L NaHCO3

To determine the effect of water-soluble polyacrylamide polymer adsorption and flow behavior on oil recovery, both steady-state and unsteady-state flow experiments were performed on Berea sandstone and reservoir cores. Berea sandstone core is strongly water-wet while the reservoir core is mildly oil-wet. Relative permeability curves and polymer adsorption measurements were made at residual oil saturation and 100% water saturation for increasing polymer concentrations. Mobility measurements were made at different polymer concentrations and shear rates. A selective reduction of the relative permeability to water with respect to the relative permeability to oil was observed for both Berea and reservoir sandstone cores. The reduction of the relative permeability to water in the presence of oil phase is more than that at 100% water saturation. As polymer concentration increases, polymer adsorption, irreducible water saturation and relative permeability reduction increase. Residual oil saturation remains almost the same. Wettability is beneficial to water-soluble polymer adsorption. In reservoir core, relative permeability reduction could be attributed to both wettability change and pore-size restriction. Polymer adsorption isotherm follows Langmuir's law. Relative permeability reduction as a function of polymer adsorption exhibits an "S-type" curve. It increases exponentially as polymer adsorption increases and eventually approaches a constant. An empirical model was proposed to correlate this characteristics. As predicted by the depletion layer and viscoelasticity theories, flow behavior of polymer solution in porous media is found to be significantly different from that in bulk solutions. A modification was made to the existing shear rate model based on mobility measurements, showing that the depletion layer effect is in direct proportion to polymer concentration, and is more significant in the reservoir core than in Berea core. New models were incorporated in UTCHEM, a chemical flood simulator developed by the University of Texas. These models can provide more accurate prediction of the combined effects of relative permeability reduction and viscosity over existing models. Case studies show that the long-lasting relative permeability reduction by polymer adsorption is likely to maximize the benefits of polymer solution in polymer flooding. Increased understanding through simulation may lead to improved field profile modification and near wellbore treatments. The effects of cross-flow and the degree of relative permeability reduction are two critical factors to determine polymer placement strategies and success. P. 293