No abstract

We have completed an exploratory economic analysis based upon systematic compositional simulations of CO2 EOR to establish how much and what type of economic incentive might be needed to encourage oil companies in the U.S. to store CO2 in oil reservoirs. The economic analysis took into account factors such as capture and transportation costs. In order to quantify the effect of flood performance, various simulations were performed by employing different reservoir types, well spacing, and injection schemes. Experimental design and method of response surfaces along with Monte Carlo simulations were utilized to perform this study in a systematic, efficient and accurate manner. Combinations of reservoir parameters and economic factors were studied to achieve comprehensive understanding of the financial performance of coupled CO2 sequestration and EOR projects. Possible CO2 credits were also quantified in a probability based distribution functions considering various uncertain economic and geologic characteristics in different projects. Introduction CO2 concentration in the atmosphere has drastically increased from 280 ppm during pre-industrial age to its current level of 380 ppm (Bryant, 1997). It is proven that this is mainly due to the dramatic increase in the fossil fuel consumption. This has caused climate change concerns among environmentalists and it is gaining more publicity as international agencies and governmental sectors in different countries seriously considering CO2 reduction policies implemented. It should be mentioned that there is no direct proven evidence showing the relation between climate change and the CO2 emissions. However, due to the greenhouse effect of CO2, it is mainly suspected that a higher CO2 concentration in the atmosphere has caused these climate changes. Geological CO2 storage as the only effective option to mitigate atmospheric CO2 emissions has been considered since the 1990's and has been implemented in large scale for the first time in Norway. Based on the data published by Moritis (2002) over 35 million tones of CO2 have been injected into the oil reservoirs for the purpose of EOR, and currently few aquifer CO2 storage projects are underway. Weyburn CO2 sequestration and EOR project is the only on-going commercial coupled EOR and sequestration project which has shown great success in terms of both objectives of the project (Malik and Islam, 2000). Carbon dioxide is transported from a North Dakota coal-gasification plant through pipelines and is injected into the to Weyburn oil field. One of the main aspects of all current EOR projects in the United States is the use of inexpensive CO2 from natural resources. These sources have high CO2 purity and there are no additional costs for capture and compression of their stream. There are also limited amounts of anthropogenic CO2 available from fertilizer, petrochemical, and coal-gasification plants which are much more expansive than natural sources of CO2. Due to the recent high oil prices and assuming it will continue in similar fashion in the future, CO2 flooding projects are expected to rapidly grow in numbers and volume in the next decade. Therefore, there will be serious need for additional CO2 sources. From another perspective, carbon emission regulations have already been set in place in some European countries as well as Japan under Kyoto protocol. If the regulations are fully implemented in the industrialized countries such as United States, it can serve as double-purpose for both providing huge additional CO2 sources for EOR processes and vast potential for geological storage of anthropogenic CO2 emissions. Among all CO2 emission sources, stationary sources such as power stations and petroleum industry facilities are main contributors.

The effect of relative permeability hysteresis on both CO2 storage and oil recovery has been studied using a compositional simulator. The effects of process parameters such as water-alternating-gas (WAG) ratio and CO2 slug size, and reservoir heterogeneity characteristics such as Dykstra-Parson coefficient and correlation lengths on CO2 storage and tertiary oil recovery were simulated using hysteresis based upon existing correlations. Three different relative permeability and capillary pressure models for three different rock types in the reservoir were carefully constructed. Reservoir fluid PVT data were used to develop the Equation-of-State (EOS) model. A grid refinement study was performed to evaluate the numerical convergence behavior of the simulation model with the hysteresis option included. In the refined cases, it was necessary to apply a higher-order approximation scheme to reduce the numerical dispersion of the simulations. In addition, due to the application of very small gridblock sizes, physical dispersion was also taken into account. Experimental design and statistical analysis were used to understand the most influential factors on oil recovery, project net cash flow, and CO2 storage. Optimization was carried out and response surfaces were constructed to quantify the effect of each parameter. Introduction The effect of relative permeability hysteresis on the geological storage of CO2 in saline aquifers has been studied in recent years (Kumar et al., 2004; Ozah et al., 2005; Spiteri et al., 2005). Only two fluid phases (gas and brine) are needed to describe the injection of CO2 in aquifers. In coupled CO2 sequestration and Enhanced Oil Recovery (EOR) processes, the degree of complexity is higher due to the nature of multiphase flow of a multi-contact miscible displacement in an oil reservoir. Depending on temperature and pressure, three or more fluid phases (Guler et al. 2001) are present in the reservoir during CO2 injection. Compositional simulation is needed to account for the phase behavior effects that occur when CO2 is injected into the oil reservoir. Compositional simulation of WAG injection for EOR purposes, with and without hysteresis-included, has been shown to predict different results in 2-D and 3-D cases (Christensen et al. 2000). In addition, Christensen et al. (1998) have shown that simulations of this process can have considerable compositional effects, therefore applying compositional simulation will give more accurate results than using black oil simulations. Moreover, when the fluid saturations experience cyclic changes, relative permeabilities and capillary pressure data show hysteresis behavior. Hysteresis is defined as path-dependent relative permeability and capillary pressure curves during drainage and imbibition cycles. The imbibition oil and gas relative permeability curves are generally lower than the drainage curves at the same saturation. But the imbibition water relative permeability curve is slightly greater than the drainage curve. Hysteresis is greatest for the gas phase and most important for WAG processes. For coupled EOR and geological storage of CO2, one of the key issues is the effect and importance of the trapping and hysteresis on the amount of CO2 remaining (stored) in the reservoir. If gas remains in the reservoir in the form of trapped gas, the risk of gas migration and its escape from the reservoir will be minimized. Some studies have been performed to investigate the effect of residual gas saturation on the amount of stored CO2 in saline aquifers using compositional simulators (Kumar et al., 2004; Ozah et al., 2005) and black oil simulators (Spiteri et al., 2005). It should be noted that the degree of complexity of fluid flow in porous media in the studies which are involved in the coupled CO2-EOR and sequestration is much higher than in the aquifer storage cases; therefore careful selection of relative permeability models and their associated parameters have an important role in the final results.

A compositional reservoir simulation study was performed to investigate enhanced oil recovery and sequestration of carbon dioxide. Maximizing profit from oil recovery and maximizing the amount of carbon dioxide stored in the reservoir are competing goals and both will be important in the future. Both depend on large number of parameters and the strategy used to flood the reservoir. A very large number of simulations are required to understand and evaluate different strategies for each reservoir and for each realization of the properties of a particular reservoir. In this study the effects of variety of flood design variables on both EOR and sequestration objectives were investigated in sandstone and carbonate reservoirs separately. Experimental design and the method of response surfaces were used as tools to perform this study in a systematic and efficient way. Design parameters such as well spacing, different injection and production schemes, various well control techniques, and different mobility control methods were selected for study. By applying fractional factorial design and D-optimal methods, simulation cases were selected to study the effect of the parameters for each scenario. The amount of CO2 stored at the end of the oil recovery process and the net present value of the each sensitivity case were considered as the two decision criteria. Economic analysis for all of these cases were performed based on the necessity to account for CO2 storage factors such as capture and transportation costs, and possible CO2 tax credits for storage. Response surface analysis was utilized to determine the best strategy based upon these decision criteria for different type of the reservoirs. The result using this approach was similar to the result from an exhaustive simulation study, but took much less computation time and effort. An approach that is both realistic and feasible, such as the one used in this study, will be needed for future simulation studies because of the increasing importance of CO2 geological storage, the extremely wide variety of reservoir conditions of potential interests, the need to understand and reduce uncertainties, the need to find better operational strategies, and the uncertainty in future economic and regulatory incentives.

Correlations are commonly used to predict CO2 multiple contact miscibility (MMP) since such correlations are generally inexpensive and easy to use. In this study, we used a novel approach based upon four dimensionless scaling groups commonly used for hydrocarbon phase behavior modeling (reduced temperature and acentric factors for light and heavy pseudo components) as well as multivariate regression analysis based on response surface methodology to develop an MMP correlation for a broad range of reservoir oils. Applying the response surface method and multivariate regression analysis made it possible to quantify and rank the effect of each one of the mentioned dimensionless groups on the predicted MMP value. Since reservoir temperature is one of the main parameters, slim tube simulations were performed at four different reservoir temperatures (90, 150, 180, and 220 °F) for all of the fluid models. Based on the results from these simulations, and by performing multi-variate regression analysis, MMP values were correlated using a response surface based on linear, quadratic, and third degree equations. Our new MMP correlation takes into account the important equation-of-state properties for heavy- and light-oil components as well as temperature. Predicted MMP values from the new correlation were compared with previously published MMP correlations and found to have a lower average error. Introduction Miscible CO2 flooding is one of the most efficient displacement processes among tertiary oil recovery methods. Based on a study by Stosur et al. (1990), on future potential of Enhanced Oil Recovery (EOR) methods in the US, miscible CO2 gas injection is gaining more popularity and eventually will be more attractive than any other EOR techniques. This can be related to higher oil prices as well as availability of more CO2 sources considering the global regulations and restrictions on CO2 emissions. In a CO2 displacement project, when full miscibility between injected CO2 and reservoir fluid is reached, capillary forces are eliminated from displacement process which ideally results in no oil trapping and consequently higher recovery values. Difference between reservoir pressure (or displacement pressure) and Minimum Miscibility Pressure (MMP) is the most important factor to determine whether miscibility has been achieved in the reservoir. Displacement pressure should be higher than MMP in order to achieve miscibility in the reservoir. Miscibility in reservoir conditions is generally achieved by two different mechanisms. When the injected fluid and reservoir crude become fully miscible, or in other words, First Contact Miscible (FCM) conditions are achieved, a single phase fluid is created and therefore, injected fluid completely displaces the reservoir fluid. One of the most obvious cases for this type of miscibility condition is FCM of Butane with some crude oils at reservoir conditions. Another type of miscibility mechanism is called multiple contact miscible displacement. Carbon dioxide generally makes multiple contact miscibility with crude oils at some reservoir conditions. This means that many contacts are necessary (in the form of mass transfer) for crude oil components and CO2 to be mixed with each other. In these contacts, CO2 first starts to be condensed into the reservoir oil, and then light oil components are vaporized into the CO2-rich phases. This continues until there is no interfacial tension between these two new phases and a single hydrocarbon phase is being produced. This process mainly depends on reservoir pressure since reservoir temperature is considered constant in the CO2 flooding processes. As reservoir pressure increases, more CO2 is dissolved in the oil and more oil components are vaporized by oil. It is known that the extraction of hydrocarbons depends greatly on the density of CO2 (Lake, 1989). As CO2 density increases, more hydrocarbon components are vaporized from crude. In general, higher reservoir pressure results in higher CO2 density. The pressure at which reservoir oil and CO2 are in extremely close contact is called MMP.