The majority of bitumen and extra heavy oil is produced by steam-based recovery processes yet these methods are energy intensive and emit large amounts of greenhouse gas to the atmosphere. Not only is the emissions intensity high due to steam generation, but also the water handling and treating facilities required for these recovery methods is expensive both to purchase and install but also to operate. A lot of focus of research has been on reduced steam processes, such as thermal-solvent techniques, as well as in situ combustion technologies such as Air Injection. Here, air injection is evaluated as a follow-up process to Cyclic Steam Stimulation in a deep thick heavy oil reservoir. The reservoir simulation model is obtained from a history-match to existing cyclic steam stimulation (CSS) field data. The results demonstrate that an additional 33% oil recovery is reached by using an air injection follow-up process. This gives a total recovery factor equal to about 55%. Based on incremental cumulative energy-to-oil ratio, the results suggest that air injection follow-up processes should be considered for post-CSS operations but that further improvement of the energy intensity is needed.
With increasing world demand for energy, greater attention has been placed on the exploitation of the huge existing resources of heavy oil and bitumen. Although thermal in-situ recovery methods such as steam assisted gravity drainage (SAGD) have been very successful in exploiting such resources, the thermal efficiency of SAGD, its greenhouse gas emissions and water requirements remain major concerns.
Co-injection of solvent with steam shows promise for enhancing oil rates as well as reducing energy and water consumption with correspondingly lower environmental impacts. In hybrid steam-solvent methods, there is a balance between the solubility of the solvent and its ability to reduce bitumen viscosity. Proper selection of the solvent for the reservoir operating conditions is key for optimizing process efficiency and maximizing performance improvement over the steam-only method.
Convective mixing at the edge of the steam chamber enhances heat and mass transfer rates which increases oil mobility and production rate. In this study, the convective mixing at the steam-bitumen interface is examined using theoretical stability analysis of the thermal-solvent boundary layer. Several alkane solvents were compared based on the time required for the onset of the buoyancy-driven instabilities in the system. The results show that there is a higher degree of convective mixing for some intermediate solvents, which is in agreement with reported laboratory and simulation results. The onset of convective mixing and the wavelength of the instabilities are obtained as a function of reservoir and fluid properties for various solvents.
These results can aid in the screening and selection of appropriate solvent additives to steam for a given reservoir and bitumen properties; also this analysis can be applied for mixtures of solvents to optimize the overall efficiency of the steam-solvent recovery method.
The analysis of fluid transport in fractured reservoirs is of great concern in petroleum and environmental engineering. The objective of this work is to study mass transfer between the matrix and the fracture network in such complex formations. In this study, the impact of adsorption on mass transfer in a fractured medium with variable fracture spacing is investigated.
Development of a mathematical model for mass transfer in dual porosity systems enhances predictions of the rate of mass transfer between matrix and fracture. In addition, it provides a tool for an appropriate design of advanced oil and gas recovery processes. Mass transfer is modeled based on the advective-dispersive transport with adsorption in an infinite acting dual porosity reservoir under radially divergent and continuous injection. The fracture spacing has been taken into account by including the variable rock matrix block size distribution in the developed model.
By employing this model, the effects of the adsorption rate on the mass transport in a subsurface environment are analyzed. The impact of the rate of adsorption on the accumulation of the injected tracer (catalyst) in a reservoir is investigated.
An understanding of the effect of adsorption on advective-dispersive mass transport with variable fracture intensity can be a key finding to develop and design advanced oil recovery processes.
A dual porosity model was introduced by Warren and Root (1963) based on the findings of Barenblatt et al. (1960). These models have been employed to define the fluid transport in fractured rocks. The dual porosity models have been used to handle the complexity of the fluid flow and transport in naturally fractured reservoirs. These models have been developed in the past to enhance the predictions of field scale mass transfer in fractured and heterogeneous formations (Aguilera, 1995; Haggerty and Gorelick, 1995; Jardine et al., 1999; Neretnieks, 1980). Based on the laboratory and field studies, analytical models were developed by many investigators (Rasmusen and Neretinieks, 1980; 1981; Sudicky and McLaren, 1992; Tang et al., 1981). In this study a model which describes the effect of adsorption on mass transfer in dual porosity media is developed. This model is then used to obtain a better understanding of the mass transfer between rock matrix blocks and fracture networks. This study can be employed for the design of ultra dispersed catalyst injection for in situ upgrading of heavy oil and subsurface mass transport in fractured reservoirs. The role of adsorption of the injected tracer or catalyst on the mass transfer in matrix and fracture is studied. Results show their importance in the mass transport processes.
A divergent radial flow system in a fractured reservoir with impermeable top and bottom boundaries is considered. Planar shape rock matrix blocks are assumed to exist in the reservoir. In this study we consider the flow of a single incompressible fluid. The physical properties of the rock and fluid are assumed to be constant. We further assume continuous injection through the total pay zone. The mass transfer in the rock matrix blocks is assumed to be diffusion dominated while in the fracture both advection and dispersion contribute to the mass transfer. The rate of adsorption in the fracture and matrix is represented by the retardation factors, which will be described in the next section. Figure 1 shows the schematic of the system used in this work.
Expanding Solvent-Steam Assisted Gravity Drainage (ES-SAGD) was invented to enhance SAGD performance by reducing energy use while increasing oil production rates and recovery factor. ES-SAGD involves co-injection of solvent and steam. The majority of energy losses occur between the steam generator and sandface and at the top of the depletion chamber (to the overburden). ES-SAGD performance improvement is traditionally ascribed to oil phase dilution which in turn leads to oil phase viscosity reduction. However, the amounts of solvent added to the process are typically very small (< 5-6% by volume) thus it remains unclear how the solvent can lead to significant lowering of the steam-to-oil ratio (~25-50%) and large enhancements of the oil rate (~25 to 100%). Here, we report on how SAGD and ES-SAGD (hexane, heptane and octane solvents) can potentially perform in the presence of in-situ emulsification at steam chamber edge. We present a numerical approach which allows incorporation of emulsion modeling into SAGD and ES-SAGD simulations with commercial reservoir simulators via a two-stage pseudo chemical reaction. Numerical simulation results show excellent agreement with experimental data for low-pressure SAGD and ES-SAGD. Accounting for viscosity alteration, multiphase effect and enthalpy of emulsification appear sufficient for effective representation of in-situ emulsion physics during SAGD and ES-SAGD in very high permeability systems. Results also show that, in-situ emulsification may play a vital role within the reservoir during SAGD; increasing bitumen mobility thereby decreasing cSOR. It was concluded that traditional approach to numerical ES-SAGD simulation can significantly over-predict incremental oil recovery. Results from this work extend understanding of ES-SAGD by examining its performance improvement over traditional SAGD in terms of multiphase behavior at the edge of the chamber, thermal efficiency and incremental recovery. Results reveal that dynamics at the edge of the chamber is more complex than simple solvent dilution model.
Cold Heavy Oil Production with Sand (CHOPS) is a non-thermal heavy oil recovery technique used primarily in the heavy oil belt in western Alberta and eastern Saskatchewan. Under CHOPS, typical recovery factors are between 5 and 15% with average ~10%. This leaves ~90% of the oil in the ground after the process becomes uneconomic. CHOPS exhibits an enhancement in production rates compared to conventional primary production, which is explained by formation of high permeability channels known as wormholes. The formation of wormholes has been demonstrated to occur in both laboratory experiments and field tracer studies. The ability to model growth of wormholes does not currently exist in commercial reservoir simulators. Here, wormholes are modelled as multi-lateral wells, which grow dynamically in the reservoir, using existing wellbore features. A module was coupled to CMG STARSTM to dynamically grow wormholes in the reservoir taking foamy oil flow, sand failure, and sand production into account. Here, we present on the results of history matches against field data to tune model parameters. The history-matched model reasonably predicts production trends of field CHOPS operations. The results provide a methodology to model CHOPS and predict under uncertainty where the wormholes will tend to grow into the reservoir. This provides a tool for placing new wells in the reservoir that will most likely not be in direct contact with existing wormholes. Multiple realizations of the reservoir can be used to mark the region of the reservoir that undergoes wormhole formation. The model can then be used for follow-up EOR processes such as cycle solvent injection as well as field scale optimization.
With the decline of conventional oil production, developing and producing heavy oil resources efficiently is becoming more important. The Liaohe Heavy Oil Field steam operation is unique - it started with cyclic steam stimulation (CSS) operation that transitioned into a continuous steam-assisted gravity drainage (SAGD) operation. With respect to oil production in China, this field is considered critical for heavy oil production and technology development. Cyclic steam injection was initially done through vertical wells. This had the benefit that it provided a good start-up of depletion chambers in the reservoir. These chambers then grew under gravity drainage after continuous steam injection (through the vertical wells) and continuous production through a set of horizontal wells was started. Controlled and deliberate transition from CSS to a gravity drainage process with the objective of optimizing energy intensity (GJ injected per unit volume oil produced) with control enabled through production and thermocouple data is a smart field operation which we refer to as a Reservoir Production Machine (RPM). In this paper, as a first step to understand the operation and its impact on the reservoir, we have history matched the CSS operation based on the injection and production data from field. The use of vertical steam injection wells (formerly the CSS wells) in combination with horizontal production wells operated in a SAGD mode of operation is explored. The history-matched model can be used to develop automated RPM technologies to optimize not only energy intensity but also emissions intensity.
The Steam and Gas Push (SAGP) process was developed to improve the thermal efficiency of SAGD process. In SAGP, non-condensable gas is co-injected with steam into the reservoir. Ideally, the non-condensable gas accumulates at the top of the reservoir and provides insulation which reduces heat losses to the overburden. This means that lower SOR can be achieved at the same recovery factor. It remains unclear how energy is distributed and transformed within the chamber and its edges when non-condensable gas is added to the injected steam. In this work, we compare conduction and convection at edge of the steam chamber during SAGD and SAGP. The results show that both oil production rate and cumulative oil are reduced in SAGP compared to SAGD when 0.8 mole% NCG is co-injected with steam. This is because the injected NCG accumulates at the upper part of the leading edge of the steam chamber and slows down the growth of the steam chamber in that area, which results in lower cSOR but with a reduction of recovery factor. If 0.8 mole% NCG is co-injected at later periods of the operation, lower cSOR results without a significant reduction of oil production rates and cumulative oil production. In this case, the injected NCG migrates directly to the upper part of the reservoir and accumulates at the side edge of the steam chamber, since the steam chamber had already grown to the top of the reservoir. The added gas slows down lateral growth of the steam chamber in the upper part of the reservoir and forces steam chamber growth in the downward direction. From an analysis of energy transport in SAGP and SAGD operations, the results reveal the optimal timing for the onset of NCG co-injection with steam.
It has been observed that the shale gas production modeled with conventional simulators/models is much lower than actually observed field data. Generally reservoir and/or stimulated reservoir volume (SRV) parameters are modified (without much physical support) to match production data. One of the important parameters controlling flow is the effective permeability of the intact shale. In this project we aim to model flow in shale nano pores by capturing the physics behind the actual process. For the flow dynamics, in addition to Darcy flow, the effects of slippage at the boundary of pores and Knudsen diffusion have been included. For the gas source, the compressed gas stored in pore spaces, gas adsorbed at pore walls and gas diffusing from the kerogen have been considered. To imitate the actual scenario, real gas has been considered to model the flow. Partial differential equations were derived capturing the physics and finite difference method was used to solve the coupled differential equations numerically. The contribution of Knudsen diffusion and gas slippage, gas desorption and gas diffusion from kerogen to total production was studied in detail. It was seen that including the additional physics causes significant differences in pressure gradients and increases cumulative production. We conclude that the above effects should be considered while modeling and making production forecasts for shale gas reservoirs.
The permeability of unconventional gas/oil reservoirs is a critical control on economic viability of unconventional plays, yet its determination, particularly in ultra-low permeability shale reservoirs, remains a challenge. Some of the difficulties in obtaining accurate permeability measurements in the lab include: recreating in-situ stress and fluid saturation conditions; establishing the appropriate sample size for measurement; correcting for sorption of gases on kerogen and clays; accounting for non-Darcy flow (slippage and diffusion); among many others. Unsteady-state measurements are most popular for establishing permeability in ultra-tight rock; both pressure-decay and pulse-decay decay techniques have been used. Analysis methods for these techniques have been established, but there remain some questions about whether these analysis methods are optimal for establishing permeability.
In this work, we investigate the use of pressure- and rate-transient analysis (PTA/RTA) methods to analyze data obtained from a new core plug analysis procedure, designed specifically to extract information (permeability and pore volume) from ultra-low permeability reservoir samples (core plugs). The new analysis procedure calls for analyzing the rate and/or pressure data analogously to larger-scale well-test/production data. During a core plug production test for example, derivative analysis of rate-normalized pseudo-pressure change is first analyzed to determine flow-regimes. For homogenous samples, linear flow is followed by boundary-dominated flow; for this scenario, permeability can be established by noting the end of linear flow and using the distance of investigation calculation to calculate permeability (knowing core length). Permeability can also be established independently from a linear flow (square-root of time) plot. Pore volume can also be established. Analytical simulation is used to verify estimates of permeability and pore volume from RTA/PTA. Our solutions allow complex unconventional gas reservoir behavior to be incorporated, including corrections for adsorbed gas and non-Darcy flow. Our new methodology is tested using various simulated cases which differ due to: 1) reservoir type (single or dual porosity, homogenous or heterogeneous); 2) matrix permeability; 3) analysis type (post injection/falloff production test, or post-injection falloff); 4) adsorption (compressed gas storage only or compressed + adsorbed gas storage); Darcy or non-Darcy flow. In all cases, reasonable estimates of permeability and pore volume were obtained, provided the appropriate corrections are made.
We believe this new technique for analyzing core data, and the proposed core testing procedure, will considerably improve on current techniques for establishing permeability and pore volume of unconventional reservoir samples.
Reservoir simulation for a full field heterogeneous model with millions of grid blocks demands significant computational time so improving the computational efficiency becomes crucial in designing a reservoir simulator. Graphics Processing Unit (GPU), a new high-profile parallel processor with hundreds of microprocessors, stands out in parallel simulation because of its efficient power utilization and high computational efficiency. Also, its cost is relatively low, making large-scale parallel reservoir simulation possible for most of desktop users.
In this paper several GPU-based parallel preconditoners, in conjunction with a new GPU-based GMRES algorithm, are proposed and coupled with an in-house black-oil simulator to speedup reservoir simulation. In particular, massively parallel ILU preconditioners (ILU(0), ILUT, block ILU(0), block ILUT), which are usually regarded as data dependence and highly sequential preconditioners, are developed on GPUs.
In the numerical experiments performed, the SPE 10 problem, a 3D heterogeneous benchmark model with over one million grid blocks, is selected to test the speedup of our GPU solver and preconditioners. On the state-of-the-art CPU and GPU platform, the new GPU implementation can achieve a speedup of over eight times in solving linear systems arising from this SPE 10 problem compared with the CPU based serial solver. Moreover, our GPU solver is successfully coupled with the in-house black-oil simulator to test the performance of the whole parallel simulation process, with a speedup of about six times. The excellent speedup and accurate results demonstrate that the GPU-based parallel linear solver and preconditioners have the great potential in parallel reservoir simulation.