Numerous steam-assisted gravity-drainage (SAGD) optimization studies published in the literature combined numerical simulation with graphical or analytical techniques for design and performance evaluation. Efforts have integrated the simulation exercise with global optimization algorithms. Some studies focused on optimization of cumulative steam/oil ratio (cSOR) in SAGD by altering steam-injection rates, while others focused on optimization of net cumulative energy/oil ratio (cEOR) in solvent-additive SAGD by altering injection pressures and fraction of solvent in the injection stream. Several studies also considered total project net-present-value (NPV) calculation by changing total project area, capital-cost intensities, solvent prices, and risk factors to determine the well spacing and drilling schedule. Optimization techniques commonly used in those studies were scattered search, simulated annealing, and genetic algorithm (GA). However, applications of hybrid GA were rarely found.
In this paper, we focused on optimization of solvent-assisted SAGD using various GA implementations. In our models, hexane was selected to be coinjected with steam. The objective function, defined on the basis of cSOR and recovery factor, was optimized by changing injection pressures, production pressures, and injected solvent/steam ratio. Techniques, including orthogonal arrays (OA) for experimental design (e.g., Taguchi's arrays) and proxy models for objective-function (F) evaluations, were incorporated with the GA method to improve computational and convergence efficiency. Results from these hybrid approaches revealed that an optimized solution could be achieved with less central-processing-unit time (e.g., fewer number of iterations) compared with the conventional GA method. Sensitivity analysis was also conducted on the choice of proxy model to study the robustness of the proposed methods.
To investigate the effects of heterogeneity in the design process, optimization of solvent-assisted SAGD was performed on various synthetic heterogeneous reservoir models of porosity, permeability, and shale distributions. Our results highlight the potential application of the proposed techniques in other solvent-enhanced heavy-oil-recovery processes.
Our studies of the underlying fundamental gas-recovery mechanisms from shale gas are motivated by expectations of the increasing role of shale gas in national energy portfolios worldwide. We use pore-scale analysis of reservoir shale samples to identify critical parameters to be employed in a gas-flow model used to evaluate well-production data. We exploit a number of 3D-imaging technologies to study the complexity of shale pore structure: from low-resolution X-ray computed tomography (CT) to focused ion beam and scanning electron microscopy (FIB/SEM). We observe that heterogeneity is present at all scales. The CT data show fractures, thin layers, and density heterogeneity. The nanometer-scale-resolution FIB/SEM images show that various mineral inclusions, clays, and organic matter are dispersed within a volume of few-hundred µm3. Samples from different regions differ sharply in the shape, size, and distribution of pores, solid grains, and the presence of organic matter. Although the samples have clearly distinguishable signatures related to the regions of origin, extremely low permeability is a common feature. This and other pore-scale observations suggest a bounded-stimulated-domain model of a horizontal well within fractured shale that accounts for both compression and adsorption gas storage. Using the method of integral relations, we obtain an analytical formula approximating the solution to the pseudopressure diffusion equation. This formula makes fast and simple evaluation of well production possible without resorting to complex computations. It ss a decline curve, which predicts two stages of production. During the early stage, the production rate declines with the reciprocal of the square root of time, whereas later, the rate declines exponentially. The model has been verified by successfully matching monthly production data from a number of shale-gas wells collected over several years of operation. With appropriate scaling, the data from multiple wells collapse on a single type curve. Pore-scale image analysis and the mesoscale model suggest a dimensionless adsorption-storage factor (ASF) to characterize the relative contributions of compression and adsorption gas storage.
The critical gas velocity and flow rate for unloading liquids from a gas well has been the subject of much interest, especially in old gas-producing fields with declining reservoir pressures. For low-pressure gas wells, Turner's model (also called Coleman's model) is judged as more suitable for predicting liquid loading in gas wells. However, field practice proves that there are still a number of low-pressure gas wells producing without loadup when the production rate is lower than Turner's minimum production rate.
On the basis of experimental results, a new approach for calculating the critical gas-flow rate is introduced in this paper, which adopts Li's basic concepts, while taking into account the impact of the changes of gas-lifting efficiency caused by the rollover of droplets in the process of rising. A dimensionless parameter, loss factor S, is introduced in the new model to characterize the extent of the loss of gas energy.
Well data from Coleman's paper (Coleman et al. 1991) were used in this paper for validation of the new model. The predicted results from the new model are better than those from Li's model, and even better than Turner's model. The new model is simple and can be evaluated at the wellhead when the pressure is less than 500 psia and the liquid/gas ratios range from 1 to 130 bbl/MMscf, which is suggested by Turner et al. (1969) to ensure a mist flow in gas wells.
This paper presents a simplified method of production forecasting for tight/shale-gas reservoirs exhibiting extended periods of linear flow, without the use of complex tools (e.g., analytical models or numerical models). The method, which is applicable to hydraulically fractured vertical wells and multifractured horizontal wells, is simple because it relies principally on a plot of inverse rate vs. square root of time, and it is rigorous in that it is based on the theory of linear flow and combines the transient linear-flow period with hyperbolic decline during boundary-dominated flow.
The dominant flow regime observed in most tight/shale-gas wells is linear flow, which may continue for several years. This linear flow will be followed by boundary-dominated flow at later times. Therefore, the method proposed in this study is applicable for forecasting production data for these wells because it considers these two important flow regimes. The derivation is presented for a hydraulically fractured well, and this simplified method can be applied both to hydraulically fractured vertical wells and to horizontal wells with multiple fractures. The application of this method to multifractured horizontal wells in the Marcellus and Barnett shale gas is also presented.
The method is validated by comparing its results with test cases, which are built using numerical simulation for hydraulically fractured vertical wells. For each case, only the first year of the synthetic production data is then used for the analysis. It is found that there is reasonable agreement between the forecast rates obtained using this method and the numerically simulated rates.
Currently, analysis techniques using material-balance time are being used in industry to analyze tight/shale-gas reservoirs. Because material-balance time is actually boundary-dominated flow superposition time, these analyses may show symptoms of boundary-dominated flow even though the reservoir is still in transient flow. The advantages of the forecasting method proposed in this study are that: (1) it is not biased toward any flow regimes because no superposition time functions are used; (2) reliable forecasts can be obtained without using pseudotime--this is an advantage because using pseudotime introduces complexities and an iterative procedure; and (3) the only major unknown is the drainage area.
Focusing on one branch of heavy industry (deep oil and gas drilling) in our work, we come to the importance of energy efficiency, or rather to the problem of energy-efficiency increase. In this paper, some major issues in the field have been highlighted, including the energy potentials in drilling rigs and a proposal on how to increase energy efficiency by applying cogeneration to diesel engines within the scope of the contribution to rational use of energy potentials.
Solvent-vapour extraction (SVX) processes offer an attractive alternative to thermal recovery processes by being less energy intensive and are more suitable for thinner, partially depleted reservoirs. A typical SVX process uses solvent injection to dilute the heavy oil by reducing its viscosity, allowing it to be mobilized for production. During this process, the injection of hydrocarbon solvents results in partial deasphalting of the heavy oil, thus reducing its viscosity and enhancing the process performance further.
This work examined the formation and growth of solvent chambers in laterally and vertically spaced horizontal injector/producer well pairs in porous media with five different permeabilities and three different solvent-vapour qualities. Consolidation of the porous media caused by asphaltene precipitation was also analyzed. Thermal-imaging and model excavation studies were performed to investigate the formation and growth of solvent chambers for seven different experiments conducted on a large 3D-physical-model apparatus.
The important findings from this study are as follows: During solvent injection, one or more solvent fingers develop between the injector and producer. The dominant solvent finger becomes a conduit that grows into a solvent chamber connected to the injection well in the upper portion of the reservoir, and develops into an oil-drainage conduit connected to the production well in the lower portion of the reservoir. Solvent dispersion layers are located on the margins of both the solvent chambers and the oil-drainage conduits. The location and development of these nonuniform solvent chambers and oil-drainage conduits are unpredictable, and the oil-drainage conduits do not grow significantly in diameter once connected to the production wellbore, limiting the wellbore inflow efficiency and conformity. Asphaltene precipitation and migration can aggravate this inflow problem, reducing the SVX process performance further.
SVX performance can be improved by increasing the number and diameter of oil-drainage connections between the solvent chamber and the production well, and by controlling the oil deasphalting process. This can be performed by optimizing injection- and production-wellbore geometries, and by optimizing solvent-injection rates and vapour quality.
Previously, nuclear-magnetic-resonance (NMR) carbon-type-analysis data were used to develop a mathematical model of mild thermal conversion (visbreaking) of Athabasca bitumen (Chan et al. 2006). In that work, the major reaction pathways followed during visbreaking were identified. This approach is being extended in the current work to model the visbreaking behaviour of five different oils from different geographical locations around the world. This paper shows the correlation of residue conversion with the contents of different carbon types for five heavy oils from four continents.
During visbreaking runs, operators intend to maximize process yields. This is achieved through increasing process severity by raising temperature. However, if the temperature is too high, coke forms. This maximum temperature varies with different crude oils; therefore, as refinery feedstock composition changes, so does the onset of coking temperature. Coke is a hydrocarbon material that has low hydrogen content and is insoluble in the oil. Consequently, this precipitates in the reactor, eventually causing an unscheduled unit shutdown. We have found that contents of specific carbon types in the feed oils correlate with coke formation. This correlation allows prediction of the quantities of coke that will form under the chosen visbreaking (mild thermal) conditions and the "maximum" quantities of coke that would form under coking (severe thermal) conditions.
The unusually high primary recovery factors (RFs) observed in numerous heavy-oil reservoirs are often attributed to foamy oil flow (i.e., the non-Darcy flow involving formation and flow of gas-in-oil dispersion). It occurs when the wells are produced aggressively at high drawdown pressures that led to conditions in which the viscous forces become sufficiently strong to overcome the capillary forces in pushing dispersed bubbles through pore throats. The role of gravitational forces in generating such dispersed flow has not been studied adequately. This work was intended to evaluate the contribution of gravitational forces in primary depletion of heavy-oil formations under foamy flow conditions.
Primary-depletion tests were conducted in a 200-cm-long sandpack that was held in either horizontal or vertical orientation. The results of horizontal depletion tests were compared with the depletion tests conducted with the sandpack in the vertical direction. Vertical depletions showed better recoveries at slower depletion rates compared with horizontal depletions.
The RFs of both horizontal and vertical depletions were correlated against the average drawdown pressure available to move the oil. It was found that the RF shows a strong dependence on the average drawdown pressure. It was also found that the curve of RF vs. average drawdown pressure moves slightly toward higher recoveries in the presence of an added foaming agent (i.e., with increased oil foaminess).
The solvent-aided process (SAP) is a solvent-based enhancement of steam-assisted gravity drainage (SAGD) in which small amounts of solvent, such as light alkanes or natural-gas liquids, are added to the injected steam to enhance reservoir performance and associated project economics. Expectedly, the economics with SAP are sensitive to the solvent recovered from the reservoir, making its measurement in a field test an important factor.
When a single-component solvent such as butane, which is not generally present in the produced heavy oil or bitumen, is used in SAP, estimation of the recovered solvent can be achieved uniquely. But when the solvent also has heavier components, some of which overlap with the lighter components of the produced oil, the measurement is not straightforward. The problem is compounded by the fact that the interaction with the reservoir changes the composition of the produced solvent and makes it time variant on account of different resident times associated with different components. The issue is further complicated by the fact that produced oil also undergoes an in-situ solvent deasphalting process (SDA), which is also time and space variant in the reservoir. If there were no in-situ SDA, one potential method to measure the amount of produced solvent would be to measure the total asphaltene content as an oil "marker" in the produced blend. Use of a tracer with injected solvent, as well as regression-based analyses for solvent fraction (using compositional analyses of solvent, bitumen, and the blend) of the produced blend, is error prone for these same reasons.
Because of the issues in these approaches, a new method is desirable for a more-robust and unique assessment of the solvent amount in the produced fluids. This paper elaborates on the current challenges and proposes a couple of workable methods, including use of maltenes/metals content as oil markers as well as the use of boiling-point curves of the produced blend compared with the boiling point of the base oil.
Such techniques of estimating the recovered solvent can facilitate a more-objective assessment of SAP field tests and enable economic evaluation of SAP application to a commercial scale.
For stratified reservoirs with free crossflow and where fractures do not cause severe channeling, improved sweep is often needed after water breakthrough. For moderately viscous oils, polymer flooding is an option for this type of reservoir. However, in recent years, an in-depth profile-modification method has been commercialized in which a block is placed in the high-permeability zone(s). This sophisticated idea requires that (1) the blocking agent have a low viscosity (ideally a unit-mobility displacement) during placement, that (2) the rear of the blocking-agent bank in the high-permeability zone(s) outrun the front of the blocking-agent bank in adjacent less-permeable zones, and that (3) an effective block to flow form at the appropriate location in the high-permeability zone(s). Achieving these objectives is challenging but has been accomplished in at least one field test. This paper investigates when this in-depth profile-modification process is a superior choice over conventional polymer flooding.
Using simulation and analytical studies, we examined oil-recovery efficiency for the two processes as a function of (1) permeability contrast, (2) relative zone thickness, (3) oil viscosity, (4) polymer-solution viscosity, (5) polymer- or blocking-agent-bank size, and (6) relative costs for polymer vs. blocking agent. The results reveal that in-depth profile modification is most appropriate for high permeability contrasts (e.g., 10:1), high thickness ratios (e.g., less-permeable zones being 10 times thicker than high-permeability zones), and relatively low oil viscosities. Because of the high cost of the blocking agent relative to conventional polymers, economics favors small blocking-agent-bank sizes (e.g., 5% of the pore volume in the high-permeability layer). Even though short-term economics may favor in-depth profile modification, ultimate recovery may be considerably less than from a traditional polymer flood.