|Theme||Visible||Selectable||Appearance||Zoom Range (now: 0)|
Abstract Uniformity of proppant distribution among multiple perforation clusters affects treatment efficiency in multistage fractured wells stimulated using the plug-and-perf technique. Multiple physical phenomena taking place in the well and perforation tunnels can cause uneven proppant distribution among multiple clusters. The problem has been studied in the recent years with experimental and computational fluid dynamics (CFD) methods, which provide useful insights but are impractical for routine designs. Simplified models that incorporated the proppant transport efficiency (PTE) correlation derived from the CFD results in a hydraulic fracture model have been also presented in literature. In this paper, we present a numerical model that simulates the transient proppant slurry flow in the wellbore, considering proppant transport and settling including bed formation, rate- and concentration-dependent pressure drop, PTE, and dynamic pressure coupling with the hydraulic fractures. The model is efficient and is designed to be an independent wellbore transport model so it can be integrated with any fracture models, including fully 3D and/or complex fracture network models, for practical design optimization. The model predictions are compared and found to agree with previously published studies. Parametric studies demonstrate sensitivity of proppant distribution to grain size, fluid viscosity, and pumping rate for fixed perforation designs. Analysis of the simulation results shows that the dominant cause of uneven proppant distribution is proppant inertia. Possible slurry stratification is less important, except for the cases with relatively low flow rates and near toe clusters. Accordingly, proppant distribution is less sensitive to perforation phasing than to the number of perforations in clusters. Alterations of the number of perforations per cluster within a stage enable achieving more even proppant distribution.
Summary In this work, we investigate the efficient estimation of the optimal design variables that maximize net present value (NPV) for the life-cycle production optimization during a single-well carbon dioxide (CO2) huff-n-puff (HnP) process in unconventional oil reservoirs. A synthetic unconventional reservoir model based on Bakken Formation oil composition is used. The model accounts for the natural fracture and geomechanical effects. Both the deterministic (based on a single reservoir model) and robust (based on an ensemble of reservoir models) production optimization strategies are considered. The injection rate of CO2, the production bottomhole pressure (BHP), the duration of injection and the production periods in each cycle of the HnP process, and the cycle lengths for a predetermined life-cycle time can be included in the set of optimum design (or well control) variables. During optimization, the NPV is calculated by a machine learning (ML) proxy model trained to accurately approximate the NPV that would be calculated from a reservoir simulator run. Similar to the ML algorithms, we use both least-squares (LS) support vector regression (SVR) and Gaussian process regression (GPR). Given a set of forward simulation runs with a commercial compositional simulator that simulates the miscible CO2 HnP process, a proxy is built based on the ML method chosen. Having the proxy model, we use it in an iterative-sampling-refinement optimization algorithm directly to optimize the design variables. As an optimization tool, the sequential quadratic programming (SQP) method is used inside this iterative-sampling-refinement optimization algorithm. Computational efficiencies of the ML proxy-based optimization methods are compared with those of the conventional stochastic simplex approximate gradient (StoSAG)-based methods. Our results show that the LS-SVR- and GPR-based proxy models are accurate and useful in approximating NPV in the optimization of the CO2 HnP process. The results also indicate that both the GPR and LS-SVR methods exhibit very similar convergence rates, but GPR requires 10 times more computational time than LS-SVR. However, GPR provides flexibility over LS-SVR to access uncertainty in our NPV predictions because it considers the covariance information of the GPR model. Both ML-based methods prove to be quite efficient in production optimization, saving significant computational times (at least 4 times more efficient) over a stochastic gradient computed from a high-fidelity compositional simulator directly in a gradient ascent algorithm. To our knowledge, this is the first study presenting a comprehensive review and comparison of two different ML-proxy-based optimization methods with traditional StoSAG-based optimization methods for the production optimization problem of a miscible CO2HnP.
von Gunten, Konstantin (University of Alberta) | Snihur, Katherine N. (University of Alberta) | McKay, Ryan T. (University of Alberta) | Serpe, Michael (University of Alberta) | Kenney, Janice P. L. (MacEwan University) | Alessi, Daniel S. (University of Alberta)
Summary Partially hydrolyzed polyacrylamide (PHPA) friction reducer was investigated in produced water from hydraulically fractured wells in the Duvernay and Montney Formations of western Canada. Produced water from systems that used nonencapsulated breaker had little residual solids (<0.3 g/L) and high degrees of hydrolysis, as shown by Fourier-transform infrared (FTIR) spectroscopy. Where an encapsulated breaker was used, more colloidal solids (1.1–2.2 g/L) were found with lower degrees of hydrolysis. In this system, the molecular weight (MW) of polymers was investigated, which decreased to <2% of the original weight within 1 hour of flowback. This was accompanied by slow hydrolysis and an increase in methine over methylene groups. Increased polymer-fragment concentrations were found to be correlated with a higher abundance of metal-carrying colloidal phases. This can lead to problems such as higher heavy-metal mobility in the case of produced-water spills and can cause membrane fouling during produced-water recycling and reuse.
Summary In this paper, we investigate the change in oil effective permeability () caused by fracturing‐fluid (FF) leakoff after hydraulic fracturing (HF) of tight carbonate reservoirs. We perform a series of flooding tests on core plugs with a range of porosity and permeability collected from the Midale tight carbonate formation onshore Canada to simulate FF‐leakoff/flowback processes. First, we clean and saturate the plugs with reservoir brine and oil, and age the plugs in the oil for 14 days under reservoir conditions (P = 172 bar and T = 60°C). Then, we measure before (baseline) and after the leakoff process to evaluate the effects of FF properties, shut‐in duration, and plug properties on regained permeability values. We found that adding appropriate surfactants in FF not only significantly reduces impairment caused by leakoff, but also improves compared with the original baseline for a low‐permeability carbonate plug. For a plug with relatively high permeability (kair > 0.13 md), freshwater leakoff reduced by 55% (from 1.57 to 0.7 md) while FF (with surfactants) reduced by only 10%. The observed improvement in regained is primarily because of the reduction of interfacial tension (IFT) by the surfactants (from 26.07 to 5.79 mN/m). The contact‐angle (CA) measurements before and after the flowback process do not show any significant wettability alteration. The results show that for plugs with kair > 0.13 md, FF leakoff reduces by 5 to 10%, and this range only increases slightly by increasing the shut‐in time from 3 to 14 days. However, for the plug with kair < 0.09 md, the regained permeability is even higher than the original before the leakoff process. We observed 28.52 and 64.61% increase in after 3‐ and 14‐day shut‐in periods, respectively. This observation is explained by an effective reduction of IFT between the oil and brine in the pore network of the tight plug, which significantly reduces irreducible water saturation (Swirr) and consequently increases . Under such conditions, extending the shut‐in time enhances the mixing between invaded FF and oil/brine initially in the plug, leading to more effective reductions in IFT and consequently Swirr. Finally, the results show that the regained permeability strongly depends on the permeability, pore structure, and Swirr of the plugs.
Summary The objectives of this study are to perform a fundamental analysis of the mutual interactions between crude oil components, water, hydrocarbon solvents, and clays, and to determine the optimum hydrocarbon solvent in solvent steamflooding for a particular reservoir type. The water/oil emulsion formation mechanism in the obtained oil for steam and solvent steamflooding processes has been studied via intermolecular associations between asphaltenes, water, and migrated clay particles. A series of 21 steam and solvent-steamflooding experiments has been conducted, first without any clays in the oil/sand packing, and then using two different clay types in the reservoir rock: Clay 1, which is kaolinite, and Clay 2, which is a mixture of kaolinite and illite. Paraffinic (propane, n-butane,n-pentane,n-hexane,n-heptane) and aromatic (toluene) solvents are coinjected with steam. Cumulative oil recovery is found to decrease in the following order: no clay, Clay 1, Clay 2. Based on the obtained produced oil analyses, Clay 1 and Clay 2 are found to have an affinity with the water and oil phases, respectively. Moreover, the biwettable nature of Clay 2 makes it dispersed in the oil phase toward the oil/water interface, stabilizing the water/oil emulsions. Paraffinic solvent n-hexane is found to be an optimum coinjector for solvent steamflooding in bitumen recovery.
Zargar, Masoumeh (University of Western Australia and Edith Cowan University) | Johns, Michael L. (University of Western Australia) | Aljindan, Jana M. (Saudi Aramco) | Noui-Mehidi, Mohamed Nabil (Saudi Aramco) | O'Neill, Keelan T. (University of Western Australia)
Summary Multiphase flowmetering is a requirement across a range of process industries, particularly those that pertain to oil and gas. Generally, both the composition and individual phase velocities are required; this results in a complex measurement task made more acute by the prevalence of turbulent flow and a variety of flow regimes. In the current review, the main technical options to meet this metrology are outlined and used to provide context for the main focus on the use of nuclear magnetic resonance (NMR) technology for multiphase flowmetering. Relevant fundamentals of NMR are detailed as is their exploitation to quantify flow composition and individual phase velocities for multiphase flow. The review then proceeds to detail three NMR multiphase flowmeter (MPFM) apparatus and concludes with a consideration of future challenges and prospects for the technology.
Summary Organic matters in source rocks store oil in significantly larger volume than that based on its pore volume (PV) due to so-called nanoconfinement effects. With pressure depletion and production, however, oil recovery is characteristically low because of the low compressibility of the fluid and amplified interaction with pore surface in the nanoporous material. For the additional recovery, CO2 injection has been widely adopted in shale gas and tight oil recovery over the last decades. But its supply and corrosion are often pointed out as drawbacks. In this study, we propose ethane injection as an alternative enhanced oil recovery (EOR) strategy for more productive oil production from tight unconventional reservoirs. Monte Carlo (MC) molecular simulation is used to reconstruct molecular configuration in pores under reservoir conditions. Further, molecular dynamics (MD) simulation provides the basis for understanding the recovery mechanism of in-situ fluids. These enable us to estimate thermodynamic recovery and the free energy associated with dissolution of injected gas. Primary oil recovery is typically below 15%, indicating that pressure depletion and fluid expansion are no longer effective recovery mechanisms. Ethane injection shows 5 to 20% higher recovery enhancement than CO2 injection. The superior performance is more pronounced, especially in nanopores, because oil in the smaller pores is richer in heavy components compared to the bulk fluids, and ethane molecules are more effective in displacing the heavy hydrocarbons. Analysis of the dissolution free energy confirms that introducing ethane into reservoirs is more favored and requires less energy for the enhanced recovery.
Summary Progressing cavity pump (PCP) is the essential booster equipment in oil–gas mixing delivery. Changes in relevant parameters in PCP operations directly affect the working performance and service life of the pump. On the basis of computational fluid dynamics (CFD) in this study, we apply dynamic grid technology to establish a 3D flow field numerical calculation model for the CQ11-2.4J PCP, which is used in the field of the Hounan Operation Area in Changqing oil field, China. The effects of several operating parameters, such as oil viscosity, pump rotation speed, differential pump pressure, and void fraction of oil, on the pressure and the velocity distribution of the PCP flow field are examined. Various performance parameters in the transport of the oil–gastwo-phase mixture are used in the analysis, including volumetric flow rate, slippage, shaft power, volumetric efficiency, and system efficiency. The results show that the pressure and speed distribution in the pump chamber of the PCP is relatively homogenous under different working conditions, whereas the pressure and speed exhibited sharp changes at the stator and rotor sealing line and adjacent areas in the pump chamber. Increasing the viscosity of the oil and the speed of the rotor can effectively improve the flow characteristics of the PCP, but extremely high pump rotation speed would cause a decline in system efficiency. Increasing the differential pressure and the void fraction of oil would result in a decrease in the volumetric flow rate and efficiency of the PCP. Considering the variation law of the PCP's performance parameters, the optimal interval for each operating parameter of the PCP is as follows: Oil viscosity at 50–100 mPa·s, pump rotation speed at 200–300 rev/min, differential pressure at 0.2–0.3 MPa, and the void fraction of oil not more than 50%. This research can provide technical support for the optimization of the working conditions of the PCP on site.
Liu, Wendi (The University of Texas at Austin) | Ikonnikova, Svetlana (The University of Texas at Austin and Technical University of Munich) | Scott Hamlin, H. (The University of Texas at Austin) | Sivila, Livia (The University of Texas at Austin (now with EnerVest Ltd.)) | Pyrcz, Michael J. (The University of Texas at Austin)
Summary Machine learning provides powerful methods for inferential and predictive modeling of complicated multivariate relationships to support decision-making for spatial problems such as optimization of unconventional reservoir development. Current machine-learning methods have been widely used in exhaustive spatial data sets like satellite images. However, geological subsurface characterization is significantly different because it is conditioned by sparse, nonrepresentative sampling. These sparse spatial data sets are generally not sampled in a representative manner; therefore, they are biased. The critical questions are: first, does spatial bias in training data result in a bias for machine-learning-based predictive models; and if there is a bias, how can we mitigate the bias in these spatial machine-learning-based predictions? The presence and mitigation of prediction with spatial sampling bias is demonstrated with tree-based machine learning due to its high degree of interpretability. In expectation, training data bias imposes bias in machine-learning predictions over a wide variety of spatial data configurations and degrees of bias, even when the model is applied to make predictions with unbiased testing and real-world data. We reduce the bias in prediction with a novel spatial weighted tree method over a variety of spatial data configurations and degrees of spatial sampling bias. The proposed method is able to improve the accuracy for reservoir evaluation. We recommend modeling checking and bias mitigation for all machine-learning prediction models with sparse, spatial data sets, because bias in, bias out.
Summary An integrated modeling of the cyclic steam stimulation (CSS) at the Peace River heavy‐oil/oil‐sand deposits in Alberta, Canada, is challenging because of the presence of compositional gradient, faulting, and bottomwater pockets, and the variations in the oil viscosity, rock dilation, fracturing, and the pay‐zone‐thickness variation. Both gravity and viscosity are marked by declining quality with depth, biodegradation, and compartmentalization. The high oil viscosity and low water mobility at Peace River cause low initial injectivity. High injectivity during the CSS is achieved by high‐pressure injection to fail the formation mechanically and trigger fracturing and pore deformation. Moreover, the pore dilation/recompaction triggers relative permeability hysteresis. History matching of high steam injectivities is challenging when reasonable fracture lengths and rock compressibilities are used. To match injectivities, most reservoir simulations have used either a larger compressibility (“spongy‐rock” approach) or long fractures. The spongy‐rock approach predicts a steady increase in injection pressure, whereas during the early time of the injection cycles, injection pressures increase and then level off for most of the cycle. We describe the enhancements made in an iteratively coupled geomechanical–flow model to incorporate the modeling of both pore deformation and relative permeability hysteresis to match the injection pressures, steam injectivity, and oil/water productions from CSS at Peace River that are otherwise difficult to reproduce. The geomechanical model explains surface heave and high injectivity caused by dilation attributable to shear failure, increase in pore pressure/formation compressibility, and decrease in effective stress. A dilation pressure is specified, below which the behavior is elastic and above which a higher compressibility is used. Above a maximum porosity, further dilation is not permitted. Also, the hysteresis model calculates gridblock relative permeabilities that lie on or between the imbibition/drainage curves, making it possible to use the laboratory‐derived two‐phase oil–water relative permeabilities and still match the field‐measured water‐ and oil‐production volumes. By combining an iteratively coupled reservoir–geomechanical model for the CSS with stochastic workflows, including the Latin‐hypercube design (LHD) and response‐surface methodology (RSM), the impacts of dilation/recompaction factors (fracturing pressure, maximum injection pressure, dilation pressure, recompaction pressure, and formation compressibility) are quantified through history matching the field results and automated stochastic sensitivity analysis and uncertainty assessment.