After hydraulically fracturing of shale gas wells, theoretical and experimental studies showed that over 75% of the injected water-based fracture fluids left unrecovered. The trapped water causes permeability damage and productivity impairment. The flowback water also tends to be highly saline, often with TDS contents of as much as 200,000 ppm. This study aims to investigate the effect of well shut-in before flowback stage (the soaking process) on the production of shale and tight sandstone formations.
Shale and sandstone samples were analyzed by X-ray diffraction (XRD). Marcellus shale and Kentucky sandstone cores were used. A modified core flood setup was used to allow porosity measurements by gas expansion method, then pulse decay permeability measurements, and fluid injection during the leak-off process. Nitrogen was used for gas expansion and permeability measurements, while 5 wt% KCl brine was used as representative of leak-off fracturing fluid. The fracturing fluid was injected under a constant pressure gradient (300 in the case of sandstone cores and 1,500 psi in the case of shale cores. After removing the pressure gradient, gas permeability was measured at different soaking times. Computed tomography (CT) was used to scan the cores during the experiment to observe the propagation of fracturing fluid in the core with time.
The results show increasing the regain permeability for sandstone formation was 60% of its initial value directly after the leak-off stage. Then, the regain permeability decreased with increasing the soaking time 38% of its initial value after the core completely invaded with leak-off fluid. The regain permeability was then increased with longer soaking time, as a result of reducing the chocking effect at the core inlet. The propagation rate of water saturation front from CT-scan data decreased with time until reaching the core outlet. The regain permeability on shale cores was 0.14 of its initial value and decreased with soaking time, due to depressed relative permeability curve on this tight pore-space cores.
This study addresses the mechanism of production enhancement or reduction as a result of the soaking process for shale and tight sandstone formations.
Despite numerous studies in the subject matter, industry has yet to resolve casing failure issues. A more interdisciplinary approach is taken in this study integrating seventy-eight land based wells using a data - driven approach to predict the reasons behind casing failure. This study uses a statistical software in collaboration with Python Scikit-learn implementation to apply different Data Mining and Machine Learning algorithms on twenty-four different features on the twenty failed casing data sets. Descriptive analytics manifested in visual 8representations included Normal Distribution Charts and Heat Map. Principal component Analysis (PCA) was used for dimensionality reduction. Supervised and unsupervised approaches were selected respectively based on the response. The algorithms used in this study included Support Vector Machine (SVM), Boot strap, Random Forest, Naïve Bayes, XG Boost, and K-Means Clustering. Nine models were then compared against each other to determine the winner. Features contributing to casing failure were identified based on best algorithm performance.
The design process of carbonate matrix acidizing treatments requires coring and conducting linear, radial core-flood experiments. With the current environment revolving around cutting costs, it becomes increasingly more important to accurately design cost effective acidizing treatments. This work aims to introduce a novel approach to predicting the performance of acid treatments in the field using log data only. A radial reactive flow simulator, using porosity distributed from logs, is utilized to provide accurate predictions without the need for experiments.
Core-flood acidizing experiments at two temperatures (150 and 200°F) with two acid concentrations were studied. A reactive flow simulator was built using porosity distribution derived from computed tomography (CT) scan and tuned to match experimental data. A new radial simulation model of 3.25 ft. radius was utilized to study acid propagation under field conditions. For accurate predictions, porosity was distributed using cores CT scan derived values. Simulation results were compared with traditional 1-D models. Different porosity distributions, including gamma distributions, were used in the radial model.
The reactive flow simulator was able to accurately capture wormhole propagation inside the linear core. A greater than 90% match between experimental and simulated acid pore volume to breakthrough (PVBT) was obtained using two different temperatures and acid concentrations. The simulation results from the radial field scale model show that the optimum velocity can be higher or lower than those predicted from lab experiments. Accordingly, caution must be taken when linear core flood data is used to predict acid propagation in the field. The simulations showed that traditional upscaling models overpredict acid volumes, as the predicted volumes are double at moderate to high injection rates. Models using statistically distributed porosity can provide accurate acid propagation predictions, with a relative percentage error less than 25% at extremely high injection rates.
This work introduces an accurate model using porosity directly from logs to predict acid performance while avoiding expensive designs. The simulation results revealed that traditional designs overpredict acid volumes required for field treatments. The statistically distributed porosity can be used as a substitute for CT scan derived porosity with low effect on model predictability. The reactive flow simulator can accurately match experimental data.
Almubarak, Tariq (Texas A&M University) | Li, Leiming (Aramco Services Company) | Nasr-El-Din, Hisham (Texas A&M University) | Ng, Jun Hong (Texas A&M University) | Sokhanvarian, Khatere (Sasol Chemical) | Alkhaldi, Mohammed (Saudi Aramco) | Almubarak, Sama (Saudi Aramco)
In order to satisfy the demand for oil and gas, it becomes increasingly necessary to produce from formations that are deeper, have low permeability, and higher temperature. Conventionally, hydraulic fracturing fluids make use of viscosifiers such as guar and its derivatives to generate the rheological properties required during the fracturing process. However, to withstand the high-temperature environments, higher loadings of polymer is required. This leads to an increase in polymer and additive concentrations. Most importantly, these higher loading fluids do not break completely, and generate residual polymer fragments that can plug the formation and reduce fracture conductivity significantly.
This work builds on previous work which introduced a new hybrid dual polymer hydraulic fracturing fluid that was developed for high-temperature applications. The fluid consists of a guar derivative and a polyacrylamide-based synthetic polymer. Compared to conventional fracturing fluids, this new system is easily hydrated, requires less additives, can be mixed on the fly, and is capable of maintaining excellent rheological performance at low polymer loadings. In this work, the fluid is further optimized to withstand even higher temperatures up to 400°F.
Total polymer loadings of 30 lb/1,000 gal and 40 lb/1,000 gal dual polymer fracturing fluid were tested in this work and were prepared in the ratio of 1:1 and 1:2 (CMHPG: Synthetic). They were then crosslinked with a metallic crosslinker and placed in a HPHT rheometer to measure the viscosity between 200 and 400°F. After observing the failure temperature of the mixtures, additives such as buffers, crosslinking delayers, and oxygen scavengers were added and tested at temperatures above that point. The type of crosslinker used was also varied to observe the effects of the rate of release of the metallic crosslinker on thermal stability.
The results indicate that the 1:2 (CMHPG: Synthetic) mixture performed better at temperatures exceeding 330°F than the 1:1 mixture. The failure point of both mixtures was observed to be 350°F for the latter while the former failed at 370°F. The addition of a crosslinker that allowed a more controllable release was observed to improve the thermal stability of the fluid mixture above 370°F by increasing the polymer's shear tolerance. The addition of additives to the mixture was shown to improve the thermal stability of the solution to varying degrees. Of the three additives, the most significant enhancement came from the addition of oxygen scavengers while the least was from the buffer solution.
This study investigates the performance of viscoelastic surfactant (VES)-based HCl stimulation fluids as a function of carbonate rock type and quantifies the response of the acid to different pore-structures. A pore-structure evaluation during stimulation design could lead to a successful field treatment.
Coreflood tests were conducted using several types of limestone cores with permeabilities ranging from 2.5 to 155 md. Intergranular pores were dominant in the Indiana limestone and Austin chalk samples investigated, whereas moldic pores were dominant in the Pink desert, Edwards yellow, Winterset, and Edwards white cores. Tracer experiments characterized the pore structure in each carbonate sample, and the tracer fluid was injected at 5 cm3/min and 75°F into the cores with dimensions of 6 in. length and 1.5 in. diameter. The tracer effluent data was used to measure accessible porosity (flowing fraction) for each core sample. After the tracer, the VES acid was injected at rates from 1 to 10 cm3/min and 150°F to determine pore volume to breakthrough (PVbt). The wormhole patterns were analyzed using computed tomography (CT) scan images, and the pattern complexity was examined by fractal dimension analysis.
A better pore connectivity showed for Indiana limestone compared to Edwards yellow, Winterset limestone, and Edwards white. The flowing fractions were 1, 0.86, 0.61, and 0.53 for Indiana limestone, Edwards yellow, Winterset limestone, and Edwards white, respectively. The PVbt of Indiana limestone ranged from 0.62 to 0.92. Cores with lower pore connectivity, such as Edwards yellow, had PVbt ranging from 0.52 to 0.81, Winterset limestone from 0.34 to 0.49, and Edwards white from 0.21 to 0.36. These results revealed that higher flowing fractions are required with a higher PVbt. Rocks that have the same dominant pore-structures usually exhibit similar wormhole behavior.
Prior to this study, the performance of VES fluids had only been studied on carbonate rocks with well-connected intergranular porosity. The results of this study show that porosity distribution of the rock affects the response to acids.
Historical horizontal completions designs have very wide cluster spacing, leaving behind significant volumes of hydrocarbons. This paper develops a workflow for optimizing cluster spacing using simulated production curves in unconventional oil and gas fields. Optimizing cluster spacing reduces unstimulated reservoir rock left between widely spaced fractures, more efficiently draining the stimulated reservoir volume and increasing expected ultimate recovery and initial production. This paper illustrates the workflow developed by finding an optimal range for cluster spacing in the retrograde/wet gas region of the Eagle Ford. It is estimated that optimizing cluster spacing in this fluid window will increase ultimate hydrocarbon recovery by 20% and the net present value of each well by 50-60%.
Chen, Rongqiang (Texas A&M University) | Xue, Xu (Texas A&M University) | Yao, Changqing (Texas A&M University) | Datta-Gupta, Akhil (Texas A&M University) | King, Michael J. (Texas A&M University) | Hennings, Peter (University of Texas Bureau of Economic Geology) | Dommisse, Robin (University of Texas Bureau of Economic Geology)
A series of earthquakes was recorded along a mapped fault system near Azle, Texas in 2013. To identify the mechanism of seismicity, coupled fluid flow and geomechanical simulation is carried out to model fluid injection/production and the potential onset of seismicity. Sensitivity studies for a broad range of reservoir and geomechanical parameters are performed and the calibrated models are used to identify controlling mechanisms for seismicity in the Azle area, North Texas and its relationship to hydrocarbon production and fluid injection in the vicinity. Geologic, production/injection, and seismicity data are gathered to build a detailed simulation model with coupled fluid flow and geomechanics. Geomechanical simulation results are used to calculate cumulative seismic moment magnitude. Sensitivity analyses for injection well head pressure and earthquake data are performed over a range of reservoir and geomechanical parameters. Influential parameters are selected to perform a pareto-based multi-objective history matching of well head pressures and seismic moments.
Geomechanical interaction has significant impact on seismicity in the Azle area. Unbalanced loading (overall injection and production) on different sides of the fault generates accumulation of strain change, resulting in the onset of seismicity. Previous studies seem to have significantly underestimated the fluid withdrawal rates, almost by an order of magnitude. The equivalent bottom-hole fluid rate used in this study suggests a drop in reservoir pore pressure which is consistent with the BHP trends. Thus, pore pressure increases may not explain the seismicity near the Azle area, as indicated in previous studies. Instead geomechanical effects and strain propagation to the basement appear to be the dominant mechanisms. The low fault cohesion and minimum horizontal stress obtained from history matching suggest that the faults must be near or at the critically-stressed state before the initiation of fluid production/injection. A sensitivity analysis indicates that the minimum horizontal stress and fracture gradient each play a critical role in the potential risk for seismicity related to fluid injection/production. Streamline flow pattern further proves that there is no fluid movement in the basement formation and the unbalanced loading from different sides of the fault is the controlling mechanism. This is the first study coupling fluid flow and geomechanics in the Azle area and the first to simultaneously calibrate the models with fluid flow and seismicity data.
A major part of the uncertainty for shale reservoirs comes from the distribution and properties of the fracture network. However, explicit fracture models are rarely used in uncertainty quantification due to their high computational cost. This paper presents a workflow to match the history of reservoirs with complex fracture network with a novel forward model. By taking advantage of the efficiency of the model, fractures can be explicitly characterized, and the corresponding uncertainty about the distribution and properties of fractures can be evaluated. No upscaling of the fracture properties is necessary, which is usually a required step in a traditional workflow.
The embedded discrete fracture model (EDFM) has recently been studied by many researchers due to its high efficiency compared to other explicit fracture models. By assuming a linearly distributed pressure near fractures, EDFM can provide a sub-grid resolution that lifts the requirement to refine near the fractures to a comparable size as the fracture aperture. Although efficient, considerable error is reported when applying this method to simulate flow barriers, especially when dominant flux direction is across instead of along the fractures. In this work, a novel discrete fracture model, compartmental EDFM (cEDFM) is developed based on the original EDFM framework. However, different from the original method, in cEDFM the fracture would split matrix grid blocks when intersecting them. The new model is benchmarked for single phase as well as multi-phase cases, and the accuracy is evaluated by comparing to fine explicit cases. Results indicate the improved model yields much better accuracy even for multi-phase flow simulation with flow barriers.
In the second part of the work, we applied the model in history matching and performed uncertainty quantification to the fracture network for two synthetic cases. We used Ensemble Kalman Filter (EnKF) as the data assimilation algorithm due to its robustness for cases with large uncertainty. The initial state does not need to be close to the truth to achieve convergence. Also EnKF performs well for the history matching of reservoirs with complex fracture network, where the number of parameters can be large. Therefore, it is advantageous compared to using Ensemble Smoother (ES) or Markov Chain Monte Carlo (MCMC) in this case. After the final step of data assimilation, a good match is obtained that can predict the production reasonably well. The proposed cEDFM model shows its robustness to be incorporated into the EnKF workflow, and benefit from the efficiency of the model, this work made it practical to perform history matching with explicit fracture models.
Zhang, Fan (Texas A&M University) | Adel, Imad A. (Texas A&M University) | Park, Kang Han (Texas A&M University) | Saputra, I. W. R. (Texas A&M University) | Schechter, David S. (Texas A&M University)
Field observations, along with experimental laboratory, exhibit evidence that enhancing production by CO2 huff-n-puff process is a potential EOR technique that improves the, commonly low, ultimate oil recovery in unconventional liquid reservoirs (ULR). As pressure goes beyond the MMP, intermediate components of oil vaporize into the CO2 and consequently condense at room pressure and temperature. In addition, Surfactant-Assisted Spontaneous Imbibition (SASI) process has been widely believed to enhance oil recovery in ULR, which has been investigated by several laboratory and numerical studies. During the hydraulic fracturing with surface active additives, surfactant molecules interact with rock surfaces to enhance oil recovery through wettability alteration and interfacial tension reduction. The wettability alteration leads to the expulsion of oil from the pore space as well as water being imbibed into the matrix spontaneously. However, the understanding of hybrid EOR technologies, combining both gas injection and surfactant imbibition, to enhance recovery in ULR is not well studied.
In this manuscript, we assess the potential of combining both CO2 huff-n-puff and surfactant imbibition techniques in optimizing oil recovery in ULR. Sidewall core samples retrieved from ULR were first cleaned utilizing the Dean-Stark methodology and then saturated by pressurizing them with their corresponding oil for three months. CO2 huff-n-puff experiments were operated on shale core samples under different pressures in a set-up integrated into a CT-scanner. Those cores were then submerged in the surfactant solution, in a modified Amott cell, to observe whether any additional oil is produced through the process of SASI. Total production from these two different methods, which was done sequentially, will provide insight into the possibility of hybrid EOR technology. CO2 huff-n-puff experiments were performed below and above the MMP which was previously determined by the slim-tube method. Contact angle (CA), interfacial tension (IFT) were also measured on the saturated shale core samples. CT-Scan technology was used to visualize the process of oil being expelled from the core plugs in both CO2 huff-n-puff and spontaneous imbibition experiments.
Experimental results provide a promising outcome on the application of hybrid EOR technology, CO2 huff-n-puff and SASI, improving oil recovery from ULR. Oil recovery was observed to reach around 50% of measured OOIP from CO2 huff-n-puff alone with an addition of 10% recovery from SASI after the CO2 treatment. A detailed description of the correlated experimental workflows is presented to investigate the hybrid EOR technology in enhancing oil recovery in ULR. In addition, a discussion on the difference in mechanism of oil production from the huff-n-puff and SASI method is also included alongside several additional novel findings regarding the color shift of the produced oil. MMP data of CO2 and oil measured as well as a change of contact angle (CA) and interfacial tension (IFT) when the surfactant is introduced into the system are also provided to support insight on the mechanism of the production improvement. All measured and compiled data deliver the required information for this study to demonstrate the possibility of combining both CO2 EOR and SASI EOR, a hybrid EOR, as a practical method to produce a significant amount of oil from unconventional shale oil reservoirs.
Liu, Tian (Texas A&M University)
The integration of seismic data into high-resolution geological model, provides great potential for calibrating reservoir parameters, which enables better understanding of the reservoir sweep and flow patterns. The efficacy of seismic inversion method based on the travel time of fluid saturation front using seismic onset times has been well demonstrated for integrating frequent time-lapse seismic surveys. However, due to the high cost associated with conducting seismic surveys, frequent seismic surveys are usually not commonly available. In this paper, I have generalized the onset time inversion method for infrequent seismic data using interpolated seismic onset times, making the method applicable for efficiently integrating infrequent seismic surveys. With the valuable information provided by seismic data, the uncertainty in the reservoir model parameters can be reduced through history matching. The history matched model can be used to optimize reservoir management and field development strategy. The proposed method is illustrated using synthetic and field applications.