|Theme||Visible||Selectable||Appearance||Zoom Range (now: 0)|
Abstract Market-induced production shut-downs and restarts offer us an opportunity to gather step-rate and shut-in data for pressure transient analysis (PTA) and rate transient analysis (RTA). In this study, we present a unified transient analysis (UTA) to combine PTA and RTA in a single framework. In this new approach continuous production data, step-rate data, shut-in data and re-start data can be visualized and analyzed in a single superposition plot, which can be used to estimate both and infer formation pore pressure in a holistic manner by utilizing all available data. Most importantly, we show that traditional log-log and square root of time plots can lead to false interpretation of the termination of linear-flow or power-law behavior. Field cases are presented to demonstrate the superiority of the newly introduced superposition plot, along with discussion on the calibration of long-term bottom-hole pressure with short-term measurements.
Observed deviations from classical diffusion (specifically transient linear flow, TLF) for Permian Basin Wolfcamp shale (oil) wells have historically been attributed to anomalous diffusion (AD) and reservoir heterogeneity. Recently, the authors have suggested that the combination of multiphase flow and variable operating conditions could create similar reservoir signatures (as quantified using classic flow-regime identification techniques). However, there are additional factors, not previously investigated, that could be affecting flow-regime identification (ID). This work aims to systematically investigate all possible causes of deviations from classical diffusion behavior.
A new, general semi-analytical model is developed that explicitly accounts for multi-phase flow, stress-sensitive reservoir properties, and variable operational conditions. This is achieved through the combination of modified pseudo-pressure and material balance pseudo-time. Additionally, reservoir heterogeneity and AD are incorporated into the exponent of the power-law relationship between rate-normalized pseudo-pressure and material balance pseudo-time.
A new workflow is introduced to improve flow-regime ID when various complex reservoir behaviors are occurring simultaneously. Using this new workflow, a comprehensive investigation into which parameters are causing deviations from TLF behavior is performed using simulated cases with inputs that are similar to those used to history-match a Wolfcamp shale well. Further, errors in the determination of reservoir and fracture properties caused by misdiagnosis of the flow regime are analyzed.
Our extensive investigations demonstrate that: 1) multi-phase flow, and in particular the appearance of mobile gas (producing gas-oil-ratio ramps up) as flowing bottomhole pressure (FBHP) drops below bubble point pressure, has the biggest impact on the deviation from TLF; 2) water production has a relatively small effect, even for cases with high producing water-oil-ratios; 3) pressure-dependent permeability has a relatively small effect, even for cases where FBHP changes strongly; 4) deviations from TLF in a Wolfcamp shale oil example may be primarily caused by the combination of multiphase flow and variable FBHP, while the AD effect is likely secondary.
Importantly, if log-log flow-regime identification plots are not corrected for complex reservoir behaviors, the reservoir flow regime may become distorted, leading to the selection of incorrect models for production analysis and errors in reservoir/fracture properties estimates.
Zhan, Lang (Shell International Exploration and Production Inc.) | Tokan-Lawal, Adenike (Shell Exploration and Production Co.) | Fair, Phillip (Shell International Exploration and Production Inc.) | Dombrowski, Robert (Shell International Exploration and Production Inc.) | Liu, Xin (Shell International Exploration and Production Inc.) | Almarza, Veronica (Shell Exploration and Production Co.) | Girardi, Alejandro Martin (Shell Exploration and Production Co.) | Li, Zhen (Shell Exploration and Production Co.) | Li, Robert (Shell Exploration and Production Co.) | Pilko, Martin (Shell Exploration and Production Co.) | Joost, Noah (Shell Exploration and Production Co.)
Summary Hydraulic fractures play a central role in the performance of multistage fractured horizontal wells (MFHWs) in tight and shale reservoirs. Fracture conductivity variations and connection quality between fractures and wellbore (i.e., choking skins) strongly affect well productivity. However, convincing and high-quality evaluations of hydraulic fractures for these reservoirs are rare in literature because quantifying fracture properties requires decoupling them from fracture geometry and formation properties, a difficult task in most field conditions. A data gathering and hypothesis testing program was implemented using six multifractured horizontal wells in a pad in the Delaware Basin to improve our ability to reliably forecast well performance. A systematic approach utilizing production, shut-ins, and bottomhole pressure measurements (BHP) was conducted and used to evaluate the apparent flow capacity of hydraulic fractures. Two independent techniques were used in the data analyses to characterize the hydraulic fractures; namely, pressure transients for individual wells and significant well-to-well interference signals. Both techniques render similar decline rate interpretations for the sets of fracture conductivity/permeability from analysis of the pressure data, but there is a large difference in the uncertainty of the estimated results from these two methods. The first method used a radial/linear flow regime in successive pressure buildups in three of the six wells. Simulations and theoretical analysis show that this flow regime allows decoupling fracture conductivity from fracture geometry and matrix properties. This flow regime yields the total apparent fracture conductivity (TAFC), which represents the lump sum effect of fracture conductivity. In addition, this technique characterizes the connection condition between the dominant fractures and borehole, which can be estimated from the early derivative horizontal line in pressure transient log-log diagnostic plots with minimum assumptions. Specifically, the estimated TAFC ranges from 1,140 to 1,630 md-ft at early time of well life to 525 to 855 md-ft after 100 to 139 days in production, or about a 45 to 61% reduction among these wells. The second method uses time-lag of pulse interference responses between an active and observation well. With assumptions of low, mid, and high values of fracture porosity, fracture compressibility, and fluid viscosity, characteristic fracture permeability can be estimated. Because of the large uncertainty related to the assumed fracture porosity and fracture compressibility, the pulse interference method is not likely to deliver the same certainty range as successive pressure buildups using the radial/linear flow regime. The results of this work provide a better understanding of the mechanisms of flow transport inside hydraulic fractures and at the connection between the hydraulic fractures and wellbore. The estimated TAFC and its significant decline help improve hydraulic fracturing designs and build representative reservoir models for more reliable well performance modeling and forecasting.
Seth, Puneet (The University of Texas at Austin) | Manchanda, Ripudaman (The University of Texas at Austin) | Elliott, Brendan (Devon Energy) | Zheng, Shuang (The University of Texas at Austin) | Sharma, Mukul (The University of Texas at Austin)
Abstract During stimulation in a treatment well, offset well pressure measurements resulting from stress-shadow related interference are often used to estimate hydraulic fracture geometry. Current pressure interference models typically assume one dominant fracture per stage in their analysis, which is an overly simple assumption and can result in erroneous estimates of hydraulic fracture geometry. This stems from the limited capability of existing models which are not equipped to interpret multi-cluster fracture propagation scenarios. In this study, we present workflows to analyze dynamic pressure time-series responses observed at offset monitor wells during injection in a nearby treatment well, to diagnose multi-cluster fracture propagation. A fully-coupled, 3-D, reservoir-fracturing simulator which models hydraulic fractures explicitly as compliant discontinuities has been used to simulate pressure interference in multi-well pads. We model dynamic fracture propagation from multiple clusters in the treatment well and analyze the corresponding pressure changes observed at an offset monitor well. We apply our dynamic pressure transient analysis model to analyze multi-cluster fracture propagation from the treatment well and contrast it with a scenario that assumes one dominant fracture per stage. We show that the simulated offset pressure response during the propagation of one dominant fracture per stage is very different compared to a multi-cluster propagation. We analyze the dynamic intra-stage offset well pressure signatures (inflections in the pressure response, slope of the pressure response, arrival times etc.) to develop workflows to test diversion effectiveness and provide insights on optimum job volumes, in a relatively inexpensive manner. We show the impact of completion design and perforation erosion on the offset well pressure response. We test different cluster designs and analyze the offset well pressure response in each case to quantify cluster efficiency. We apply our dynamic pressure interference testing model to field data from the Permian Basin to test for diversion effectiveness and diagnose dominant cluster variability during stimulation.
Abstract Using planar fracture models to match treatment pressure and improve understanding of the fracture geometry generation is not a new concept. Knowledge gained from this exercise has historically been used to improve engineered fracture completions and production, and maximize net present value (NPV); however, at some point during the progression from vertical to horizontal wellbores, many within the industry have forgotten about the learnings that can still be gained from current fracture models. Engineered completions have been largely replaced by spreadsheet efficiencies relevant to operations rather than production in too many cases. Some images of unconventional well stimulation treatments portray fractures growing in every direction, forming patterns that resemble shattered windshields, and have often excluded the known physics related to rock geomechanics, reservoir properties, and geology. Excuses to dismiss modeling are numerous and are gaining the reasoning of conformists. Unconventional resource plays might or might not contain large numbers of natural fractures; but, current fracture models can still be used to gain insight into the fracture geometries being generated. While the development of complex fracture models continues to evolve, the industry can still gain insight to fracture geometry and resulting production using current planar fracture modeling. Caveats to this process are that it requires: Valid measured data to establish model constraints. The engineer to understand the basic physics of how fractures are generated and when (and when not) to twist the "knobs" in the model. The engineer to understand which "knobs" should be used based on real diagnostics information. The actual single well production to be an integral part of the process. This paper demonstrates the results of honoring data measurements from a multitude of potential sources, including downhole microseismic data, downhole deformation tiltmeters, offset pressure monitoring, DTS, DAS, diagnostic fracture injection test (DFIT) analysis, injection as well as production data with bottomhole pressure measurements, etc., and the resulting observations and conclusions. Several industry examples are discussed to help frame the vast amount of information possible to help engineers do a better job of including more diagnostics into routine operations to provide additional insight and ultimately result in improved models and completion designs. This paper is not intended to merely demonstrate the results of the work but to spark an interest in bringing more intense engineering back to fracture stimulation modeling for horizontal completions.
Abstract A minimal mathematical model for transient diffusion in fractured rocks at the well scale with options to link topological and geometrical properties of fractured rocks where the geology is complex is presented. We simulate transient diffusion in systems with complex geology through a transient, interporosity model consisting of three constants to consider both complex structures that may exist at very fine scales and forms that result in long-range connectivity across multiple links in a network. Fractional constitutive flux laws that address super- and subdiffusive influences and result in transient propagation of the form r~ta, with a≠1, are considered. The spatiotemporal evolution of the transient is documented, and well-codified Flow Regimes and sub-Flow Regimes are cataloged into stages with associated typical characteristics often visible in both short- (buildup) and long-time (flow) well-behaviors. We arrive at interpretations well beyond the ‘linear-flow-regime’ that is the norm. Because the conceptual interpretation is based on a physics-based model, the ideas we present are both scalable and reproducible. We recover classical formulations through a subset of the model considered here. New behaviors, yet to be reported, are discussed.
Abstract Pressure- and rate-time data at wells producing the Wolfcamp shale are evaluated by a model based on a framework using subdiffusive concepts. Quantitative measures to estimate heterogeneities in the fracture- and matrix-systems are provided. Multiple transfer mechanisms and complex structures govern the dynamic performance of the reservoir. Long-term depletion is governed by the matrix system; our evaluations indicate that excellent coverage is obtained in draining the lateral extents of the reservoir rock. As a physics-based model is used to evaluate responses, the suggested procedures are both extendable and scalable.
Abstract Diagnostic Fracture Injection Test (DFIT) responses in some shale reservoirs, such as the Duvernay shale in western Canada, are not consistent with those interpreted through traditional analysis methods. Indeed, interpretation with traditional techniques may result in significantly incorrect estimates of closure pressure, pore pressure and formation permeability. The goal of this paper is to explain the observed DFIT behaviours for selected Duvernay shale wells in terms of low leakoff of fracturing fluid to the formation, activation of pre-existing fractures, and tip extension during the test. DFIT data in the Duvernay shale are analyzed using pressure transient analysis methods. Two scenarios are presented to explain the overall falloff behavior; moving-hinge closure with tip extension, and activation of secondary natural fractures. The validity of each scenario is examined using rigorous coupled flow-geomechanical simulation, geological information and geomechanical settings in the Duvernay Formation. Due to extremely low leakoff, the main mechanism affecting pressure falloff during the DFIT is pressure dissipation through the primary fracture created during injection. This results in significant tip extension or activation of secondary fractures. The fluctuations and spikes observed on G-function or pressure derivative plots are explained in the context of these scenarios. The leakoff rate varies with the pressure change, and the enhanced fracture surface area, during tip extension. Therefore, the assumption of Carter leakoff, and the traditional closure picks based on a straight-line tangent to the semi-log derivative on a G-function plot or 3/2 slope on Bourdet-derivative plot are not valid. Due to very low matrix permeability and the additional fracture length created through tip extension, it is unlikely that formation radial flow is established during the test, compromising the ability to obtain a valid pore pressure or formation permeability.
Abstract Hydraulic fracturing (HF) is a very complex engineering process. It involves rock and fluid mechanics, mixed mode rock failure and transportation of individual proppant particles. This process is multiscale both in time and spatial domains that is why it is almost impossible to create a fully coupled 3D model with a detailed description of physical and chemical processes even with significant assumptions. Recent data indicates well completion becomes more expensive than the drilling itself for several unconventional reservoirs. The reason is an increase of fluid and proppant pumped at high rates. Thus, the critical importance of engineering optimization of fracture spacing and individual pumping schedule for maximization of Net Present Value (NPV), Estimated Ultimate Recovery (EUR), and other metrics in the "lower for longer" price environment. Unfortunately, unconventional operators see little value in fracture modeling because of its complexity and amount of data required for any reasonable predictive power. That is why many companies consider geometric design as the cheapest and hence the most efficient option thus resulting in trial and error to optimize HF design. In this paper we use publicly available well logs and completion data covering the Midland Basin from the University Lands website. We also demonstrate that HF model calibration with microseismic data and stochastically generated Discrete Fracture Network (DFN), although is a challenging task, may improve our understanding of fracture design pitfalls and to become an essential step for optimum engineering design. Insights from our work can be useful to increase the predictive power of in-house models and reduce total cost and effort involved in this complex modeling.
Microfracturing is an excellent method of obtaining direct, in-situ stress measurements, not only in shales, but in conventional reservoirs as well. Recent advances have shown that microfracturing can help improve reservoir management by guiding well placement, optimizing injection rates, and managing perforation strategy.
Microfracturing consists of isolating small test intervals in a well between inflatable packers, increasing the pressure until a small fracture forms and then by conducting a few injection and shut-in cycles, extend the fracture beyond the influence of the wellbore. Results show that direct stress measurements can be successfully acquired at multiple intervals in a few hours and the vertical scale nearly corresponds to electric log resolution. Therefore, microfracture testing (generally performed in a pilot / vertical well) is an appropriate choice for calibrating log derived geomechanical models and obtaining a complete, accurate, and precise vertical stress profile.
This paper describes the microfracturing process and presents several examples that led to increased hydrocarbon recovery by efficient stimulation and/or completion design. Case studies presented range from optimizing hydraulic fracturing in unconventional (Delaware Basin, USA), determining safe waterflood injection rates in brownfields (Offshore UK), and helping improve perforation placement in ultra-deepwater reservoirs.