The downhole monitoring of strain using Fiber Optics (FO) can reveal unique information about the propagation and geometry of hydraulic fractures between nearby wells during stimulation and production. This work aims at creating a catalogue of commonly observed strain-rate signals captured in a not yet stimulated nearby observation well equipped with either a permanently or temporarily installed FO cable. This catalogue is the result of an informal collaboration between experience FO users from academia, service providers, consulting companies, and operators. In the creation of this first edition of a strain-rate catalogue, we considered two main types of stimulation categories (single and multi-entry) as well as the angle between the hydraulic fractures and the segment of the well where the strain-rate signals are observed (horizontal vs. vertical segments). In the catalogue we show a series of representative examples of two main types of far-field strain Fracture Driven Interactions (s-FDI) commonly encountered in frac diagnostics: 1. Vertical hydraulic fractures being monitored in a lateral portion of a horizontal well and 2. Vertical fractures being monitored in a vertical observation well. The catalogue is organized around commonly observed s-FDI motifs. Because interpretation of observed strain-rate signals can be subjective, when possible, we included observed examples with a brief description of our interpretation, as well as synthetic signals from geomechanical models of similar motifs. The strain-rate motifs were modeled based on first physical principles for rock deformation. These models serve to support the proposed interpretation of the observed signals. FO strain rate monitoring is changing our understanding about the hydraulics fracturing process. The information from FO strain is not available by other commonly used fracture diagnostic techniques. Strain- rate fractures driven interactions between wells occur in predictable patterns (Frac Domain and Stage Domain Corridors – FDC & SDC respectively) which are typically in line with the cluster spacing and stage length in the borehole being stimulated. Using FO strain monitoring, we now know that hydraulic fractures are larger than first anticipated, both in length and height. Many examples indicated that there is a direct correspondence between the near-field and far-field stimulation geometries. The lack of isolation due to cement quality and or plug failure manifests in the far-field geometries observed via FO strain-rate in nearby wells. The use of FO strain monitoring has also revealed that reopening of hydraulic fractures is common not only between prior and infill wells but also between wells from the same stimulation vintage. All these observations and conditions must be considered when interpreting new strain-rate datasets and more importantly when designing new hydraulic fracturing operations and considering different stimulation order (zipper schedule), as well as when making decisions about the vertical and lateral spacing of adjacent wells. The purpose of this industry-first edition strain-rate catalogue is to aid, new and experienced FO users, on the interpretation of strain-rate datasets. Ultimately, the accurate interpretation of FO strain data will not only help calibrate geomechanical and reservoir models but also directly influence where and how we complete unconventional wells. Nowadays, many s-FDI examples exist in scattered publications with formats that aren’t easily comparable for new users of the technology. In this project, we expand upon those publications to create an encompassing analysis with up-to-date interpretations where we have formalized the formatting of figures for better readability (color scheme, scales, etc.). What has resulted from this collaborative effort is a novel catalogue not available before in the FO published literature.
Summary Low-frequency distributed acoustic sensing (LF-DAS) has recently received much attention for its ability to monitor fracture propagation at offset wells. Hydraulic Fracture Test Site-2 (HFTS-2), a Department of Energy–sponsored field-based research experiment, has acquired LF-DAS data sets during the stimulation of many horizontal wells in the Wolfcamp Formation of the Permian Basin. Over 100 stimulated stages with different completion designs in four horizontal wells were monitored by two horizontal offset wells and one vertical pilot hole with permanent fibers. The parent well depletion affected all four horizontal wells in almost half of their lateral section. Several studies have been performed on the acquired comprehensive data set. In this study, we apply our novel Green’s function-based inversion algorithm to calculate the fracture width of each stage and investigate the impact of parent well depletion on fracture geometry. The Green’s function-based inversion algorithm has been successfully applied to a few stages of field case studies. The Green’s function matrix was built based on linear elasticity theory (3D displacement discontinuity method) to relate the fracture openings with strain responses measured along the length of the fiber. This novel algorithm only relies on measured strain to solve fracture geometry at the monitoring well. Therefore, it is independent of the physics of the fracture propagation process and can be used to validate hydraulic fracture modeling results. Using our inversion algorithm, we can efficiently and quantitatively interpret LF-DAS data to provide information on fracture geometry and completion efficiency. We analyze more than 100 stages of the LF-DAS measurements obtained at the fiber wells B3H and B4H during B1H, B2H, and B4H stimulations. We apply our inversion algorithm for four data sets covering the stimulation of the abovementioned three wells. The fracture growth at stages in the parent well’s depleted zone is biased more toward upper formation and in the easterly direction. The fracture width at the stages in the parent well’s depletion zone is reduced compared to fracture widths at stages in the nondepleted zone irrespective of the monitoring well location relative to the treatment wells. This difference in fracture widths will affect the proppant distribution to a great extent, thereby affecting the effectiveness of the stimulation. We also illustrate the application of the inversion algorithm for stages that have both conventional fracture hits with “heart-shaped” signals as well as fracture reopening signals. Our inversion algorithm gives a reasonable estimate of the fracture width that aligns with the qualitative analysis of microseismic data sets and statistic summary of fracture hit numbers of HFTS-2. We believe that this quantitative study provides us with insights into the fracture geometry due to parent well depletion effects.
Summary Distributed acoustic sensing (DAS) has recently gained importance in monitoring hydraulic fracturing treatments in the oil and gas industry. DAS data contain critical information about the fracture geometry as linearly relatable induced strain variations during the stimulation. The low-frequency components of the DAS (LF-DAS) data are known for their complexity as they exhibit various characteristic signals—caused by several mechanisms—that complicate their interpretation. LF-DAS data from horizontal monitoring wells (HMWs) have been used to detect fracture hits and characterize fracture geometry. However, the LF-DAS data from vertical monitoring wells (VMWs) have not been studied extensively as a means to infer fracture geometry. The major limitation of VMWs is the number of monitored stages, but the data contain more information about fracture height compared with LF-DAS measurements from HMWs. Hence, it is necessary to have a physical rock deformation model to simulate the strain rate responses in offset VMWs during fracture propagation to understand and interpret the various patterns that are observed in the field data sets. The objective of this study is to simulate strain rate signals in VMWs during hydraulic fracturing and to analyze the measurements to obtain information on the fracture geometry, especially the fracture height. The fracture boundary can be directly related to the strain rate signals. In this study, we propose a workflow to determine fracture height at different fiber-to-fracture (dff) distances for fracture heights ranging from 20 m to 300 m. We conduct a detailed sensitivity analysis to understand the impacts of the dff, the perforation location, the fracture passing time, and the well inclinations on the measured strain rate signals. The analysis helps interpret the various patterns observed in field data and the underlying mechanisms. Interpretation of field data from the Hydraulic Fracture Testing Site 2 (HFTS-2) using the results from our forward physical model provides valuable information on the fracture characteristics that can be captured by the physical model. The results of this study are expected to provide better interpretations of LF-DAS signals from VMWs.
Low-frequency distributed acoustic sensing (LF-DAS) has recently received much attention for its ability to monitor fracture propagation at offset wells. Hydraulic Fracture Test Site-2 (HFTS-2), a DOE- sponsored field-based research experiment, has acquired LF-DAS datasets during the stimulation of many horizontal wells in the Wolfcamp Formation of the Permian Basin. Over 100 stimulated stages with different completion designs in four horizontal wells were monitored by two horizontal offset wells and one vertical pilot hole with permanent fibers. The parent well depletion affected all four horizontal wells in almost half of their lateral section. Several studies have been done on the acquired comprehensive dataset. In this study, we apply our novel Green's function-based inversion algorithm to calculate the fracture width of each stage and investigate the impact of parent well depletion on fracture geometry. The Green's function-based inversion algorithm (Liu et al., 2021) has been successfully applied to a few stages of field case studies. The Green function matrix was built based on linear elasticity theory (Three- Dimensional Displacement Discontinuity Method) to relate the fracture openings with strain responses measured along the length of the fiber. This novel algorithm only relies on measured strain to solve fracture geometry at the monitoring well. Therefore, it is independent of the physics of the fracture propagation process and can be used to validate hydraulic fracture modeling results. Using our inversion algorithm, we can efficiently and quantitatively interpret LF-DAS data to provide information on fracture geometry and completion efficiency. We analyze more than 100 stages of the LF-DAS measurements obtained at the fiber wells B3H and B4H during B1H, B2H, and B4H stimulations. We apply our inversion algorithm for four datasets covering the stimulation of the above mentioned three wells. The fracture growth at stages in the parent well depleted zone is biased more towards upper formations and in the east direction. The fracture width at the stages in the parent well depletion zone is reduced compared to fracture widths at stages in the non- depleted zone irrespective of the monitoring well location relative to the treatment wells. This difference in fracture widths will affect the proppant distribution to a great extent, thereby, affecting the effectiveness of the stimulation. We also illustrate the application of the inversion algorithm for stages that have both conventional fracture hits with "heart-shaped" signals as well as fracture reopening signals. Our inversion algorithm gives a reasonable estimate of the fracture width that aligns with the qualitative analysis of microseismic datasets and statistic summary of fracture hit numbers of HFTS-2. We believe that this study provides us with insights into the fracture geometry due to parent well depletion effects quantitatively.
Fiber Optic monitoring has become invaluable for analyzing stimulation effectiveness and fracture geometries in unconventional reservoirs. Distributed Acoustic Sensing (DAS) is a Fiber Optic-based technique that provides dynamic strain variations along the length of the fiber during well stimulation. The strain measurements can provide critical information about the fracture characteristics. Most of the studies on Low-Frequency DAS (LF-DAS) focus on interpreting data measurements from horizontal monitoring wells. The LF-DAS measurements from vertical monitoring wells can provide valuable information about fracture propagation. However, strain response measurement from vertical monitoring wells is not fully understood. Therefore, it is necessary to model LF-DAS signals from vertical offset wells to understand the mechanisms and fracture characteristics. In this paper, we use a complex hydraulic fracture simulator to generate dynamic fracture propagation and obtain strain variations during fracture propagation at vertical monitoring wells located at different offsets. The mechanism of various strain-rate patterns with fiber-to-fracture distances is illustrated using the spatial distribution of induced stresses during stimulation. A 1D feature, the absolute sum of strain rate along the fiber length is developed to determine the fracture passing time. This study provides valuable insights into the LF-DAS signals obtained from vertical monitoring wells and helps to better interpret the field LF-DAS measurements.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.