Patent Application
20150088479
The present invention relates to the field of hydrocarbon resource processing, and, more particularly, to a method for recovering hydrocarbons and related apparatuses.
Energy consumption worldwide is generally increasing, and conventional hydrocarbon resources are being consumed. In an attempt to meet demand, the exploitation of unconventional resources may be desired. For example, highly viscous hydrocarbon resources, such as heavy oils, may be trapped in sands where their viscous nature does not permit conventional oil well production. This category of hydrocarbon resource is generally referred to as oil sands. Estimates are that trillions of barrels of oil reserves may be found in such oil sand formations.
In some instances, these oil sand deposits are currently extracted via open-pit mining. Another approach for in situ extraction for deeper deposits is known as Steam-Assisted Gravity Drainage (SAGD). The heavy oil is immobile at reservoir temperatures, and therefore, the oil is typically heated to reduce its viscosity and mobilize the oil flow. In SAGD, pairs of injector and producer wells are formed to be laterally extending in the ground. Each pair of injector/producer wells includes a lower producer well and an upper injector well. The injector/production wells are typically located in the payzone of the subterranean formation between an underburden layer and an overburden layer.
The upper injector well is used to typically inject steam, and the lower producer well collects the heated crude oil or bitumen that flows out of the formation, along with any water from the condensation of injected steam. The injected steam forms a steam chamber that expands vertically and horizontally in the formation. The heat from the steam reduces the viscosity of the heavy crude oil or bitumen, which allows it to flow down into the lower producer well where it is collected and recovered. The steam and gases rise due to their lower density. Gases, such as methane, carbon dioxide, and hydrogen sulfide, for example, may tend to rise in the steam chamber and fill the void space left by the oil defining an insulating layer above the steam. Oil and water flow is by gravity driven drainage urged into the lower producer well.
Operating the injection and production wells at approximately reservoir pressure may address the instability problems that adversely affect high-pressure steam processes. SAGD may produce a smooth, even production that can be as high as 70% to 80% of the original oil in place (OOIP) in suitable reservoirs. The SAGD process may be relatively sensitive to shale streaks and other vertical barriers since, as the rock is heated, differential thermal expansion causes fractures in it, allowing steam and fluids to flow through. SAGD may be twice as efficient as the older cyclic steam stimulation (CSS) process.
Many countries in the world have large deposits of oil sands, including the United States, Russia, and various countries in the Middle East. Oil sands may represent as much as two-thirds of the world's total petroleum resource, with at least 1.7 trillion barrels in the Canadian Athabasca Oil Sands, for example. At the present time, only Canada has a large-scale commercial oil sands industry, though a small amount of oil from oil sands is also produced in Venezuela. Because of increasing oil sands production, Canada has become the largest single supplier of oil and products to the United States. Oil sands now are the source of almost half of Canada's oil production, while Venezuelan production has been declining in recent years. Oil is not yet produced from oil sands on a significant level in other countries.
U.S. Published Patent Application No. 2010/0078163 to Banerjee et al. discloses a hydrocarbon recovery process whereby three wells are provided: an uppermost well used to inject water, a middle well used to introduce microwaves into the reservoir, and a lowermost well for production. A microwave generator generates microwaves which are directed into a zone above the middle well through a series of waveguides. The frequency of the microwaves is at a frequency substantially equivalent to the resonant frequency of the water so that the water is heated.
Along these lines, U.S. Published Patent Application No. 2010/0294489 to Dreher, Jr. et al. discloses using microwaves to provide heating. An activator is injected below the surface and is heated by the microwaves, and the activator then heats the heavy oil in the production well. U.S. Published Patent Application No. 2010/0294488 to Wheeler et al. discloses a similar approach.
U.S. Pat. No. 7,441,597 to Kasevich discloses using a radio frequency generator to apply radio frequency (RF) energy to a horizontal portion of an RF well positioned above a horizontal portion of an oil/gas producing well. The viscosity of the oil is reduced as a result of the RF energy, which causes the oil to drain due to gravity. The oil is recovered through the oil/gas producing well.
U.S. Pat. No. 7,891,421, also to Kasevich, discloses a choke assembly coupled to an outer conductor of a coaxial cable in a horizontal portion of a well. The inner conductor of the coaxial cable is coupled to a contact ring. An insulator is between the choke assembly and the contact ring. The coaxial cable is coupled to an RF source to apply RF energy to the horizontal portion of the well.
Unfortunately, long production times, for example, due to a failed start-up, to extract oil using SAGD may lead to significant heat loss to the adjacent soil, excessive consumption of steam, and a high cost for recovery. Significant water resources are also typically used to recover oil using SAGD, which impacts the environment. Limited water resources may also limit oil recovery.
Hydrocarbon resources may also be removed from other subterranean formations, such as oil shale, for example. Oil shale (kerogen shale) is sedimentary rock containing hydrocarbons, i.e. shale oil. One typical approach to shale oil extraction is to heat the oil shale to vaporize the shale oil, which can be subsequently collected. Nevertheless, this approach can he quite costly, especially when compared to other hydrocarbon resource extraction methods. It is desirable to more efficiently recover these hydrocarbon resources.
In view of the foregoing background, it is therefore an object of the present invention to provide a method for hydrocarbon recovery that is efficient and robust.
This and other objects, features, and advantages in accordance with the present invention are provided by a method for predicting hydrocarbon recovery in a subterranean formation. The method may comprise generating pre-heating data associated with fracturing in a sample from the subterranean formation, determining at least one first fractal pattern based upon the pre-heating data, and heating the sample with RF power to cause additional fracturing in the sample. The method also may include generating post-heating data associated with additional fracturing in the sample after heating with RF power, determining at least one second fractal pattern based upon the post-heating data, detecting change between the at least one first fractal pattern and the at least one second fractal pattern, and predicting hydrocarbon recovery from the subterranean formation based upon the detected change. Advantageously, the predicted hydrocarbon recovery is accurate and may reduce recovery costs.
More specifically, the detecting of change may comprise detecting a volumetric change in fracturing. The pre-heating data may comprise three-dimensional mapping pre-heating data, and the post-heating data may comprise three-dimensional mapping post-heating data. The determining of the at least one first fractal pattern may be based upon at least one neighborhood of voxels in the three-dimensional mapping pre-heating data, and the determining of the at least one second fractal pattern may be based upon at least one neighborhood of voxels in the three-dimensional mapping post-heating data.
In some embodiments, the generating of the pre-heating data and post-heating data may comprise using computed tomography (CT) scanning. The detecting of the change may comprise performing mutual information calculations. Each fractal pattern may comprise an associated resolution and an associated fractal pattern type.
Additionally, the method may comprise generating a fractal pattern look-up table (LUT) including a plurality of values for the at least one first and the at least one second fractal patterns. The method may also comprise displaying at least one image based upon the at least first and at least one second fractal patterns. The method may also include applying RF power to the subterranean formation based upon the predicted hydrocarbon recovery, and recovering hydrocarbons from the subterranean formation after applying RF power thereto.
Another aspect is directed to an apparatus for predicting hydrocarbon recovery in a subterranean formation from a sample thereof. The apparatus may include an RF antenna, a sensor, and a processor coupled to the RF antenna and the sensor and configured to generate pre-heating data associated with fracturing in the sample from the sensor, and determine at least one first fractal pattern based upon the pre-heating data. The processor may also be configured to heat the sample with RF power from the RF antenna to cause additional fracturing in the sample, generate post-heating data, via the sensor, associated with additional fracturing in the sample after heating with RF power, determine at least one second fractal pattern based upon the post-heating data, detect change between the at least one first fractal pattern and the at least one second fractal pattern, and predict hydrocarbon recovery from the subterranean formation based upon the detected change.
The present invention will now be described more fully hereinafter with reference to the accompanying drawings, in which preferred embodiments of the invention are shown. This invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art. Like numbers refer to like elements throughout, and prime notation is used to indicate similar elements in alternative embodiments. For example, although descriptions of the pre and post-RF fractures are all RF heating related, the present disclosure is envisioned to include any induced deformation (pressure fractures, thermally induced fractures, etc.).
Referring initially to
The apparatus 30 includes an RF antenna 33, a sensor 34, and a processor 31 coupled to the RF antenna and the sensor. The apparatus 30 also illustratively includes a display 32 coupled to the processor 31. In some embodiments, the sensor 34 may comprise a CT device comprising an X-ray source for directed radiation to the sample 35, and a radiation sensor for detecting reflected X-ray radiation from the sample. Of course, in other embodiments, the sensor 34 may utilize other forms of radiation and/or particles. The radiation sensor may be positioned opposite the X-ray source in some embodiments.
Referring now additionally to
The method also includes a data conversion (Block 42) where the data ingested from the sensor 34 is converted to an image file format. For example, the sensor 34 may output data in a Digital Imaging and Communications in Medicine (DICOM) format, and the conversion changes such data to a bitmap image file or a raster image file. The method includes a volumetric processing at Block 43.
The method includes generating pre-heating data associated with fracturing in a sample 35 from the subterranean formation. In other words, the sample 35 is examined for a baseline state via the sensor 34, so that subsequent states are readily comparable. The method also includes heating the sample 35 with RF power from the RF antenna 33 to cause additional fracturing in the sample. As will be appreciated, the RF heating causes the hydrocarbon resources (e.g. shale oil) in the sample 35 to expand and cause fractures therein.
Subsequent to the RF heating, the method includes generating post-heating data (via the sensor 34) associated with additional fracturing in the sample 35 after heating the sample with RF power. More specifically, the pre-heating data comprises three-dimensional mapping pre-heating data (e.g. point cloud data), and the post-heating data comprises three-dimensional mapping post-heating Hata.
Both the pre-heating data and post-heating data are processed though Blocks 41-43, and at Block 44, the method includes detecting change between the pre-heating data and post-heating data. The detecting of change may comprise performing mutual information calculations. In the illustrated embodiments, the detecting of change comprises detecting a volumetric change in fracturing.
The method further comprises displaying at least one image based upon the pre-heating data and the post-heating data (Blocks 45, 57). In some embodiments, the method includes displaying only changes between the pre-heating data and the post-heating data. The displaying of the change data may be slice based or as a volume (i.e. multiple slices).
The method also includes predicting hydrocarbon recovery from the subterranean formation based upon the detected change. In short, by leveraging change detection processes, the apparatus 30 is able to provide a prediction of the commercial viability of large scale extraction of hydrocarbons from the subterranean formation by performing scanning on the sample 35.
Advantageously, the predicted hydrocarbon recovery is accurate and may reduce recovery costs. In some embodiments, the method may comprise applying RF power to the subterranean formation based upon the predicted hydrocarbon recovery, and recovering hydrocarbons from the subterranean formation after applying RF power thereto (i.e. actual extraction process). The disclosed method provides an automated method for modeling shale fractures as a result of RF wave propagation techniques. Also, the method may show a qualitative two-dimensional (2D) and three-dimensional (3D) visualization of fractures.
The disclosed image processing algorithms automatically extract fracture features accurately in 3D to allow for efficient evaluation. Quantitative results are discussed using statistical attributes as well as using 2D plots for analysis. The method may quantify the increase in the fracture network due to RF heating and provide a method for change detection and analysis. By using statistical analysis, the relationship between fractures and increased oil recovery can be learned.
The disclosed approach can be used to analyze shale structure and infer increased oil extraction to increase cost effectiveness. In shale, natural fractures provide permeability and access to the oil and gas in the well, and inducing fractures helps improve oil and gas recovery in a variety of shale formations and other formations containing shale layers. Inducing fractures allows for the liquid to flow freely and therefore can be extracted easier and with less expense. Shale has low permeability, so gas production in high quantities requires fractures to provide permeability.
The apparatus 30 ingests 3D shale images using 3D scanning from samples taken before and after RF exposure. Change detection between these samples is quantified through calculating mutual information. The design of this apparatus 30 is based on the assumption that the changes that are interesting are distributed irregularly and throughout an area of interest. In short, the apparatus 30 provides an approach to the problem of modeling and visualizing shale fracture structures, and providing an approach to accurately predicting the amount of hydrocarbon resources that can be extracted.
Referring now additionally to
The method includes ingesting the volume slices from the sensor 34 (Block 51). For example, each volume slice may comprise an 8-bit image slice. Images 60, 65 illustrate exemplary volume slices at both pre-heating and post-heating, respectively. As can be seen in image 65, the number of fractures has increased. The method further comprises performing at least one filter process on each of the pre-heating data and post-heating data (Block 52). In the illustrated embodiment, the at least one filter process comprises a median filter process.
The method includes generating the pre-heating data by performing at least one first edge detection process, and generating the post-heating data by performing at least one second edge detection process (Block 53). In some embodiments, the at least one first and at least one second edge detection processes each comprises detecting edges of the fracturing and the additional fracturing in the sample, masking borders in the pre-heating data and post-heating data, and performing a gamma correction operation on the pre-heating data and post-heating data.
The generating of the pre-heating data further comprises performing at least one first blob filter process after the at least one first edge detection process, and the generating of the post-heating data further comprises performing at least one second blob filter process after the at least one second edge detection process (Block 54). The data is then subjected to a conversion at Block 55, where the plurality of slices is converted into a volume (Block 62).
Images 70, 75, 80, 85 illustrate the pre-heating data during different stages of the volumetric processing (Block 43). Image 70 represents the ingested volume slice, and image 75 represents the volume slice after the at least one first edge detection process. Image 80 represents the volume slice after the at least one first blob process, and image 85 represents the finished volume (assembled volume slices).
Images 90, 95, 100, 105 illustrate the post-heating data during different stages of the volumetric processing (Block 43). Image 90 represents the ingested volume slice, and image 95 represents the volume slice after the at least one second edge detection process. Image 100 represents the volume slice after the at least one second blob process, and image 105 represents the finished volume (assembled volume slices). Advantageously, the finished volume images 85, 105 clearly demonstrate the detected change in the sample 35.
Referring now additionally to
As shown in Tables 1 and 2 herein, the quantitative results of the scanning demonstrate that post-heating has 76090/6122=12.3 times more structure than pre-heating, or that the pre-heating volume has only 8.1% the amount of structure compared with post-heating volume. Each voxel in the shale sample has a length of 173.63 microns.
All of the X-ray related settings are the same for all CT scans. All settings (such as gain, X-ray energy etc.) are the same for all samples. Post-heating volumes have 37,004/313,282=8.5 times more structure on average, or pre-heating volumes have only 12% the amount of structure compared with post-heating volumes. Table 3 below shows some detailed data regarding sample 1.
Referring now additionally to
Table 4 provides more quantitative results for samples 1-4, including both pre-heating and post-heating.
The quantitative results for fracture size and vertical length are on a per fracture basis. The size of the fracture is measured in a number of edge voxels. The vertical distance of a fracture is measured in length in voxels. For all of the sample shale data volumes, the size and vertical length increased with RF energy applied.
Table 5 shows data relating to the use of mutual information as a measure of change or similarity for samples 1-4. The mutual information value increases with increasing similarity. As the statistical relationship between the two vectors becomes more dissimilar, the joint entropy, H(X, Y), increases. The vectors X and Y correspond to histograms with twenty bins each for fracture volumetric size and fracture vertical length. The most change occurred between the pre/post RF fracturing in sample #3, since it has the lowest value for mutual information.
Referring now additionally to
Also, this apparatus 30′ may be used with a method for predicting hydrocarbon recovery in a subterranean formation. The method comprises generating pre-heating data associated with fracturing in a sample 35′ from the subterranean formation, determining at least one first fractal pattern based upon the pre-heating data, and heating the sample with RF power to cause additional fracturing in the sample. The method also includes generating post-heating data associated with additional fracturing in the sample 35′ after heating the sample with RF power, determining at least one second fractal pattern based upon the post-heating data, detecting change between the at least one first fractal pattern and the at least one second fractal pattern, and predicting hydrocarbon recovery from the subterranean formation based upon the detected change. Of course, the determining of the at least one first and second fractal patterns is based upon a plurality of potential fractal patterns.
The determining of the at least one first fractal pattern is based upon at least one neighborhood of voxels in the three-dimensional mapping pre-heating data, and the determining of the at least one second fractal pattern is based upon at least one neighborhood of voxels in the three-dimensional mapping post-heating data. In some embodiments, each fractal pattern comprises an associated resolution and an associated fractal pattern type. The method also comprises displaying at least one image based upon the at least first and at least one second fractal patterns.
Referring now additionally to
Advantageously, the fractal pattern LUT replaces the 3D data ingested from the sensor 34′. Indeed, the data is compressed via this conversion to the fractal pattern LUT. Once compressed, the data may be transmitted off site, for example, in in situ applications (Block 180). Once the compressed data is received, the method includes ingesting the compressed data, and expanding the data using the fractal pattern LUT (Blocks 181-182). The method includes decompressing the data, and extracting voxels to regenerate the original 3D point cloud data (Blocks 183-184, 186).
An exemplary fractal pattern LUT is shown herein as Table 6.
Referring now to
Referring now additionally to
In the following, a discussion of exemplary embodiments follows.
Change detection is a key task in remote sensing studies where temporal data is analyzed to detect changes over a period of time. Some applications of remotely sensed change detection related to Earth sciences include environmental impact assessment, flood damage assessment and land-use and land-cover change analysis. Assessment of change is also important for military applications like surveillance of a particular area, target detection, and damage assessment. The basic principle of change detection from remote sensing images is based on the difference in reflectance or intensity values between the images taken at two different times due to changes on the Earth's surface. Some commonly used image change detection algorithms are image differencing, image ratioing, image regression, principal component analysis (PCA), and change vector analysis.
A typical requirement for change detection from remote sensing images is the accurate registration of temporal images. In other words, the images must be aligned with each other such that corresponding locations in the images are present at identical pixel positions. Any registration errors (or mis-registration) present in the images may lead to incorrect change detection.
The present disclosure discusses a robust method for quantifying change without requiring affine registration, consisting of six degrees of freedom, three translations and three rotations. The present disclosure considers mutual information in the fractal domain as the change detection method since it takes advantage of the fact that the entropy of a discrete distribution is invariant to rotation and translation.
The present disclosure demonstrates a method for shale fracturing, modeling, and change analysis. The present disclosure uses radio wave heating to penetrate oil shale to heat up water deposits to cause fractures. Modeling the size and orientation of these fractures can be used to infer financial benefits and increased oil production. The present disclosure discusses a method to assist analysts by using 3D image processing and visualization. The method of the present disclosure provides quantitative change detection metric using mutual information. Graphical 2D plots as well as 3D volumetric rendering of fractures provide qualitative change detection.
The lab setup for the disclosed process uses computed tomography (CT) as the remote sensor to collect the volumetric data for the shale samples. This lab setup is scalable for real world applications and serves as a demonstration. The present disclosure's goal is to extract the shale fractures before and after RF energy is applied to the shale samples. The present disclosure's goal is to quantify the number, size, and vertical distance of the fractures. The presence of fractures is an important for optimizing recoverable oil to be extracted in an area of interest.
The present disclosure workflow overview to extract shale fractures is shown in
Volumetric processing consists of a 3D edge detector and blob filter. The blob filter finds all fractures with a given intensity threshold that are connected with a minimum size. The blob filter considers all eight neighbors in the local neighborhood around each voxel in the fracture. Extracted fractures can be exported to all the standard 3D formats for visualization. The current total processing time on SUN Blade Dual Processor is less than eight hours per 512×512×512 voxel space. However processing time with a general purpose graphics processing unit (GPGPU) is much faster. Using GPGPU processing, the current achieved fractal finding and pattern labeling operation with 100 million points is approximately 30 seconds.
The main components of the flow exploit the 3D edge and 3D blob filters. The edge detector looks in three dimensions to return edge strength for each voxel in the feature space. The blob filter finds all fractures with a given intensity threshold that are connected with a minimum size. The blob filter considers all eight neighbors in the local neighborhood around each voxel in the fracture. Extracted fractures can be exported to all the standard 3D formats for visualization.
The present disclosure has results showing that the number of fracture points increases significantly when applying RF energy to the shale samples. Table 1 shows the number of points for the fractures that have been extracted and modeled. Post-RF has 760906/61652=12.3 times more structure or Pre-RF volume has only 8.1% the amount of structure compared with Post-RF volume. Each voxel in the shale sample has a length of 173.63 microns.
Visualizations of these volumes are shown in
Fractures have a big impact on oil production but are inherently difficult to quantify. The flow rate is dependent on the permeability of the shale. Desirable attributes are length, spacing, and connectivity. Table 2 shows quantitative results for fracture structure before and after RF energy applied. All of the X-ray related settings are the same for all CT scans. All settings (such as gain, x-ray energy etc.) are the same for all samples. The Pre-RF volumes have only about 11% the amount of structure compared with Post-RF volumes.
Table 4 shows quantitative results for several shale samples. Fracture mean and standard deviation of size and vertical length are shown on a per fracture basis. The size of a fracture is measured in number of edge voxels. The vertical distance of a fracture is measured in length in voxels. For all of the sample shale data volumes, the size and vertical length increases with RF energy applied.
One method for detecting change or similarity can be accomplished through the application of a mutual information algorithm. Mutual information (MI) between two variables is based on information theory. It measures the amount of information between two variables. The use of MI as a similarity measure on computed tomography data has been performed on medical images. Mutual information considers the joint entropy and the individual entropies and is defined as:
I(X;Y)=H(X)+H(Y)−H(X,Y); (1)
where I is mutual information and H is entropy. When the data vectors are more similar, a lower numerical value is obtained for joint entropy. The mutual information value increases with increasing similarity. As the statistical relationship between the two vectors becomes more dissimilar, the joint entropy, H(X, Y), increases. The vectors X and Y in the experiments correspond to histograms with twenty bins each for fracture volumetric size and fracture vertical length. Entropy is defined as:
H(pi)=Σpi log(pi); (2)
where p is the probability vector for which a uniform distribution is assumed. A good implementation in Matlab is available. The present disclosure's solution for calculating mutual information makes use of fractal analysis of the extracted shale fractures. This is based on spatial information of patterns in a local neighborhood.
Table 5 shows results of calculating the mutual information and entropy for the fractal metrics of volume and vertical distance. The volume of a fracture is really the count of its edge voxels. The most change occurred between the pre/post RF fracturing in sample #3 since it has the lowest value for mutual information.
Note that a lower MI value, when comparing the pre/post RF volumes, means that the volumes are very different and quantifies the amount of fracturing performed. Recently there has been some work on calculating mutual information from 2D images by directly considering spatial information in local neighborhood regions.
Three dimensional scanning systems can automatically measure a large number of points forming the surface of the earth, or some other object, and can output a point cloud as a data file. The point cloud represents the set of points in a three dimensional space that the device has measured with respect to a set of coordinate axes. The present disclosure is interested in extracting the point cloud describing the shale fracture. This is accomplished by using the points from the 3D blob filter.
The present disclosure use of fractals is based on self-similarity. The flow diagram is shown in
From this basic element, larger patterns can be constructed. As the 2×2×2 fractal block is passed through the entire volumetric scene, similar voxel (3D pixel) patterns are identified. Here, the data are true 3D point clouds with a range or Z value.
The present disclosure's process requires considerable compute capability, such as GPGPU architecture to find the fractal patterns at multiple resolutions in near real-time. There are several formats where this method could be used besides 3D points or vertices which can be in a Common LiDAR Data Exchange (.LAS) as well as ASCII or binary format. Some popular 3D graphic polygonal formats are: Virtual Reality Modeling Language (VRML, GeoVRML), Drawing Exchange Format (DXF), Shapefiles, OpenFlight, Maya Computer Aided Design (CAD), Stereo-Lithography (STL), and Keyhole Markup Language (KML). The present disclosure's process allows for routine data conversions between the above formats listed.
GPUs are assuming a more prominent role in PC system architecture. GPUs deliver more cost-effective and energy-efficient performance for applications that need it. NVIDIA's next-generation CUDA architecture (code named Fermi), is the latest and greatest expression of this trend. With many times the performance of any conventional CPU on parallel software, and features to make it easier for software developers to realize the full potential of the hardware, Fermi-based GPUs will bring supercomputer performance to more users than ever before. GPUs have become more flexible and programmable such as NVIDIA's CUDA which is considerably faster than a CPU based algorithm. Each Fermi GPU is equipped with an L2 cache (768 KB in size for a 512-core chip). The L2 cache covers GPU local DRAM as well as system memory.
The Fermi generation CUDA architecture is the most advanced GPU computing architecture ever built. With over three billion transistors and featuring up to 512 CUDA cores, Fermi delivers supercomputing features and performance at 1/10th the cost of traditional CPU-only servers.
Different resolutions of data within the point cloud can be identified by building upon the 2×2×2 fractal pattern.
The use of a lookup table provides for an efficient bookkeeping method for each fractal position and pattern type. By partitioning the fractal block into eight different zones, regardless of the resolution level, it is possible to efficiently track the levels of self-similarity in a code book. Since there are only 256 different patterns in a 2×2×2 block, it is relatively straightforward to use a binary set of numbers to indicate the presence or absence of a voxel point, as shown in
From this binary representation, the look-up table is generated. The table contains the position of the fractal, the pattern type, and the level of repeatedness of the pattern. Note that with this information, for a level 3 fractal pattern, the amount of data represented is 8×8×8 or 512 individual LiDAR points. A part of a sample lookup table is shown in Table 6.
The first three columns contain the row, column, and height coordinates. The fourth, fifth and sixth columns show the fractal pattern for fractal levels one, two, and three. The present disclosure's system contains up to 256 possible fractal patterns. According to the lookup table, a zero placeholder is inserted in a column if there exists a higher fractal dimension for that position. For example the fourth row in the table contains a level three fractal pattern (8×8×8) of a level 3 pattern of type eleven at the x,y,z location of 291, 469, 97.
Decoding is required for locating the voxels that make up the identified fractal pattern. Efficient decoding is accomplished by expanding the fractal data patterns back to the original x, y, and z format. Each fractal in the look-up table consists of an x, y, and z position as well as the level of fractal repeated patterns found during the compression process. Since the position and pattern type is known, the pattern decoding is much faster than encoding since the information.
The results shown in
Fractal geometry can be useful for explaining the fracture behavior and rock properties. It was found that a peak fractal dimension value exists and corresponds to porosity. Previous studies involve correlating the fractal dimension of fractures. The information can be useful to earthquake studies. Fractal dimension describes the irregularity of the surface roughness. A higher fractal dimension corresponds to rougher fracture surface and is relevant to hydraulic fracturing of oil, gas or geothermal wells.
The fractal fracture networks in rock possess some structured features which have high cumulative production and better drainage patterns. In shale, natural fractures provide permeability and access to the oil and gas in the well. By using statistical analysis, the relationship between fractures and increased oil recovery can be learned.
The present disclosure presents a method for demonstrating and quantifying the change in shale fracturing due to the RF wave propagation. The method of the present disclosure allows a field engineer to make robust change detection calculations without the need to register the data. The data may be collected at appropriate periods to determine a cost effective stopping point since MI is correlated to cost savings. So far, processing on several different shale samples has been performed. The present disclosure's fractal method minimizes the amount of data that needs to be considered while incorporating positional and regional neighborhood data.
The present disclosure's RF antenna and CT scanner were used as a sensor to collect volumetric data of shale samples. The use of 3D blob filter extracts the significant individual fractures of interest. The present disclosure uses mutual information to quantify change for pre/post RF shale fracturing volumes. Modeling of fracturing techniques can be used to estimate cost savings and increased recoverable barrels from potential oil reserves. The present disclosure's process increases the energy returned on energy invested (EROI) by assisting with calculating when the RF heating requirement is satisfied. Future work would be to perform calculations on increase in oil liquid flow rate due to fracturing.
Other features relating to hydrocarbon extraction are disclosed in co-pending application: titled “METHOD FOR HYDROCARBON RECOVERY WITH CHANGE DETECTION AND RELATED APPARATUS,” Attorney Docket No. GCSD-2663 (62030), incorporated herein by reference in its entirety.
Many modifications and other embodiments of the invention will come to the mind of one skilled in the art having the benefit of the teachings presented in the foregoing descriptions and the associated drawings. Therefore, it is understood that the invention is not to be limited to the specific embodiments disclosed, and that modifications and embodiments are intended to be included within the scope of the appended claims.
