Combined microseismic mapping and fiber-optic sensing assess fracture effectiveness in the Barnett shale
Using both analysis tools allowed the operator to determine how near-wellbore conditions influenced the overall fracture geometry.
ERIC HOLLEY, Pinnacle, a Halliburton service; and ROBERT HULL, Pioneer Natural Resources
Since the advent of hydraulic fracturing, significant efforts have been made to discover what is actually occurring when fluids are injected into a target reservoir. Various technologies are used to determine the dimensions, orientation and conductivity of a hydraulically induced fracture. Direct and near-wellbore techniques use temperature logs and radioactive tracers to obtain data about fracture height and proppant placement. Beyond direct measurements made within the wellbore, tiltmeter and microseismic data may be used to monitor the orientation and dimensions of hydraulic fractures. These in-situ reservoir studies, away from the wellbore, have the potential to image the areal extent of the fracture network as it develops through time. New technologies used to estimate stimulation effectiveness include fiber-optics-based distributed temperature and acoustic sensing tools. Each of these analysis tools has its own set of strengths and weaknesses and requires its own assumptions to provide valid and confident results. A combined analysis using both microseismic fracture mapping and fiber-optic distributed temperature sensing (DTS) was conducted for a single well completed in the Barnett shale. The lateral well was completed in 2007 using perforations for reservoir entry and valves for isolation between treatments. A microseismic dual-geophone array was used, which included a vertical monitoring well to the southwest of the treatment lateral. On the well studied, fiber-optics was used only to acquire frac fluid injection data, not production data. The analysis performed on data from the first five frac stages completed in the well helped the operator to better evaluate both the near-wellbore environment and far-field fracture geometries, and to identify correlations between the two environments. The analysis demonstrated the usefulness of combining microseismic and DTS data both to improve the understanding of the current completion’s effectiveness and to help the operator ascertain the proper adjustments needed to optimize future stimulations.
INTRODUCTION Over the past decade, microseismic mapping has become a prevalent method to understand the far-field geometry of a hydraulic fracture. It is regarded as a cost-effective approach for understanding and quantifying stimulation effectiveness and interpreting fracture geometries with reasonable confidence. Microseismic events take place when the normal stress is reduced along pre-existing planes of weakness until shear slippage occurs. When a reservoir is hydraulically fractured, these emissions can be monitored and the arrival of their waves can be used to locate where changes in the fluid pressure indicate reservoir stimulation. The acoustic emission travels through the reservoir and is then recorded by a downhole geophone array placed in an observation well near the treatment well. Once the microseismic activity is acquired, it can then be processed using geophysical techniques to produce a complete fracture map. Fiber-optic diagnostic tools are quickly gaining recognition as a viable method for acquiring and quantifying accurate near-wellbore results. Under the umbrella of fiber-optics-based information fall numerous methods of data acquisition and analysis, one of which is DTS. Through data transfer from a fiber-optic cable extending from the toe of the well to surface, the operator can view a temperature trace along the entire wellbore at any moment in time at 1-m spatial resolution. The most pertinent information acquired in terms of temperature analysis relates to monitoring injection and fluid production. DTS data acquisition typically begins before any fluid injection so a geothermal temperature profile for the entire well can be established. During fracturing, the fluid pumped downhole is typically cooler than bottomhole static temperatures, and DTS monitors the movement of this cool fluid down the casing by comparing the temperature readings during injection with the original geothermal temperature data. As the stimulation fluid exits the casing and directly contacts the fiber-optic cable, fluid movement outside of casing can be identified. This type of analysis has proven helpful in analyzing perforation effectiveness, plug or packer integrity, or cement integrity behind casing. To understand the true diagnostic potential of combining microseismic data with DTS results, the strengths and weaknesses of both technologies must be understood. Microseismic fracture mapping can provide an estimate of the far-field fracture geometry and azimuth. One challenge inherent to fracture mapping is that it is commonly not accurate on the 1-m scale, because a number of factors—including velocity and tool orientation errors—affect the location of microseismic events. DTS technology, in contrast, is very accurate on the 1-m scale and requires very little processing to output the final results. However, DTS cannot view activity that occurs away from the wellbore. Combining the near-wellbore accuracy of DTS with a geometric map of the created fracture network via microseismic data offers the possibility of obtaining very conclusive results in terms of stimulation effectiveness, completion efficiency and, most importantly, how these results can impact overall well production. ANALYSIS The Barnett shale, when cored or logged with full-bore formation microimager logs, typically demonstrates a significantly dense number of near-vertical natural fractures that are often oriented N45°E in the study area. In the Barnett and other naturally fractured shale plays in North America, cemented and non-cemented natural fracture orientations often vary, and additional complexities related to faults in the reservoir often exist. Frac fluid moving through the reservoir may not travel along a simple trend or well-defined principal stress direction. For well placement and development drilling, it is important to characterize and understand the natural and stimulated fracture geometries. Furthermore, attempting to correlate production trends and treatment results without the aid of diagnostic tools in this type of heterogeneous reservoir environment can be exceedingly difficult. For the study well in Parker County, Texas, each treatment interval consisted of 10-ft perforated sections at five shots per foot with treatment interval spacing of 287–313 ft. Height growth in this reservoir was minimal and very well-contained. The data indicated a large fracture network with multidirectional fracture growth. In Fig. 1, representing a point in time roughly halfway through the Stage 1 stimulation, there is noticeable variability in the temperature data uphole from the Stage 1 perforated interval (blue). The occurrence of these types of spikes in the temperature profile is quite common. The spikes are a function of the intermittent binding of the fiber-optic cable to the production casing, which is typically at every cross-coupling and at a point in the middle of each casing joint. At points where the cable is not directly bound to the casing, the temperature recorded will vary because there is no direct conductive heat transfer from the casing to the cable. This inconsistency does not interfere with the ability to assess fluid movement through the perforations, since at these points the fluid will directly contact the fiber-optic cable. The Stage 1 stimulation is represented in Fig. 1 by the blue diamond (for perforations) and blue points (for microseismic events). The DTS results indicate that fluid is exiting the wellbore at the desired perforation interval, as determined by the fact that the temperature at the Stage 1 perforation interval is slightly cooler than at any other point along the displayed wellbore section. This indicates direct fluid contact with the fiber-optic cable. The DTS results also show that no fluid is moving through the toe of the well, since the temperature downhole of the Stage 1 perforations is significantly warmer than the rest of the displayed wellbore. The microseismic map, as anticipated, reveals broadening complex fracture geometry, which is interpreted as fracture networking. For reservoirs in which complex fracture geometries are observed, stimulated reservoir volume (SRV) data can be quantified using the microseismic results, for the entire well or for each individual stage. This value is then correlated to production results and compared to other SRV values in the general area. For this case study, the stimulated reservoir area (SRA) is analyzed, as opposed to direct stage-by-stage comparison of SRV values. Typically, SRV values are compared on a well-by-well basis and then normalized to overall well production. In the well under study, simply referring to SRA values has merit since the vertical height growth of fractures stays very well-contained and consistent throughout the treated intervals, ranging from 200 ft to 250 ft. Figure 2 highlights one of the major contributions DTS provides with respect to fluid injection analysis. The two images define the near-wellbore fluid movement for Stages 2 and 3. It is apparent that fluid is moving through the annular area between the casing and the wellbore once it exits the Stage 2 perforated interval (red). This is evidenced by the smooth temperature profile indicating that the fluid is contacting the fiber-optic cable directly. As a result, roughly 150 ft of reservoir is exposed to the treatment fluid, as opposed to the 10 ft designed in the perforation strategy. This can indicate a cement isolation issue in this region. When DTS data from Stage 3 (light blue) is acquired, it is apparent that fluid is moving behind casing on either side of the perforated interval. In this case, roughly 500 ft of reservoir is exposed to the fluid. Comparing these two images, it is obvious that fluid entry points into the reservoir were overlapping with respect to Stages 2 and 3.
DTS analysis during the stimulation cannot determine how many points of fluid entry exist behind casing in scenarios such as this. It can only determine that at least two points of fluid entry exist, one at either end of the consistent temperature decline. The most accurate method for determining the number of entry points into the reservoir using DTS is to watch thermal recovery after the stimulation job is completed, Fig. 3. The points where fluid has entered the reservoir can be identified through this warm-back image. Whereas fluid stayed very well-contained as it entered the formation during Stage 1, as reflected by the very narrow temperature restoration signature, there is a very broad thermal cooling signature between the perforations of Stages 2 and 3, indicating that a significant portion of the fluid volumes for these two fractures entered the reservoir at roughly the same place. This conclusion is further supported by the fact that the injection temperature profiles for these two stages displayed significant overlap near the wellbore, and by the temperature decline observed right at the Stage 3 perforated interval. Now that DTS analysis has established that overlap exists between perforated intervals near the wellbore, microseismic results can be used to help define the effect this will have on the stimulation geometry. Figure 4 shows the microseismic results for both Stage 2 (red) and Stage 3 (light blue). As this well was completed in the Barnett shale with some structural dip to the southeast, it is expected that the fracture geometry will be fairly complex. This expectation is confirmed by the significant amount of SRA overlap between the two stages. Typically, this amount of overlap can be attributed to the reservoir characteristics or completion strategy. In this case, however, the SRA overlap also appears to be influenced by inadequate treatment isolation near the wellbore. Often in a mature field, desired values for SRV and subsequent SRV percentage overlap are determined prior to the fracture treatments based on previous work done in the field, resulting in completion strategies to optimize these values. In this case study, if minimal values for SRV overlap were targeted and these results displayed very high SRV overlap percentages, then perhaps stimulation or perforation strategies could be adjusted in the future. However, with the added benefit of the DTS information, it appears that perforation spacing, stimulation design and pre-existing natural fracture patterns were not the only influences to dictate hydraulic fracture overlap in the reservoir. Clearly, the overlap was also influenced by the lack of isolation in the near-wellbore environment. Stage 5, in this particular case, may provide additional evidence that the near-wellbore region directly impacts hydraulic fracture growth away from the wellbore. The DTS data for Stage 5 (green) in Fig. 5 shows that during the stimulation, the treatment fluid in the near-wellbore zone stays reasonably well-contained near the designed 10-ft perforated interval. Comparing the Stage 5 and Stage 4 (orange) microseismic geometries, it appears— especially on the north side of the lateral—that the overlap is limited, suggesting that Stage 5 will provide added value to production by contacting virgin rock outside of the geometry created during Stage 4. When reviewing the microseismic SRA overlap on the south side of the lateral, there appears to be more overlap near the microseismic geophone observation well. This is most likely attributable to a previous fracture network that exists as a result of a previous stimulation in the observation well. The fracture network on the south side of the treatment lateral could be heavily influenced by the previously established fracture network in the area. The outcome is that we see more network overlap between Stages 4 and 5 on the south side of the lateral. To help quantify this overlap, the SRA was calculated, as well as SRA percentage overlap. The SRA values for Stages 3, 4 and 5 are 1.8 million sq ft, 2.1 million sq ft, and 2.2 million sq ft, respectively. These values include all microseismic points mapped for each stage in the Barnett zone. SRA percentage overlap between Stages 3 and 4 is surprisingly similar at 51% (Fig. 6) compared with Stages 4 and 5 at 48%.
CONCLUSIONS The combination of DTS and microseismic analyses clearly suggests that fluid movement in the near-wellbore zone, and the resultant points of fluid entry into the reservoir, can significantly impact fracture network development. If only one of these tools were utilized, the evaluation of the completion effectiveness would be much less accurate. For instance, if only microseismic mapping were used on this well, then it would be more challenging to identify the main source of the geometry overlap. The operator might have concluded that, besides reservoir characteristics, the perforation strategy or stimulation technique was the sole or main contributor to total stimulation overlap. With the additional aid of DTS data, it can be seen that poor isolation near the wellbore may have also contributed to the geometry overlap. Having both diagnostic tools certainly helps to define the required corrective actions moving forward to complete future wells as designed. Unconventional reservoirs will continue to challenge operators as they look at ways to optimize reservoir productivity while reducing costs. Combining multiple diagnostic tools such as fiber-optic DTS and microseismic fracture mapping is one method to improve the overall understanding of these challenging plays. ACKNOWLEDGMENT The authors would like to thank the management at Pioneer Natural Resources, Halliburton and Pinnacle for the opportunity to publish this article.
|
|||||||||||||||||||||||||||||||||||
ROBERT HULL is a geophysicist for Pioneer Natural Resources, currently assigned to the Geosciences Technology group, and is a leader in 3D and microseismic acquisition. He has worked in the Barnett shale, West Africa and South Africa for Pioneer. He previously worked for Maxus Energy/YPF-Repsol. Mr. Hull earned a bachelor’s degree in geology from the University of Rochester and a master’s degree from the University of Texas at Dallas. / 
