Hydraulic fracturing is key to the economic success of many oil and gas fields around the world and has improved production in low permeability reservoirs for more than 50 years. Successful stimulations are engineered to place the proper type and volume of slurry based on estimating the dimensions of the optimal fracture to be created in a specific wellbore. Several commonly used technologies are available which determine important fracture parameters such as fracture dimensions, fracture orientation, fracture conductivity and proppant placement effectiveness. Fracture models are today's most widely used tool to predict the optimal frac geometry based on conditional inputs such as closure stress, pore pressure, permeability, fluid saturation and numerous other mechanical and petrophysical properties of the reservoir. In many cases, these parameters are based on assumptions rather than hard data, and incorrect assumptions then lead to sub-optimal stimulation results. Direct near-wellbore diagnostics such as radioactive tracers and temperature logging are often used to gather information about fracture height and proppant placement effectiveness, while direct far-field diagnostics such as tiltmapping and microseismic fracture mapping are used to determine hydraulic fracture dimensions and orientation. Direct fracture diagnostics alone only tell the story of what happened after the fact on a given well, but they can also be used to build a calibrated fracture model which accurately predicts fracture growth in a reservoir. Depending upon the critical information needed for specific fracture stimulation, one or more diagnostic tools may be applied. These diagnostic tools will be discussed and compared in order to provide a reference of widely used diagnostic tools with strengths and limitations discussed along with examples of each in use in fracture optimization. Fracture Complexity For many years, fractures were assumed to be bi-wing, single planar features (mostly for easier numerical modeling) that would stay primarily within the pay zone and grow very long (Why would they want to grow anyplace else?). More than 10 years of direct fracture diagnostics and over 6000 mapped fractures have proven those assumptions to be mostly incorrect. Fractures in the real world are very complex. Numerous cases have been documented in the literature with direct fracture diagnostics, mine-backs and core-throughs where fractures are seen in multiple parallel planes, in multiple directions, and in "T-shaped" fashion with both horizontal and vertical components. Existing literature is rife with cases of incomplete coverage of pay zones where fractures may miss entire perforated intervals, only partially cover some intervals, grow primarily out of zone in others, deviate significantly from the wellbore causing connection or link-up problems, and grow into unwanted water or gas intervals nearby. Fracture mapping can be used in real-time to evaluate whether the entire pay is being sufficiently stimulated, whether the design calls for enough or too many stages, whether the optimal fracture length has been achieved and whether adjustments need to be made to the existing treatment design. Diagnostics Groups Fracture Diagnostics can be broken into 2 main groups (see Figure 1): Indirect and Direct Diagnostics. Indirect techniques include fracture modeling, well testing and production data analysis, while Direct measurements are further subdivided into Near-Wellbore (such as radioactive tracers, temperature, and production logs) and Far-Field (tiltmeter and microseismic mapping).

In developing a new field or reservoir, many parameters are important in understanding the success or possible areas for improvement in hydraulic fracturing. Estimating fracture geometry is essential to effectively calibrate a reservoir model to production results. Radioactive (RA) tracers have been used in hydraulic fracturing treatments to infer fracture dimensions. Three stable isotopes (i.e. Scandium, Iridium and Antimony) were used in various parts of the treatment to understand the progression of hydraulic fracture growth. Advanced sonic anisotropy logging tools, using a broader range of frequency acquisition, were used to enable shear measurement in cased hole environments over a wide range of interbedded coal, shale and sandstone sequences both before and after the hydraulic fracture treatment. Amplitude and anisotropy changes after a hydraulic fracture have been measured using sonic anisotropy logging and used to infer fracture height. Finally, the sonic anisotropy can be evaluated above and below the perforated interval and investigate hydraulic fracture height growth away from the wellbore, potentially visualising a greater distance than available with RA tracers. We will show how sonic anisotropy and radioactive tracer logging methods can be used to better understand the fracture geometry and aid further design work. The paper will present data from two (2) wells in the Walloon Coal Measures of the Surat Basin where both RA tracers and sonic anisotropy logs were used to infer fracture dimensions. Both wells used a combination of treated water stages, containing low concentrations of proppant, followed by borate-crosslinked gelled water stages with higher concentrations of proppant. This project contained a large amount of other hydraulic fracturing diagnostics including treatment pressure history- matching, microseismic monitoring and surface tiltmeters. In this paper we will note how those diagnostics compared with the results presented herein, but their results are discussed in greater detail elsewhere (Johnson et al. 2010a). Generally, the results indicate good agreement between these two fracture diagnostic methods and the authors will illustrate the complimentary nature of these diagnostics in gaining a fuller understanding of fracture height, especially in environments of complex fracture development.

Summary This paper compares flowback efficiencies using polymer concentration and frac fluid tracer methods. Results are presented for the flowback efficiency of each frac fluid segment using non-radioactive chemical frac tracers injected in a well, along with the results for the total flowback efficiencies using polymer concentration and frac fluid tracer analysis methods. Two wells were fraced and traced with various chemical frac tracers. Upon commencing flowback, samples of produced aqueous fluid solution were collected according to a pre-designed sampling schedule that lasted for 72 hours. Samples were analyzed for tracer, polymer, calcium, potassium, sodium, and chloride concentrations. With the use of the mass balance technique, the total flowback volume and flowback efficiency for each fluid segment were calculated by use of the tracer method. In addition, total flowback and flowback efficiency were calculated by use of both polymer concentration and tracer methods. To better evaluate and compare the results of polymer concentration and frac fluid tracer analyses, dynamic fluid leakoff tests were conducted in a laboratory environment by use of both low and high permeability core samples. Detailed laboratory and field results are presented along with a comparison of flowback results from both polymer concentration and frac fluid tracer methods. Introduction Chemical frac tracers (CFT) are from the family of halogenated organic acids and were originally developed in an effort to bolster the level of understanding regarding the dynamics of hydraulic fracture placement, subsequent fluid flowback and proppant bed cleanup. Borrowing from many years of experience with interwell tracing in which non-radioactive chemical tracers have been successfully used to evaluate interwell communication, several groups of these chemical compounds were identified that could potentially be placed in each segment of the frac fluid so as to more directly measure the flowback efficiency of each fluid segment. Armed with this flowback profile data together with the treatment pressure history of the frac treatment, it was believed that much could potentially be learned both about the dynamics of segmented fluid placement as well as segmented fluid flowback and cleanup. Given the established formation/fracture damage potential for conventional proppant transport fluids, those fluid segments not adequately recovered following the treatment could, in principle, detrimentally affect the flow capacity of the propped fracture. Chemical frac tracers were designed to be placed in chemically-differentiated and/or proppant-differentiated fluid segments of the fracturing fluid so as to assess the cleanup of the fracture as a function of segment fluid chemistry and/or fracture geometry. In so doing, it was believed that the sufficiency or insufficiency of addition rates for key frac fluid additives such as polymers, breakers and gel stabilizers could be assessed. It was also believed that the relative cleanup of individual frac treatment segments in a multiple stage completion procedure could be monitored. It was further hoped that inferences could be made from these data regarding lateral placement effectiveness of proppants and vertical communication between zones. Furthermore, the tracer analysis results could be used to assess the amount of each injected segment recovered and hence to calculate flowback efficiency. To fully investigate frac fluid compatibility of these chemical tracers, a series of rheology tests were designed and conducted with the Fann Model 50. Two generic frac fluids were selected to evaluate the effects of these chemical tracers on the viscosity of these frac fluids. These fluids are zirconate-crosslinked 35 lb/Mgal CMHPG (carboxymethyl hydroxypropyl guar) and borate-crosslinked 40 lb/Mgal guar gel. The first two tests were designed to establish a baseline for the viscosity of these two fluids without the addition of any chemical tracers. The viscosities of borate-crosslinked guar and zirconate-crosslinked CMHPG, the two generic fracturing fluids, at 250ºF and after 60 minutes at various shear rates, did not change with the addition of chemical frac tracers at concentrations of up to 100 ppm for zirconate-crosslinked CMHPG and more than 10 ppm for borate-crosslinked guar. The percent change in the pH of borate-crosslinked guar and zirconate-crosslinked CMHPG for the before and after rheology tests at 250ºF and 60 minutes with the addition of tracer is well within the percent change of fluid pH without the addition of tracer under similar testing conditions (Sullivan et al. 2004). Background Fluid flowback can be either of a fracture-tip or near-wellbore type. If flowback is of the near-wellbore type, it indicates extensive near-wellbore leakoff owing to a highly permeable zone around the wellbore. This causes much of the pad fluid segment to leakoff near the wellbore and, therefore, the pad fluid is first to be recovered. In a low permeability formation, pad flows to the fracture tip owing to low permeability and/or damaged permeability around the wellbore resulting in minimal leak-off near the wellbore. Once the well is subjected to flowback under this condition, what is injected first flows back last, if fluids are formulated properly. If some segments of gelled frac fluid are not broken effectively before the well is subjected to flowback, the early injected fluids could potentially finger through the late injected unbroken fluids and flowback first. In general, the flowback order of each segment depends on a number of factors, such as fluid type and polymer concentration, crosslinker concentration, breaker loading, pumping schedule, closure pressure, and flowback schedule, to name a few. Therefore, comprehensive diagnoses of flowback can only be accomplished with complete injection and flowback information. The detrimental effects of reduced fracture conductivity as a result of poor flowback are well documented in the literature. Most references have focused on the effects of using improper flowback procedures on well performance (Veatch and Crowell 1982; Hickey et al. 1981; Robinson et al. 1988; Barree and Kukherjee 1995). The associated effects are proppant movement into the wellbore, proppant crushing at or near the wellbore, and fracture plugging yielding reduced conductivity and productivity. Although fluid flowback is an important part of the fracture treatment, it has been overshadowed by proppant flowback concerns in recent years. The conventional method of quantifying fracture cleanup has been to report load water recovered during flowback. This value, however, is greatly influenced by the volume of formation water produced. It also, at best, provides information on the total recovery rather than the flowback of each frac fluid segment.

The majority of discoveries in the Cooper basin have been structural traps on regional highs, with significantly less information available from the deep tight sand and shale plays. Recently, exploration and appraisal of the large, prospective gas resources in the deep troughs began. While these plays have been known for some time (Hillis et al. 2001), data has only recently been presented detailing the unconventional targets in the Nappamerri trough (Pitkin et al. 2012). This paper focuses on unconventional reservoirs within the southern Cooper basin, specifically, a southern extension of the Nappamerri trough and the Mettika embayment, part of the Tenappera trough. Senex Energy began to review the potential of the unconventional reservoirs in its Cooper basin acreage in 2010. This paper focuses on three wells, Skipton 1, Kingston Rule 1, and Hornet 1, which were hydraulically fractured in early 2013. In the well planning stage of the project, it was identified that sufficient data needed to be collected to characterize both the formation and hydraulic fracturing behavior. The requirement for fracturing diagnostics was critical because of the complex nature of hydraulic fracturing within the Cooper basin. To manage these unknowns, the data that were collected included core testing, diagnostic injection tests on most intervals, surface tiltmeters to measure fracture azimuth and orientation, radioactive (RA) tracers to infer fracture height, and chemical tracers to estimate the flowback contribution from each stage. The results showed high treating pressures and near-wellbore pressure loss (NWBPL) in many stages of the fracture treatments. Generally, the lower the fracture gradient, the easier the job was to place and the better the reservoir quality. The RA tracers showed good containment in the one well in which they were used, with heights of 20 to 60 ft. The main barriers to height growth appeared to be changes in lithology. The surface tiltmeter results showed an azimuth that was different than expected from borehole breakout data in the basin. This has been seen in other studies in the Cooper basin. While the exact reason for this requires additional work, possible explanations are discussed within the paper.

Activity in the Barnett shale of north Texas has continued to surge over the past several years, and with this surge in activity has come a steady evolution of completion strategies. Most operators agree that the best Barnett shale wells are those that have the most extensive fracture network development. One of the biggest challenges facing operators has been determining which completion strategy will create the largest fracture network along with identifying the most cost-effective diagnostic methods for evaluating those strategies and ultimately optimizing the completions. This paper will describe an integrated completion diagnostic methodology for assessing and potentially optimizing Barnett shale completion strategies. The methodology described employs radioactive tracing and logging together with chemical tracing and conventional completion metrics to evaluate the effectiveness of Barnett completions. This methodology will be utilized to assess the effectiveness of four pairs of simultaneously-fractured parallel horizontal laterals located in three different areas stretching from the western edge of the Barnett "Core Area" to the western edge of the play's expansion acreage. Key completion parameters were identified that can be used to optimize future Barnett completions in this and potentially other areas within the basin. Introduction The Barnett shale is a successful development target as an unconventional reservoir in north Texas' Fort Worth Basin. The majority of the production from the Barnett is located in the Railroad Commission Newark, East field. Total gas production from all north Texas Barnett fields is in excess of 2.5 bcf/d from approximately 6800 wells, with cumulative production of 2.9 tcf as of April 2007. The Barnett is a Mississippian-aged marine deposit. Within the basin, the Barnett ranges in thickness between far less than a hundred feet in the far south and west (as it is affected by the Bend Arch and Llano highs), to over a thousand feet of thickness next to the northeastern edge (near the Munster Arch). The Barnett is typically seen as a black organic-rich shale. It is unique in its relatively low clay content and high quartz content. This mineralogical anomaly is thought to contribute to the good fracture conductivity that can be created during the hydraulic fracturing completion of Barnett shale wells. Unconformably underlying the Barnett shale is either the Ordovician-aged Ellenberger (limestone and dolomite) or a thin Simpson and Viola section. The distinction in the underlying formation is controlled by the position of the erosional limit of the Ellenberger limestone and dolomite or Simpson and Viola sections. While both instances represent a long period of sub-aerial exposure, the cave creation and subsequent karst topography development of the Ellenberger surface is of present importance and significance as it relates to the development of the Barnett shale reservoir. While most of the basin was relatively quiet through Barnett deposition, local karsts represent relatively active features that did see significant structural "development" during Barnett deposition. It is thought that this karst topography developed slowly as overburden weight collapsed the cave systems. This collapse was accommodated by the filling of Barnett lows, resulting in "ring-faulted" collapse structures with locally thickened Barnett sections. Areas of the basin that have preserved sections of the Simpson and Viola were largely "protected" from the Ellenberger's karsting effects. These areas are found in the far north and east portions of the basin, toward the Muenster Arch and Ouachita Thrust front. The Viola, where present, provides the additional mechanical benefit of providing a frac barrier during hydraulic fracture stimulation of the overlying Barnett rock.