CSUG/SPE 138145

Petrophysical Characterization of the Eagle Ford Shale in South Texas J. Mullen, SPE, Halliburton

Copyright 2010, Society of Petroleum Engineers This paper was prepared for presentation at the Canadian Unconventional Resources & International Petroleum Conference held in Calgary, Alberta, Canada, 19–21 October 2010. This paper was selected for presentation by a CSUG/SPE program committee following review of information contained in an abstract submitted by the author(s). Contents of the paper have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers, its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written consent of the Society of Petroleum Engineers is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.

Abstract The Eagle Ford shale play is an emerging shale play that extends from the Mexican border in south Texas all the way to the East Texas Basin. Developing the play into an economically viable venture encompasses numerous challenges to include:

• The shale production characteristics vary across the play. • The shale is producing dry gas in some areas and wet gas or oil in others. • Some regions are naturally fractured, while others are not. • The play must be hydraulically fractured to be economically productive; however, what completion techniques have been

successful in one well will not necessarily work in another, even in the same field. Thus, it is critical to consider the local-area reservoir characteristics when trying to complete each well.

This paper focuses primarily on understanding the reservoir by integrating various data-acquisition and reservoir-

characterization techniques (i.e., mudlogs, basic openhole logs, and advanced logs, such as Dipole sonic, geochemical, NMR (magnetic resonance-imaging log), and core analysis) to determine the shale’s petrophysical characteristics and to thus build a locally validated petrophysical model (shale log) that can be applied to future wells with reduced data-acquisition programs to grade the reservoir. The model is used to ascertain the surrounding lithology and clay typing in addition to the hydrocarbon resource potential of the well. Furthermore, this tool can be used to answer completion questions, such as where the organic-rich zones are located, where to perforate, what the geomechanical issues are, how “fracable” the rock is, and how plastic the rock is.

Introduction The Eagle Ford shale has long been known as a shale resource rock, but only recently has it been “discovered” as a viable shale play formation.. Technological advances in hydraulic fracturing and horizontal drilling have been one of the drivers, as well as the high gas prices prevalent in 2007 and early 2008. These factors lead to a boom in unconventional-reservoir development (particularly in the Haynesville play). The first few exploration wells in the Eagle Ford shale were drilled in late 2008 in LaSalle County, in the gas window of the play (Fig. 1). Early on, the completion designs emulated Barnett-style, water-frac stimulation treatments with varying degrees of success. We now realize that one of the major challenges when trying to design a completion in this shale play is to know what type of hydrocarbons will be produced. Fig. 1 is for illustrative purpose only and displays the general areas where the reservoir produces oil (green), high liquids (yellow), and predominately dry gas (red).

Historically, the Eagle Ford was thought to only be the source rock for the Austin Chalk, Buda, and other shallower formations; it is now seen as a reservoir in its own right. This means understanding the local reservoir is critical to successfully developing this unconventional reservoir, which needs to be effectively stimulated to make the play economically attractive. Geology The Eagle Ford shale lies above the Buda limestone and is uncomfortably overlain by the Austin Chalk. This late-Cretaceous shale formation covers a laterally extensive area from Maverick County in the west, all the way across the state to the eastern county of Burleson, and beyond (Fig. 1). Fig. 2 illustrates how the stratigraphic column through the Eagle Ford varies across the play.

2 CSUG/SPE 138145

Fig. 1—Lateral extent of Eagle Ford shale in south Texas, for illustrative purposes only.

Fig. 2—Stratigraphic column through south Texas (Condon and Dyman 2006). Condon et al. (2006) comprehensively described the geology, structural features, and depositional environment of the Eagle Ford, as well as the hydrocarbon-migration mechanism. Basically, reactivated faults provided pathways and barriers to updip migration. In some downdip areas, they acted as barriers leading to large hydrocarbon accumulations, whereas, in other areas, they acted as corridors allowing the oil to migrate updip. Additionally, the Cretaceous period was quiet, tectonically speaking, in this northern Gulf of Mexico area—the shelf-edge reefs (black line labeled “reef trend” in Fig. 1) were formed in the early Cretaceous period

CSUG/SPE 138145 3

where the continental shelf met the Gulf of Mexico basin and only diverged at the western edge. The southern branch was formed much earlier than the northern branch, which in turn formed landward above the position of the earlier reef. Following the building of the southern reef, lime mudstone and shale were deposited by marine transgression. In the late Cretaceous period, the shelf edge also affected the deposition of the Eagle Ford shale because shelf slope, basin turbidities, and deltas formed on and along these areas leading to depocenters (which are called this because certain areas had much higher volumes of sandy sediment compared to other sedimentary deposits in other portions of the play). Consequently, the Eagle Ford reservoir quality and thickness varies laterally and vertically, depending on its location in the play and whether it is updip or down dip of the reef. Fig. 3 illustrates how some basic structural characteristics of this shale vary significantly across the play; for example, its gross height ranges from 20 to 500-ft thick. Even its depth varies from 2,500 to 14,000 ft. In addition to the aforementioned variation in the predominate type of hydrocarbon produced, some parts of the play are overpressured, generally downdip of the Edwards reef trend.

Fig. 3—(a) Map of the Eagle Ford top; (b) Map of the Eagle Ford bottom.

4 CSUG/SPE 138145

Fig. 4—Lateral variation in basic log response across the Eagle Ford play.

Three wells are described in this petrophysical characterization— Well 1 is in the gas-condensate window, Well 2 is in the dry-gas window and is at a similar depth to Well 1, and Well 3 is in the oil window. The distance between Well 1 and Well 2 is about 55 miles; the distance between Well 2 and Well 3 is about 135 miles. Fig. 4 illustrates significant changes that are seen when just looking at two basic log responses across the play (gamma ray is the green line in Track 1 and the red line is the deep resistivity in Track 2). It can be clearly seen that, traveling westward into the gulf, the thickness of the Eagle Ford increases from ~220 to ~420 ft, and more lithofacies are differentiated (Well 3). Between Well 1 and Well 2, it can be seen that the lithofacies are persistent in both wells, as is the transitional zone between the Buda and Eagle Ford (which has thinned in Well 2, probably as a result of shallower seas). The height is also relatively stable, even though the distance between the two wells is about 55 miles. The Buda and Eagle Ford transition zone is totally absent from Well 3; instead, it can be seen that the upper Eagle Ford (the reservoir between the red line labeled “Eagle Ford” and the blue line labeled “Eagle Ford target”) has increased substantially in thickness by a factor of seven.

Looking at core data verifies these logs responses. It can be seen from the ternary diagrams (Fig. 5) that the mineral composition changes as one moves across the play— the more western well is more quartz rich, the other wells are more carbonate and clay rich. This is explained by the geology. In the late Cretaceous Period, the eastern part of the play subsided less than the western side; consequently, the upper Cretaceous rocks in the eastern part of the play contain more shale and carbonates and less sandstone in comparison with time-equivalent rocks in the western part of the play (Condon and Dyman 2006). The core data was used to validate in each specific area the petrophysical shale-log model and also the mineralogy interpretation derived from the geochemical tool.

CSUG/SPE 138145 5

Fig. 5—(a) Well 1 (gas-condensate window) X-ray diffraction (XRD); (b) Well 2 (gas window) XRD; (c) Well 3 (oil window); (d) all 3 wells XRD. Mineralogical Interpretation using Geochemical Tool The geochemical tool has a theory of operation based on neutron-induced, capture gamma ray spectroscopy to measure elemental yield. The geochemical tool uses a chemical source, Americium-Beryllium, to emit neutrons into the surrounding environment. The neutron energy decreases with time and interactions (e.g., scattering and absorption that in turn cause the formation, borehole and geochemical tool itself to create gamma rays whose energies are characteristic of the excited element). The emitted gamma ray spectra are then analyzed. The acquired neutron-induced, capture gamma ray spectra undergo spectral fitting to give the relative elemental yields of eight different elements (magnesium, aluminum, silicon, potassium, calcium, titanium, manganese, and iron). Then, an oxides-closure model is applied to give the elemental weight fractions. Finally, mineral volumes are interpreted from the elemental weight fractions, coupled with the bulk density, neutron porosity, and volumetric photoelectric absorption data by means of a probabilistic error-minimization approach. All the logging data is used concurrently to solve for mineralogy and porosities, and the solution can be refined further if resistivity data is added, which also enables the saturations to be determined. The inelastic gamma ray spectra also undergoes spectral fitting for the elemental yields of oxygen, carbon, silicon, calcium, and iron. Galford et al. (2009) details the tool theory and mathematical techniques used to derive the mineralogical interpretation from the measured gamma spectra.

Thus, the geochemical tool can be used to identify complex mineralogy by measuring the most commonly found elements in the earth’s crust and then estimating the mineral volumes. Core XRD data is used to validate the local geochemical mineralogy interpretation (discussed in a later section) so that, on future wells in that region, no coring will be needed unless significant variation in lithology is seen, thus leading to savings in developing costs and potentially reducing the time to drill and complete the well. Some of the most important features of the geochemical tool are its ability to predict the different types and volumes of clays

6 CSUG/SPE 138145

present as well as quantifying the volume of kerogen present without the need to take core. The geochemical tool was run on Well 2 and Well 3.

The Petrophysical Shale-Log Model Data from a variety of different sources, such as electrical logs, core data, and mud logs, on each vertical well or a direct offset well were analyzed using a well-defined systematic approach (Rickman et al. 2008; Kundert and Mullen 2009; Mullen et al. 2007) to develop an accurate petrophysical model to characterize the Eagle Ford shale in each specific area that could then be applied to other wells in the region to aid in developing a completion strategy based on the well’s petrophysical characteristics. The model used the electrical logs to first indentify the organic-rich shale using the spectral gamma ray, which was also used for clay typing if geochemical data were unavailable. The geochemical tool was used for the mineral identification, which is critical for completion design because the fluid effects on the formation must be considered. If geochemical data were unavailable, the other conventional logs were used (i.e., photoelectric absorption for pyrite volume, bulk density for limestone volume, neutron porosity for clay volume and porosity, and resistivity for water saturation) to determine the mineralogy. The sonic logs combined with multiple other petrophysical relationships (where both mineralogy and shale type were considered) were used to calculate rock properties (Poisson’s ratio (PR) and Young’s modulus (YM)), and this was then calibrated to both dipole sonic data and core data. Borehole image logs were used to identify fracture type and frequency, bedding planes, dip orientation, laminations, facies, faults, etc.—thus giving a detailed description of the wellbore circumference. The magnetic resonance imaging logging (NMR) tool was run in two of the wells to determine a lithology-free porosity and the volume of free fluids present. Mullen et al. (2005) described in detail the theory of NMR T1 and T2 measurements in unconventional reservoirs. The T1 is used for volumetric analysis because, in the T1 inversion space, all the bulk gas will appear much later (above 1000 ms) than the bulk volume irreducible (BVI) water (0 to 32 ms) and will thus appear on the far right of the spectrum. In contrast to this, the T2 bulk gas signal lies in a much earlier window (typically 20 to 80 ms for small pores), thus leading to too much BVI and no free fluid. Additionally, the shale-log model was used to calculate the total organic carbon (TOC), kerogen content, free-gas volume, and rock-brittleness factor (Rickman et al. 2008), thereby grading the shale reservoir into potential completion intervals (Mullen et al. 2010). The brittleness factor is both a function of PR and YM and is used to identify which rock is ductile and which is brittle. The more ductile rock is indicated by a green color spectrum, while the brittle rock is colored red in the brittleness factor track (Fig. 6). The ductile rock will tend to act as a frac barrier as well as making a good seal.

Fig. 6—Offset well calibrated to core data.

Well 1: Gas-Condensate Window In Well 1, the log-generated stress profile in conjunction with the rock properties was used for the completion design to simulate the sensitivity to frac rate and fluid type. The shale-brittleness factor, mineralogy, and the shale-classification analysis (Fig. 6) were used in the frac simulator to model the laminations and bedding planes (Stegent et al. 2010). This was incorporated in the frac model to simulate any bounding-layer properties that could affect fracture-height growth. The closure pressure, along with the embedment test data, was used to select the proppant type and mesh size. Finally, the mud log in the vertical pilot was reviewed to see the extent of any hydrocarbon shows.

Having evaluated the logs, the core data from an offset well (Fig. 5a, Fig. 7, and Table 1) was then used to calibrate the model for Well 1 (Fig. 8). This is a critical component of the analysis because, without measured core data, the log interpretation can be inaccurate. The entire petrophysical-analysis process loops around between logs and cores and back to the logs; thus, validating an accurate petrophysical model of this particular area. The mineralogy derived from the logs was calibrated to the core samples XRD analysis mineralogy, both of which were used to tailor the fracture fluids used for the completion. Core XRD analysis showed an

CSUG/SPE 138145 7

average total clay content of ~20% (ranging from 5 to 45%) with about half (10%) of that being mixed-layer clay; thus, clay-control material would be needed in the completion treatment and also in drilling the lateral in the well. The low YM (1-2e106 psi) indicated that the rock is relatively “soft” and prone to proppant embedment and possible fines migration at the high expected closure pressures; therefore, these characteristics would need to be factored into the completion design.

Fig. 7—Well 1 (a) average mineralogy and (b) clay-component percentages.

TABLE 1—SUMMARY OF RESERVOIR PROPERTIES FROM CORE-DATA ANALYSIS FROM OFFSET WELL TO WELL 1 — Minimum Maximum

TOC, % 2 6 Porosity, % 8 18

Water saturation, % 7 31 Permeability, nanoDarcies 1 800

Young’s Modulus, psi 1.00e106 2.00e+106 Poisson’s ratio 0.25 0.27

Core geochemistry identified the kerogen type and maturity to be in the mature-oil window, with some core sitting in the

condensate wet-gas window with fair hydrocarbon potential. Thus, the completion design would need to focus on an oil-producing shale rather than dry gas, unlike the other early Eagle Ford wells. Fluid-sensitivity tests and proppant-embedment tests were conducted on the core, and the findings were used in the stimulation design (Stegent et al. 2010). Thin sections of the core characterized the Eagle Ford shale visually as a planar, laminated shale with numerous bedding-plane fractures, planktonic foraminifera, and organic-rich matrix. Occasional bedding normal fractures and pressure-release fractures were also observed (Fig. 9). Whole core indicated that the overburden stress was normal. This information was used in the completion strategy because the potential existed for the swelling clays identified by XRD and the log to plug these fractures. No fractures were seen by the openhole image log from the offset well, but this is not uncommon for shale reservoirs. Image logs from the offset well showed the Eagle Ford shale to be a finely laminated shale with bedding planes, all with dip angles of less than 10°, indicative of a low-energy depositional environment, with the azimuth primarily SSE, and that neither fractures, nor faults, nor any folds were discernable in this part of the Eagle Ford play. The identification of natural fractures and natural-fracture orientation in the core, together with understanding the key reservoir characteristics (high possibility of the clays swelling, plugging the pore throats, hence impeding liquid hydrocarbon recovery in a fairly ductile rock) to tailor the stimulation design to the reservoir, are fundamental to developing commercially successful shale plays.

8 CSUG/SPE 138145

Fig. 8—Well 1 shale-log analysis.

CSUG/SPE 138145 9

Fig. 9—Normal and bedding-plane fractures from Eagle Ford cores.

Fig. 8 shows the shale-log analysis on Well 1, a vertical pilot hole. Raw log data are to the left of the depth track. It can clearly be seen in the mineralogy track that organic-rich shale is present throughout the entire 176-ft Eagle Ford interval and that the Eagle Ford shale predominantly consists of limestone with varying amounts of shale. A 22-ft thick, clay-rich facies was seen at the base of the Eagle Ford and is identified as organic ductile shale (shaded light green in the shale classification track) that has hydrocarbon-storage capacity. It has a much lower YM (2e106 psi) and higher stress than the Eagle Ford above it (YM of 3e106 psi), so it might act as a fracture barrier to prevent downward growth into the underlying Buda formation. The low YM across the entire interval also indicated that embedment and fines migration might be an issue at these predicted stresses. Another thinner and ductile clay-rich facies is seen at top of the interval. Directly above the Eagle Ford, there is a clay-rich facies that has high closure stress, which might help to prevent fracture growth out of the interval. The shale classification track shows the interval to be primarily a laminated shale (shaded yellow) and a calcareous brittle shale (shaded red), with the two aforementioned organic ductile shale facies at the upper and lower part of the interval. The shale brittleness track shows that the brittleness factor (black line) varies throughout the interval—it is much more ductile at the bottom in the clay-rich facies (shaded green) and is much more brittle (shaded red) through the main body of the interval. In the main body, the brittleness factor varies between ductile/brittle because of the laminated nature of the shale. This track also displays the water saturation—all the facies are saturated with hydrocarbons and only a little irreducible water. The rock-properties track show the stress (red line), YM (green line), and PR (blue line). The main body of the interval has a similar lower stress profile compared to the upper and lower clay-rich facies, and thus will be easier to fracture. This main body also has more kerogen and TOC than the upper and lower portions. Free gas is shown throughout the interval (shaded pink in the gas-content track). The shale-perm and shale-porosity tracks both confirm that the higher quality reservoir rock is located in the main body of the interval. The mud logs show that the highest rate of penetration (ROP) and the highest gas shows are in the lowest two thirds of the interval, indicative that natural fractures might be present.

10 CSUG/SPE 138145

Thus, any future lateral drilled from this wellbore should be targeted to land in the “sweet spot,” in the window between 12,860 to 12,880 ft, to maximize recovery from the interval. This would allow the upper ductile facies to act as a seal for the reservoir and as a potential upper-frac barrier. The lower ductile facies has the best chance of keeping the stimulation treatment from breaking into the Buda below. Stegent et al. (2010) describes in detail the successful hybrid completion treatment of the lateral borehole of this well that, sixty days after treating, had produced more than 50,000 bbl of oil and 150 MMscf gas (~80,000 BOE), and only ~10BWPD, far exceeding the production of an offset, slickwater completion.

Well 2: Dry-Gas Window Well 2 had a full suite of openhole-logging tools run and sidewall cores pulled because it was this operator’s first venture into the Eagle Ford play. The geochemical interpretation was solved for the common reservoir rock minerals: quartz, limestone, dolomite, and feldspar, and shale/claystone rock minerals: illite, smectite, kaolinite, and chlorite. Then, the geochemical interpretation was fine-tuned to the core XRD data (Fig. 5b). In Fig.10, the little circles represent the core data points for the different minerals, kerogen, effective porosity, apparent matrix density, and effective water saturation. It can be seen that there is good correlation between the core data and the log data, thus there is confidence in this mineralogy interpretation. The geochemical mineralogy interpretation shows the volumes of chlorite and kaolinite were fairly constant through the interval and that high illite volumes (averaging 10% and increasing up to 40% at the base of the interval) were seen; thus, some type of clay-control material (e.g., KCl) would be needed to control clay swelling, inhibition, and hole cleaning, and prevent washout or borehole erosion and instability when drilling the lateral or completing the vertical pilot well. The geochemical analysis shows the kerogen-rich material is located in the 235 ft from 85 to 320 ft. The upper Eagle Ford has less kerogen-rich material than the lower Eagle Ford (starting around 162 ft). It can be seen that the effective porosity averages about 8% in the lower Eagle Ford and increases slightly in the upper Eagle Ford. The geochemical total porosity averages about 8% in the lower Eagle Ford, whereas at the upper Eagle Ford, it averages 10%. Fig. 11 shows the detailed NMR NMR T1 and T2 fluid-distribution analysis for this well. There is about a 6% undercall on the NMR porosity and the lower Eagle Ford appears to be gas-charged. This interpretation is further confirmed by looking at the overlaying geochemical porosities (Fig. 12). The NMR porosities are significantly undercalling the porosity, indicative of the reservoir being gas-filled. This was confirmed by T1-T2 diffusion plots. The NMR T1 analysis indicates that ~235 ft of the reservoir contains hydrocarbon. Thus, using both the NMR and geochemical data porosity data together gives a good indication of what the porosity distributions are within the reservoir. Little movable water was identified in the reservoir by the NMR tool. NMR T1 data were used to identify free porosity, bound fluid (both claybound and BVI), and to give a qualitative indicator of permeability in this extremely tight unconventional reservoir. Core data showed the permeability to be in the order of nanoDarcies, which is beyond the sensitivity of the NMR tool, and is probably not reflective of the bulk-system permeability. Small-volume diagnostic fracture-injection tests are the most reliable method to derive the permeability in the system. This technique has been proven in various shale plays across the USA— most recently, the Haynesville.

Fig.12 shows the shale-log model that was calibrated to core data (represented by the colored triangles and circles) using the complete suite of tools. Dipole sonic logs were run and thus anisotropy could be calculated (colored orange in the anisotropy track). The clay-rich facies at the top and bottom of the interval show the largest degree of anisotropy. Dipole data was also used to calculate a brittleness factor in the fast and slow directions (blue and yellow lines respectively on the brittleness track), as well as from the compressional data (black line). Differences between the fast and slow shear brittleness factors give an indication of potential fracture complexity that can develop when stimulating the well. In Well 2, it can be seen that the upper and lower Eagle Ford intervals appear to be fairly brittle and are separated by a ductile facie that is about 20-ft thick. This must be considered when trying to complete this vertical well because the stresses in this ductile facie are greater than the overlying and underlying rock and it might then act as a frac barrier. Additionally, the YM in this ductile facie is ~2e106 psi; whereas, the rock beneath it has a higher YM of 3-3.5e106 psi, thus embedment and fines issues must be addressed in the completion design. The kerogen-rich material (shaded black in the DeltaLogR track) was determined from the geochemical data, as was the TOC— both of these are much higher in the lower Eagle Ford than the upper Eagle Ford, indicating that that interval has better quality of reservoir rock. This is also seen by the free gas that is present in the whole interval, but is much higher in the lower Eagle Ford. The clay type and plasticity tracks are used together to identify potential wellbore-stability issues that need to be addressed before drilling a lateral. Ideally, the lateral should be targeted in the best quality rock as determined by the geochemical data, NMR, rock properties, brittleness factor and mudlog. In this case, around 210 to 230 (Fig. 11), and then verified using the plasticity and clay type that the well will not be landed in a zone that has high plasticity (shaded green in the plasticity track) and high volumes of mixed-layer clays (shaded red in the clay type track i.e., swelling clays).

Image logs in this well showed the Eagle Ford to be deposited in a low-energy environment, with dips less than 10° in a SSE direction. No faults or natural fractures were identified. It was seen that the texture and composition changed as the different facies were identified. The mudlog showed that the largest gas shows were seen in the lower Eagle Ford interval, indicating the presence of a natural fracture pathway. Core geochemistry in Well 2 identified the kerogen type and maturity to be in the dry-gas post-mature window, and the kerogen quality to be organically rich. Core rock-properties experiments found the rock to have a higher

CSUG/SPE 138145 11

YM (3-4e106 psi) than that in Well 1, so the shale-log model was calibrated to these core data. Thus, the completion design for this operator’s first Eagle Ford vertical well would need to focus on a dry-gas producing reservoir similar to the other early Eagle Ford wells, but with sustainable production. The key to the latter is tailoring the treatment to this reservoir to avoid issues of swelling clays, plugging pore throats, and embedment in “soft,” ductile rock (Stegent et al. 2010).

Fig. 10—Well 2 geochemical data calibrated to core data.

12 CSUG/SPE 138145

Fig. 11—Well 2 NMR T1 and T2 data.

CSUG/SPE 138145 13

Fig. 12—Well 2 shale-log analysis calibrated to core (black dots, triangles, andcrosses).

Well 3: Oil Window The only core data available was XRD (Fig. 5c), which was used to calibrate the geochemical tool in this area. The amount of quartz has increased, with a corresponding decrease in the amount of clay when compared to Well 1. Average values in the Eagle

14 CSUG/SPE 138145

Ford are given in Fig. 13. It can be seen that, in Well 3, the Eagle Ford average clay composition is significantly different than that of Well 1 (Fig. 7b). The amount of kaolinite has increased from 7 to 32%. It contains ~70% swelling clays, which make up about 7% of the total rock composition on average; thus, clay control will still be needed in the fracture treatment to minimize the effect of swelling clays on the pore throats in the reservoir.

Fig. 13—Well 3 (a) average mineralogy and (b) clay-component percentages.

Fig. 14 shows the shale-log analysis on Well 3, which used the geochemical interpretation for mineralogy, the NMR T1 analysis for fluid distribution (Fig. 15), and the dipole sonic data to calculate anisotropy and rock properties. Detailed borehole image-log analysis on Well 3 is given in Figs. 16 and 17, which show that, in this part of the play, open fractures of random orientation are evident in both the upper and lower Eagle Ford, which is now ~420 ft. Borehole-breakout analysis determined the maximum stress direction to be NE-SW. The blue tadpoles in Fig. 16 represent the natural open fractures, and the black tadpoles represent the drilling-induced fractures. The identification of these natural fractures is important to consider in the stimulation treatment because they, combined with the dominant hydraulically induced fracture, can possibly lead to pressure-dependant leakoff during the fracture treatment. Furthermore, the natural fractures can create a complex fracture network that should be identified early in the treatment. The Eagle Ford is seen again to be a laminated shale with bedding planes that all had low dip angles (less than 10°), indicating a low-energy depositional environment. No major or abrupt changes in the bedding planes were observed and no folds or faults were identified. The relative colors of the borehole-image log can also be used qualitatively because higher organic content and lower total-clay volumes result in lighter colors on the image; thus, the lower Eagle Ford appears lighter and more organically rich than the upper Eagle Ford, based on the image log’s colors. Mud-log gas shows were present throughout the gross interval with the exceptions of the clay-rich facies at the bottom of the Eagle Ford and in the middle of the Eagle Ford. Both the geochemical interpretation and the borehole analysis show many more facies are present in the upper Eagle Ford section than in the first two wells, as well as some additional facies appearing in the lower Eagle Ford. The geochemical interpretation is validated by the XRD data. The predominant clay was illite throughout the interval; the lower kerogen-rich Eagle Ford was more quartz rich than the upper Eagle Ford, which was more lime rich. Hydrocarbons were estimated across the entire Eagle Ford interval by the geochemical analysis. NMR T1 analysis found the total porosity range to be from 5 to 14%, averaging about 10%, with more free-fluid porosity in the upper Eagle Ford than the lower Eagle Ford, which has more storage capacity. It should also be noted that this reservoir’s geochemical total porosity correlates closely to that from the NMR T1 analysis. In this well, the Eagle Ford does not appear to be gas charged. The more lime-rich upper Eagle Ford is clearly seen to be more brittle (red in brittleness track) than the lower Eagle Ford (green). The lower Eagle Ford has more hydrocarbon-storage capacity (shaded pink in the gas-content track) and displays a higher degree of plasticity (green in plasticity track) and more swelling clays than the upper interval. Interestingly, the clay-rich facies that separates the upper and lower Eagle Ford display the largest amount of anisotropy. Consequently, the optimal completion should be made where natural fractures are present, where the image log appears lighter, and where there is more storage capacity in the reservoir, whilst ensuring any plasticity/swelling-clay issues are addressed when drilling the lateral. Embedment is unlikely to cause issues for this low-YM rock because the expected stresses are not high.

CSUG/SPE 138145 15

Fig.14—Well 3 shale-log analysis.

16 CSUG/SPE 138145

Fig. 15—Well 3 NMR T1 and T2 data.

CSUG/SPE 138145 17

Fig. 16—Well 3 borehole-image analysis showing fracture frequency.

18 CSUG/SPE 138145

Fig. 17—Well 3 borehole-Image analysis. Conclusion The thickness of the Eagle Ford increases from ~220 to ~420 ft moving west from Well 1 to Well 3, with Well 3 having many additional new facies that are not present in Wells 1 and 2. Well 3 has a more quartz-rich lower Eagle Ford than the other two wells that are more carbonate and clay rich. The geochemical tool was used to differentiate these facies and represent a cost-effective way to characterize mineralogy and locate kerogen-rich rock without taking core. The Eagle Ford shale has core permeability in the order of nanoDarcies (beyond the range of NMR technology) thousands of times smaller than that of tight-gas reservoirs; thus, it must be hydraulically fractured to be productive. A more representative method to characterize the bulk system permeability is using small-volume diagnostic fracture injection tests (DFIT’s).

The Eagle Ford varies significantly in petrophysical characteristics, such as thickness, mineralogy, and hydrocarbon produced, etc. across the play; thus, understanding and characterizing the local area using logging and core data are key to developing the play for sustainable long-term production, and to reduce the learning curve in this new play. The shale-log petrophysical model integrates all available data from multiple sources, such as NMR, geochemical, dipole sonic, image log, mudlog, fracture frequency, and core analysis into one comprehensive plot for a well that can be used for decision making, tailoring the completion design, and further geologic and stratigraphic interpretation, either well-to-well or basin wide. The model provides information, such as:

• Location of organic-rich zones. • Location of brittle zones. • Location of permeable zones. • Frac-design parameters (i.e., volume, additives, proppant strength, and acid solubility). • Clay typing. • Organic content. • Volumetric assessment. • NMR porosity and free fluid and a qualitative indicator of permeability. • Plasticity.

CSUG/SPE 138145 19

Acknowledgements The author thanks Mani Azerli, Mke Mullen, South Texas Tech Team, and Scott Shannon for their help as well as management of Halliburton for permission to publish this work. The author also thanks those who permitted their well-data to be published. Nomenclature BVI = Bulk volume irreducible GR = Gamma ray KCl = Potassium chloride NMR = Magnetic resonance-imaging log NMR = Nuclear magnetic resonance PR = Poisons’ ratio ROP = Rate of penetration TOC = Total organic carbon XRD = X-Ray diffraction YM = Young’s modulus References Condon, S.M. and Dyman, T.S. 2006. 2003 Geologic Assessment of Undiscovered Conventional Oil and Gas Resources in the Upper Cretaceous

Navarro and Taylor Groups, Western Gulf Province, Texas: U.S. Geological Survey Digital Data Series DDS–69–H, Chapter 2 42 p. http://pubs.usgs.gov/dds/dds-069/dds-069-h/.

Galford, J., Quirein, J., Shannon, S., Truax, T., and Witkowsky. J. Field Test Results of a New Neutron-Induced Gamma-Ray Spectroscopy Geochemical Logging Tool. Paper SPE 123992 presented at the Annual Technical Conference and Exhibition, New Orleans, Louisiana, 4–7 October. doi: 10.2118/123992-MS.

Mullen, M., and Roundtree, R. 2007. A Composite Determination of Mechanical Rock Properties for Stimulation Design (What to Do When You Don’t Have a Sonic Log). Paper SPE 108139 presented at the Rocky Mountain Oil and Gas Technology Symposium, Denver, Colorado, 16–18 April. doi: 10.2118/108139-MS.

Kundert, D. and Mullen, M. P2009. Proper Evaluation of Shale Gas Reservoirs Leads to a More Effective Hydraulic-Fracture Stimulation. Paper SPE 123586 presented at the SPE Rocky Mountain Petroleum Technology Conference, Denver, Colorado, 14–16 April. DOI:10.2118/123586-MS.

Mullen, M., Gegg, J., Bonnie, R., Cherry, R., Riggert, G. 2005. Fluid Typing With T1 NMR: Incorporating T1 and T2 Measurements for Improved Interpretation in Tight Gas Sands and Unconventional Reservoirs. Paper presented at the SPWLA 46th Annual Logging Symposium, New Orleans, Louisiana, June 26–29.

Mullen, J., Lowry, J. and Chukwuemeka Nwabuoku, K. 2010. Case Study: Lessons Learned Developing the Eagle Ford Shale. Paper SPE 138446 to be presented at the Tight Gas Completions Conference, San Antonio, Texas, 2–3 November.

Rickman, R., Mullen, M., Petre, E., Grieser, B., and Kundert, D. 2008. A Practical Use of Shale Petrophysics for Stimulation Design Optimization: All Shale Plays Are Not Clones of the Barnett Shale. Paper SPE 115258 presented at the Annual Technical Conference and Exhibition, Denver, Colorado, 21–24 September. doi: 10.2118/115258-MS.

Stegent, N., Leotaud, L., and Prospere, W. 2010. Cement Technology Improves Fracture Initiation and Leads to Successful Treatments in the Eagle Ford Shale. Paper SPE 137441 to be presented at the Tight Gas Completions Conference, San Antonio, Texas, 2–3 November.

Stegent, N.,Wagner, A., Mullen,J, and Borstmayer, B. 2010. Engineering a Successful Fracture-Stimulation Treatment in the Eagle Ford Shale. Paper SPE 136183 to be presented at the Tight Gas Completions Conference, San Antonio, Texas, 2–3 November.