Study re Shale Formations WESTERN
ATLAS
Atlas Wireline Services
September 30, 1994
Western Atlas
Mr. Sanford Dvorin International
V~rDS RESOLIRCES 12001 North Central
Expressway, Suite 870
5025 Arapaho Rd., Suite 350 Dallas, Texas 75243
Dallas, TX 75248 (214) 960-9494
Fax (214) 960-1915
Mr. Dvorin,
As per our conversation, the following is a brief log analysis of the Magnolia Petroleum
Corporation, Trigg Estate #1 well in Dallas County, Texas. The analysis will be over the
Barnett Shale zone from approximately 8,606' through 9,050'.
Barnett Shale Zone:
8606'-8666' 25-40 Ohms resistivity indicates dense brittle shales compared to normal
shales of less than 10 ohms; microlog shows laminations and positive
separation (micro normal > micro inverse). These laminations may
represent a reservoir rock that, with vertical fractures, will create
permeability throughout this zone and thus produce if hydrocarbon is
present.
8666'-9050' Over 50 ohms resistivity is indicative of very dense and possibly brittle
shale; the microlog continues to indicate laminated sequences throughout
this interval.
As the enclosed article states, as long as these shales have vertical fractures and conditions were
right for making hydrocarbon, they should be productive. These logs give good indication of
the factors necessary for a producing interval - high resistivities for hydrocarbon and laminated
interval for reservoir rock. If vertical fractures are present interconnecting the various lithology
and shale laminations (the hydrocarbon source) this well should produce as should offsets in this
same vicinity.
Mr. Dvorin, if I may be of further assistance, please let me know.
Thanks,
Bruce R. Noblett
Senior Log Analyst
BRN:hmm
Attachment
SPWLA Thirtieth Annual Logging Symposium. Sun¢ li-14, 1
NEW RESISIIVIlW AND FRACTURE DETECTION CONCEPTS
FOR THE EVALUATION OF SHALE FORMATIONS
Bruce R. Noblett and L. Pat Howard
Arias Wireline Services
Western Atlas International, Inc.
Dallas, Texas
For decades, shales such as the Eagle Ford, Bakken, and Barnett have been explored for commemial pro-
duction. Although some wells were prolific, others were disappointing. Conventional log analysis would
indicate possible zones of interest in the shales but gave no concrete evidence of producibility. Because these
black carboniferous shales are considered good soume rocks, if indication of movable hydrocarbons and
vertical fracturing can be detected, the shales may be productive. With the onset of Circumferential
Acoustilog~ fracture identification/orientation and Dielectric resistivity measurements, log analysis of
these black shales can now be facilitated.
This paper will show the use of S~o measurements and movable oil plots with data from the Dielectric
resistivity tool and fracture e~aluation from the Circumferential Acoustilog. Dielectric measurements provide
an indication of.hydrocarbon saturation and movability. Shales typically have bedding plane or "horizontal"
permeability. This permeabili .ty is usually too Iow for commercial production even il'the shales show hydro--
carbon saturations. Using the Circumferential Acoustilog, vertical fracturing through the shale beds can be
detected. This increase in permeability., due to fracturing, provides the necessary. "pipeline" to produce exist-
ing hydrocarbons. A new computer interpretation of Circumferential Acoustilog data will be shown that more
efficiently evaluates fractures, fracture frequency., and fracture trends. Examples of logs run in the Barnett
shale of nor-da central Texas ',,,'ill be presented in combination with a core analysis and geochemical analysis
of the shale. Evaluation of productive and non-productive sections of the Barnett shale will be summarized.
INTRODUCTION
Carbonaceous shales have alv, ays been a difficult log analysis problem. While shales have a l'figh porosity,
the semipermeable nature of the shale causes the effective porosi~ to be very low or nonexistent. Log analysis
has therefore treated shale and the shale-associated ~,ater-filled pore space as solid rock matrix. In shales
with a high hydrocarbon content, the possibility exists for production if an associated fracture network
occurs within the shale body. If the shale interval is analyzed in the traditional manner, the standard porosity
and resistivity tools prcMde little help in evalu.ating the formation. This has been evidenced ~ the lack
of data to effectively e',aluate shales such as the Bakken, Eagle Ford, and Barnett. It has generally been
accepted that the more evidence of vertical fracturing in the Bakken shale of the ~511iston basin area. the
more productive the overlying formations. Likewise, the less fracture indication, the lower the production
capabilities. In the Bakken shales, detection of fractures and hydrocarbons has been very difficult when using
standard log analysis. These problems have also occurred in the Eagle Ford shale of central Texas and the
Barnett shale of north central Texas. Some wells in the Austin Chalk trend have had completions attempted
in the underlying Eagle Ford shale. Many geologists consider the Eagle Ford to be the source rock for the
Austin Chalk. Abundant mudlog flourescence and gas shows do occur in this shale, thus, these completion
procedures were warranted. Still. standard log analysis could not support the evaluations. Until the use of
new resistivity and fracture detection concepts in the north Te~s area. the Barnett shale formation also was
difficult to evaluate for production. To perform proper evaluation, special techniques were required.
For production to occur in these shales, at least two criteria must exist: I) a fracture network must be present
and 2) movable hydrocarbons must be present. High uranium deposits olden occur within the fractured shales,
resulting in unusually high gamma ray readings. To detect these fractures and any additional fractures not
containing high concentrations or' uranium salts, circumferential acoustic measurements should be used.
)
SPWLA Thirtieth Annual Logging Symposium, June 11-14. 1989
Furthermore, it has generally been accepted that shales have bedding-plane permeability due to the clay
structure of the rock. This permeability alone is rarely enough to produce hydrocarbons in vast amounts due
to the overburden pressures. If the shale is sufficiently brittle and becomes vertically fractured during
geological tectonic events, this increase in "fracture" permeability may help the hydrocarbon production.
Hydrocarbons present in the shales need to be defined as movable or residual. Evidence of flushing of detected
hydrocarbons around the wellbore suggests movable hydrocarbons and is seen in the dielectric resistivity
measurements. Thus. the actual open fractures in the shales may be observed using the circumferential
acoustic tool and fluid invasion may be determined using the dielectric tool. This paper will briefly describe
the tools and the techniques used to perform this evaluation.
CFRCIo~D-ERENTL4~L ACOUSTILOG LOGGING
The Circumferential Acoustilog incorporates the use of generating and detecting circumferentially-
propagated acoustic boundary ~.aves. Development and interpretation of the Circumferential Acoustilog and
acquired data has been extensively discussed (Setser, 1981; Guy et al., 1986; Noblett et al., 1987; Fertl, 1988).
As shown in Fig. I, the logging tool is a four-arm pad device that contains two transmitters opposite each
other and two receivers opposite each other on a horizontal plane. Each pad is at 90 degrees with respect
to the adjacent pad. The logging instrument is designed to enhance the recording of Rayleigh and guided-fluid
boundary waves by eliminating any direct-fluid wave components. The Circumferential Acousti]og has
already been proven successful in detecting vertical or near vertical fracture systems and also has the ability
to delineate a fracture trend (Noblert et al.. 1987).
A voluminous amount of data is generated by the Circumferential Acoustilog due to its sampling rate
(8 samples per foot, thus a ISA-inch vertical bed resolution) and the acquisition of four waveforms per sample,
representing each 90 degree quadr-,mt of the borehole. A variable density plot and orientation data is presented
at the well site. The acoustic v, aveforms are digitally recorded on tape for additional computer processing.
If open vertical fracturing is encountered, the Rayleigh wave will ~'pically not propa~te across the fracture
zone. The guided-fluid wave. however, will propagate and only minimally anenuate. Therefore, as described
previously, attenuations of the Rayleigh reave in two of the four quadrants (opposite or adjacent) is an
indication of open vertical fracturing. Figure 2 shows the Circumferential Acoustilog well-site data. W~.vetrain
data from each 90 degree quadrant of the wellbore with a correlation gamma ray, X-Y calipers, and borehoIe
orientation data including the azimuth direction of the reference pad is presented. Evidence of fracturing can
be seen at XX166-XX184 (Quadrant A and C attenuation), X.X200-XX214 (Quadrant B and D attenuation),
and XX220-XX237 (Quadrant A and C attenuation).
The new computerized interpretation is shown in Fig. 3.
Tractr I ~ Contains a gamma ray for correlation
Track 2 -- Shows the borehole rugosity as recorded by the X-Y calipers on the logging instrument.
These data are represented by a picture of the wellbore. In addition, zones showing Rayleigh v, ave
attenuation on the raw data are plotted next to the wellbore. This indicates not only the zones that are
fractured but also the vertical fracture extent and relative fracture frequents' in the fracture system. The
more fractured an interval, the higher are the number of quadrants attenuated (up to four). Two quadrant
attenuation, showing vertical fracturing, exists in the intervals XX166-XXi84, XX200-XX214, and
XX220-XX237. Single quadrant attenuation, which may be due to rugs, porosity changes, or
fracturing, is shown from XX162-XX166. XX190-XX200. and XX218-XX220. Therefore, it can be
interpreted that the vertical fracture extent is from XXI62 to XX237, with the greatest fracturing in the
two quadrant attenuation intervals. Superimposed in Track 2 is the directional trend of the fracture
system as detected by the orientation data in the logging tool. In this example, the fracture trend is 64-71
degrees or ENE-WSW.
Track 3 -- Shows a histogram every 50 ft of the average vertical fracture trend orientation throughout
that 50-ft interval.
SPWLA Thirtieth Annum Log_~ing Symposium. June 11-14. 1~.
DI]ELECTRIC LOGG EN'G
Dielectric resistivity, measurements have been recorded in various formations since the inception of the tools
in the late seventies. Tv, o instrumentation systems are available: a mandrel-type device with a 47 MHz
transmitter and two receivers, and a pad-type device with two 200 MHz transmitters and t'v,~o receivers. The
instrument specifications are shown in Fig. 4. Various applications for this logging system include the location
and quantification of hydrocarbons (virtually independent of salinity), thin bed analysis, sand count
determination, and evidence of mo,table vem~s residual hydrocarbon. Additional applications include salinity
estimation, cementation factor esth'nation using the "W' factor, and irreducible water saturations and shallow
resistivity deaenn/nation in oil-base mud systems. These and other interpretive concepts have been discussed
previously (Arias Wu'el/ne Services, 1.988).
The 200 MHz dielectric resistivity measurement is most useful in evaluating shales. Due to its vertical
bed resolution of less than 3 inches, the shales which often contain thin laminations are easily detected.
The high frequency dielectric tool is capable of measuring the dielectric property of the rock matrix at a
shallow depth of investigation as well as measuring the R,~ resistivity at this shallow depth. By comparing
the S,,, value from the dielectric permitivity measurement with the S,~ value of the dielectric resistivity
measurement, an estimation of the resistivity value of the water fluid (R0 at a shallow depth can be made.
When contrasts between R~ and R,,, occur, thi~ P~ value indicates the points at which fluid invasion occurs.
To calculate S,,, within the shale using normal An:hie relationships, it is necessary to assign a porosity
value to the shale interval. This calculation is completed by calculating the apparent porosity from the
compensated neutron and compensated density logs and using this porosity value for the S,~ calculation.
This results in a visual presentation for moved hydrocarbons while reco~m'fi.zing that the effective matrix
porosity within the shale is really zero. Differences between apparent water-filled porosiL'y from the deep
induction response and the apparent v. ater-filled porosity from the dielectric response indicate areas where
probable flushing has occurred. The dielectric tool calculation of water resistivity values (RO will indicate
kigh resistivity where cormate v, ater has been replaced by mud filtrate v, ater.
An example of the dielectric analysis is shown in Fig. 5. Althou~ the presentation looks simple and is easy,
to read, it provides information not ava.ilable without the dielectric instrument.
Track I -- This u-ack includes the gm ray and textural parameter, W. The W calculation does not
apply to this shale evaluation but is shown in Appendix A. The higda ga my readin~ occur as a
result of uranium salt deposits with. in the shale wi'rich often indicate fluid flow.
Traclc2 -- This track shows the v, ater resistivity calculation from the dielectric log. In this ex,ample, R~,
is 0.04 ohm-meters and ~ is approximately 0.4 ohm-meters. Where invasion is low, the water resis-
tivity calculation approximates P,.,,. As invasion increases, the v,~ter resistivity calculation approaches
R~, depending upon the mL'dng of R~, and R.~. In Fig. 5, X.548-X.572 and X583-X600 indicate v,~ter
resisfivifies approaching IL.. and thus Iow flushing, while the zones from X520-X5~. 7.573-2(580 and
X63¢X642 indicate higher v, ater resistMfies and therefore greater R~- invasion.
Track ..3 -- This track shows the apparent v, ater saturation based on the apparent porosity, in the zone;
S,~ is from the dielectric data and S~, from the dual induction R~. These are apparent S,~ values rather
than true S~, since the effective matrix porosity is zero.
Traclc 4 -- These are the apparent porosity values. The zones where fluid invasion has occurred are
indicated by the stippled shading.
Track5 -- This is the lithology track. Since the interval is 100% shale, only a shale lithology is shown.
DISCUSSION
During the Mississippian Age, a black, carboniferous shale v,~s deposited over a wide area in north central
Texas. This shale, the Barnett. has long been considered a major source rock to many of the conglomerates
and limestones deposited during this same geologic age. Because the Barnett shale is such an outstanding source
for hydrocarbons, many operators are scrutinizing the data recorded through this interval. Mud logs typically
describe the Barnett as a black-gray shale containing carbonate streaks, fissures and, coal inclusions. Traces
) 3
SPWLA Thirtieth Annual Logging Symposium, June 11-14. 1989
of' mc[stone and free pyrite also are detected. A rock e~aluation pyrolysis taken in the Barnen shale showed
the following:
Total organic carbon 6.96%
Free hydrc~::arbons 2.53 mWg
Pyrolyzable kemgen 42.40 m~g
Hydrogen index 609
Maximum temperature 438°C
By definition, the values for the total organic carbon and pymlyzable kerogen indicate a very goed hydrocarbon
potential zone. The maximum temperature and hydrogen index numbers also support a mature hydrocarbon-
bearing forrrmtion. Gas shows tend to increase when drilling through the shale but some of these shows are
due to shale "bremkout" gas while others rrmy prove to be productive intervals. It is cases such as this where
the use of the Dielectric Log and Circumferential Acoustllog is most advamageous. The following three case
studies were all performed in the north central Texas region of the Barnett shale.
Case Study I:
Well/fl is an excellent example ora Barnett shale zone showing evidence of both vertical fracturing and movable
hydrocarbons (Figs. 6 and 7). Abnormally high gamma ray readings from X355-X383 and X412-X440 are the
first indications of the fractured shale intervals. These zones correspond well with vertical fracture zones
detected by the Circumferential Acoustflog. Vertical fracturing extends from X352-X4~_. X412-X420, and
X426--X480. The identification of vertical fracturing through~t the shale satisfies the first cdt,efta for a possible
productive shale formation. Using the data from the Dielectric Log, zones indicating that flushing has occurred
can be identified. These intervals are X352-X362, X368-X374, X381-X384, X390-X396, X406-X418,
X423-X433, and X436-X442. Note that the R~ curve reads higher than the formation waters. R~. The shale
below X442 shows some fracturing on the Circumferential Acoustilog but no flushing, and thus, may be closed
fractures and not as productive. Th. is evidence of flushing corresponds to the movable hydroc~"bon zones, seen
as the stippled areas in the porosity section of Track 4. In this example, the fracture indic, atom Gamma ray,
Circumferenthl Acoustilog) and rnoxable hydrocarbon indicator (flushing_, tLO all indicate the same zones of
interest for production. Therefore, 'q~ll ~ should be analyzed as a probable product/vt Barnett shale inter,'al.
This well was recently perforated in the previously mentioned zones of interest and tested econom/c production.
It is interesting to note that if the apparent density/neutron porosity and kigh resistivity within the interval from
X350-X488 are used, water saturation calculations of less than 50% v, vuld occur and indicate productivity.
Th/s would imply a much larger interval in which to attempt a completion versus that of the newly introduced
analysis. Lcd'er than expected production would occur due to the lack of flushing and greater residual
hydroca.d~ns apparent in most of the additional shale inter~al. A final note on Well ~ would be that the average
fracture trend, as seen Mth the Circumferential Acoustilog, in the zone of interest gould be a~roximately
103-126 degrees or ESE-W.~P,V. TI'ds information allows the operator to choose the best offie,ning location in
order to stay in the same fracture trend for optimum production in the zone of interest.
Case Study 2:
As can be seen in Figs. 8 and 9, the Barnett shale section of'C,~Ll fi2 is much thinner than the previous example.
High gamma my readirg~s are evident from X424-X440, X447-X450, X464--X470, and X480-X487. This
indication, in combination with the vertical fracture detected by the Circumferential Acoustil~, suggests
sufficient fracturing for a prcriuctive interval. Even the flushing, as indicated by the kigher R: curve on the
dielectric analysis from X426-X432, X436-X440, X472-X476, and X480.X486, suggests evidence of fracturing.
There is little evidence, however, of actual movable h~rocarbons in th/s well. Very lirde residual hydrocar~n
is shown at X442-X4-48 but there is a lack of fracturing through this interval. Furthermore, the v, ater saturations
are higher and thus suggest that whatever movable fluid is in the zone may be water as opposed to h>ctrocarbons.
It was recommended that this Barnett shale zone, even though it is h.ighly fractured (trend 10-55° or NNE-
SSW), should not be completed due to the lack of detectable movable hydrocarbons.
Case Study 3:
When vertical fracturing cannot be detected throughout the shale zone of interest, hydrocarbon production will
probably not be commercial. The Circumferential Acoustilog, shc~'n in Fig. 10, shov,~ no evidence of any
vertical fractures. Gamma my readings are high in various sections and flushing, shown with the Dielectric Log
SPWLA Thirtieth Annual LoEging Symposium, June l l-l.t, i
(Fig. II), is probably being due to bedding plane permeability. This conclusion is suggested due to the lack
of vertical fracturing evidence. There is considerable residual hydrocarbon in this Bar'net~ shale as seen in
Track 4 of the Dielectric Log. Movable hydrtx:arbon is also apparent but will probably not be productive due
to the lack of vertical fracturing and thus a lack of reservoir extent and permeability. This well was analyzed
as hydrocarbon bearing but non--commercial due to lack of vertical permeability.
CONCLUSION
As has been shown, there are at least two criteria needed for commercially productive shale intervals. First,
vertical fracturing is required in order to increa.se effective permeability. Secondly. there must be evidence of
movable hydrocarbons. New fracture detection analysis techniques using the Circumferential Acoustilog and
resistivity concepts using the Dielectric Log have made these shale analyses routine.
Atlas W'tre Services, 1.988, Dielectric Logging Service, p~blication number 9578.
Fertl, W.H., 1988, Circumferential Acoustic Logs Detect Natural Fractures and Determine Their Orientation:
Log Analysis Line, November.
Guy, J.O., Fertl, W.H.. and Oliver, D.W., 1986, The Use of Circumferentially Propogated Acoustic Waves in
Well Logging: Paper 5PE tS~.~ presented at the SPE of A.hME 61st Fall Meeting. Oct. 5- 8, New Orleans. LA.
Noblett, B.R., Feril, W.H., and Guy, J.O., 1987, Recent Ad~ances in Fracture E~-aluation: Paper SPE 16226
presented at the SPE 1987 Production Operations Symposium, March 8-10. Oklahoma City., OK.
Noblett, B.R., and ~rtl, W.H., 1987, Circumferential Acoustilog Finds Vertical Fractures: V~rfd Oil,
November 1987, pp. 37-38.
Setser, G.G.,1981 Fracture Detection by, Circunfferenfial Acoustic Erie _rgy: Paper SPE 102_04 presented at the
56th Annual SPE Fall Meeting. Oct. 5-7. San Antonio, TX.
ABOUT THE AUTHORS
Bruce R. Noblert L. Pat Ho~atcf
Brace R. Noblett received his B.S. d~ree in geology )'rom Allegheny College in 1977. He joined Atlas W'mfiine
Services in the same year and later did extensive ~rk with fracture identification to~ng tools as a log analyst.
Currently a senior log analyst in Dallas, he has published paper~ and given talks on new fracture analysis
techniques to various SPE groups and the CWLS. Brace is a member of SPWLA. AAPG. and the Dallas
Geological Society.
L. Pat Ho,~ard attended Rice Universi~ and West Texas State University and received a B. S. degree in
chemistry. He joined Atlas x3,qreline Se~'ices in 1974 as a logging engineer. He has since held the positions
of area engineer and area technical manager in west Texas and Oklahoma. Pat is currently a senior log= analyst
in Dallas and is a member of SPWI_,~ and SPE.
SPWLA Thirtieth Annual Logging Symposium, June 11-I-I, 1989
APPEN])LX A
Dielectric Analysis Derivation of "W" Factor
Using Dual Induction Focused Log Using Dielectric and Minilog®
S,~ = S.,o =
· R, ' Rxo
Calculations from:
Dielectric = S~o, Rz(R.~f)
Minilog = R~o
Kev: Depth of investigation of Minilog and
200 MHz Dielectric approximately the same
Unknown: m,n values. Solve for m = n or W
v'a.lue
(Rz/Rxo)I;w
Sxo - or
~,~ 1
Sxo = ov,-~- x Rz
Rxo
Solving for textural para,meter, W:
In (Rz/R~o)
W=
In ($~o x ~)
Place W value into S~ equation to correct for
%~rying m,n values
I
S,~ = '~' x R~
R,
-6-
SP\VL,-~ Thirtieth Annual Logging Symposium, June 11-14, '
Length 18 t~15.5 rn
O.D. 4 in./102 rnrn
Max. temp. 350°Fi175°C
Max. press. 20.000 psi/137.9 MPa
Recommended 25 to 30 fprnlT.6 to 9.1 rnprn
logging speed
Figure 1
Configuration and specifications of Circumferential Acoustilog
C24 (INCH) I TEN (LBS)
C13 (INCH}
GR (APl) DEV
SPWLA Thirtieth Annual Logging Symposium, June l l-l-t. 1989
47 MHz 2C0 MHz
Inst~mer~ Type Mandret Pad
Diameter 4.125 in. (105 mm) 45 irt (114 mm)
Lerx:jm 12.0 ft (3.66 tn) 98 ft (2_~9 tn)
V'~ght 175 113 (79.4 h:j) 250 lb (1'~34
Numl::er d '~-~qs~n~ers 1
Numl:ler o' ReCav~ws 2 2
Max. Oper~ng Tern~gera~ure 3503F (1770C) 3503F (17'/~')
Max. O~era~ng Pressure 20.CC0 ~ (137~C0 kPa) 2(10C0 I:~ (!37,800 kPa}
Minimum Hc~e Size 8,0 in. (152 mm) 60 in, (~52 mtn)
Reccrnmended b:x~ging $Ceecl 60 It/rain (183 rr/m~n) 30 ffJmn (9.1 tn/tnin)
high-resob~x~ tncde
~ Rescluto'~ 7.87 in. {02 tn) 3.0 in. (Off/'5 tn)
T~-~n~er to Recaver 1 3150 in, (0.8 tn) 100 in. (025 m)
Transmr~er to Rece,,~' 2 3~L37 ~n (I.0 tn) 130 ~n. (033 tn}
Figure 4
Dielectric logging tool specifications
F i'~A~3 Uf~£ II1DIC/.T IOInl I F RE {~U E NI~Y TEXTURAL
:::: :::: ~;~' ~ :__ ~ .....
~ ~ ': ~__..
:;:: ~ ~ ij~ ~.,...~ ............
:~ ~ , ~ ...................
~:::: :::(, j ~ :: ~
....... ~ ...... 9-'.
Il
. :, ~
--~--' /~..,I,...',.~. ::::~- .... ~ ...............
Figure 6 Figure 5
Case Sludy I Circumlerenlial Acouslilog Dielocldc Epilog®
[FR~URE IND~TI~IFflE~JEN~ DE.ti......... '---~
GAMMA RAYI CALIPER I ~LIPE~ 2 1~102 (OIMM)
;;1 - N
Figure 10 Figure ~
Case Sludy 3 Circumlorenlial ~ouslilog Case Sludy 2 Dielectric Epilog
SPWLA Thirdeth Annual Log~.in.~ Symposium. June 11-14. 19Sg
.-.-.
Case S~uc¥ 3 Oie~ec~r,c Emlog