This document discusses advances in using high-resolution seismic reflection methods for coal exploration in the United States. It provides examples of how seismic data acquisition and processing techniques have improved, allowing for better imaging of coal seams. The use of vibrator sources and nonlinear sweep parameters are shown to provide higher quality seismic data compared to traditional small explosive charges. Computer modeling and attributes analysis are also discussed as ways to better interpret seismic data and identify subtle geologic features important for mine planning. The integration of seismic data with borehole geophysical logs and computer modeling is demonstrated to improve geologic interpretations.

1. By LAWRENCE M. tiOC’HIO( ‘0
Consolidation Coal Cornpam
Libm-v. Pennsvhlnicl
T . . .
he utillzatlon of seismic reflection
methodsin US coal explorationincreased
duringthe lastdecade,asevidencedby the
number of publications and papers
presentedat technical meetings. The in-
creasedapplication of innovative techni-
quesadaptedfrom the petroleumindustry
has resulted in improvementsin seismic
data acquisition, processing, and inter-
pretation.The objectiveof this paperis to
presentsomeadvancesin the applicationof
high-resolutionseismicreflection methods
to US coal exploration since they were
adoptedin tbe mid-1970s.
US coal companiestraditionally have
minedreserveswith soundgeologiccondi-
tions and have avoided known problem
areas.Asthesebetter reservesaredepleted,
coal companieswill have to mine through
geologicallymore difficult areasto gainac-
cessto other good reserves. Also, more
coal companiesare utilizing the longwall
mining method because it allows large
blocksof coal to be mined very efficient-
ly. However, highly productive longwall
mining requires large reserveswhich are
fairly level to be free of major geologic
anomalies.The US Bureau of Mines es-
timatesthat an unscheduledinterruptionof
a longwall face advancecostscoal com-
paniesan averageof $250per minute.This
translatesinto a lossof about$120000 per
eight-hourshift. A few weeksof downtime
could cost a coal company millions of
dollars. If tbe downtime was causedby
significant geologic anomalies or distur-
bances,then the resultingloss might have
been preventedif an adequateseismicand
drilling explorationprogramhadbeencon-
ductedin advance.
Evolution. Seismic surveying for coal
explorationrequires some refinements in
methodologybecausethe targets fall be-
tweenthevery shallowobjectives,lessthan
15.2m (50 II), normallyencounteredin en-
gineering applicationsand deeper targets,
40 45 50 55 sp 65 70 75 80 ehotpoint
(4
OS_-
z
8>
B.+
0.3 -
40 45 50 55 60 65 70 75 81) shotpoint
(b)
0.2-
F
;>
‘S
0.3-
Figure 1. Resultsof linear (a) and nonlinear (b) sweeps.
- 0.2
( coal
- 0.3
GEOPHYSICS: THE LEADING EDGE OF EXPLORATION DEC‘EMHtK 1991

2. Conocoprototypevibrator beingusedas a high-frequencysourcein coalexploration.
_-_ _..
0 3w
91:4m
Figure 2. Seismicsectionshowinga robust and continuouscoal-seamreflection. This indicatesuniform seamthicknesswith no
detectablegeologicdisturbance.
GEOPHYSICS: THE LEADING EDGE OF EXPLORATION DECEMBER 1991 25

3. depthblbelow
surface
198m
sandyshale
sandstone
sandyshale
ehale
fire&y
ehale-limestoneunite
shale
coal
8ilt8tone/claystone
sandyshale
213m
228m
243m
Figure 3. Lithologic cross-sectionof the two boreholesin Figure 2.
161 107’ 26’7’
t 9 t
I 1x5 145 155 165 175 165 195 205 215 k25 235
greaterthan914 m (3000 ft), of petroleum
exploration. Undergroundmining of bitu-
minouscoalin the majorUS basinsoccurs
at depths ranging from 45-762 m (150-
2500 ft) beneaththe surface. Fortunately,
seismographsthatweredesignedfor engin-
eeringapplicationsandpetroleumexplora-
tion can be adaptedfor coal exploration.
Data acquisitiontechniqueshave evolved
from the early ’70s when 12-channel,lo-
bit, fixed-gain engineering seismographs
were commonlyusedin field surveys.As
microcomputertechnologyimproved dur-
ing the l!%Os, 24-channel, 15bit, instan-
taneousfloatingpoint engineeringseismo-
graphsbecamethestandard.Currentengin-
eeringseismographsandrecordingsystems
are capableof yet higher digital sample
rates, broader dynamic range, and more
recording channels. This has resulted in
better signal-to-noiseratios due to higher
fold, whicheffectsanimprovementin both
vertical and lateral resolution.Additional-
ly , some conventionalrecording systems
used for petroleum exploration have the
capabilityto recordseismicdataat sample
rates of 1 ms and 112ms with up to 48
channelsper record.
Somecurrentengineeringseismographs
and conventional recorders have built-in
software packages that can be used to
evaluatethe data while still in the field.
Digital filtering, correlation, spectrum
analysis, normal moveout corrections,
commonoffsetgathers,andbrutestackare
commonutilitiesin thesesystems.Thisuse-
ful technologyis beingimplementedin the
field to enhanceanalysisof the recorded
seismic data to ensurequality control in
dataacquisition
F*leldtesting and data acquisition. The
typical seismic sourcesused for Coalex-
ploration in Europe, Asia, Canada, and
Australia are small explosive chargesor
mechanicalweight-dropdevices.However,
2s6
?
&
245 255 265 23 265 hotpoint
0.0
a?lcoal
0.2
figure 4. Seismicsectionshowing shallow coal-seamreflection interrupted by multiple faults.

4. therearealimitednumberof publishedca$c
historiesin which a vibrator was effective
for shallow high-resolution work. Thlz
paper focuseson resultsobtainedusing .I
high-frequency vibroseis source for coal
exploration. The Conoco prototype unit
shown at the beginningof the article can
generatesweepfrequenciesashigh as400
Hz with up to 33 000 lb peakforce. When
utilizing a vibroseis source,proper selec
tion of sweep parameters-like frequency
sweep range, sweeplength, linear and/or
nonlinear sweep-is an important part ~1’
the acquisitionprocess.
Example 1. Two seismic sectionsare
presentedin Figure 1 to demonstratefield
testingof different vibroseissweepparam-
eters to optimize the recording of high
resolutionseismicdata.The testswerecon-
ducted in a deep Appalachian coal field
where the target coal seamlies about610
m (2000 ft) beneaththe surface and the
averageseamthicknessis 1.8 m (6 ft) Syn-
thetic seismogramsgenerated from sonic
anddensitylogsindicatea datumcorrected
arrival time of approximatly246 msfor the
coal seamreflection. Figure la is a seismic
sectionobtainedusinga linear sweep.The
reflection associatedwith the target coal
seamis rather weak and hard to interpret.
A secondtest over the same interval was
madewith a nonlinearsweep. The result-
ing sectionis shownin Figure lb. Similar
dataprocessingsequenceswere employed
to assembleboth sections.The seismicdata
obtainedwith the nonlinearsweepshow ;I
robustandcontinuousreflectionassociated
with the coal-seamhorizon; this suggests
uniform near-seam conditions. The non-
linear sweep method concentratedmore
energyin the desiredfrequencybandwidth
and improved the signal-to-noise ratio
Moreover, coherentreflectionsfrom great-
er depthswere alsorecorded,as indicated
bycomparisonof reflectorcontinuityat345
ms.The depthof thisdeeperreflectorises-
timatedto be 823 m (2700 fi).
The experiencesgained in conducting
thesefield testswill assistin proper selec-
tion andusageof sweepparametersto op-
timize the recordingof good quality seis-
mic data. The following two examples
demonstratecase studieswhere collection
of seismicdata with very good signal-to-
noise ratio provided assistancein evaluat-
ing subsurfacegeologicconditions.
Example2. Figure 2 is a seismicsec-
tion from a high-resolution survey con-
ductedto evaluateseamcontinuitybetween
boreholes. The reflection associatedwith
thetargetcoal seamis indicated,aswell as
the locationsof two boreholes(A and B)
which are separatedby 439 m (1440 ft).
Boreholedata revealed seamthicknessto
be about3 m (10 ft) andseamdepthabout
228.6 m (750 ft). Lithologic information
from the two boreholes is presented in
Figure 3. Informationfrom thesetwo bore-.
holes alone could not guaranteethat the
seamwas continuousand uniformly thick
overthisinterval. Geologicdisturbancesor
anomaliesthatmightcreateadversemining
VSP data
Checkshot da!a
CDP seismicdata
Sonic & density logs
Geologic interpretation
Geloglc logs
Syntheticseismogram
Time-depth conversion
Well-logcorrelation
Seismicprocessing
Strabgraphicmodeling
Structural modeling
Complex attributes
Acoustic log studies
Mapping
Figure 5. Schematicdiagram of interactive workstation environment in coal mining.
Data types are on the left and typical applicationson the right.
60
Reflection
Density {g/cc) Depth (R) coefficent Timn fn~
-0 .25 0.25
...-- -I
0 100
-0.150
Figure 6. Synthetic seismogramgenerated from available geophysicallogs.
conditionscould be present between the
boreholes.The sectionshowsthecoal-seam
reflectionto be robustandcontinuousfrom
SP-20 to SP-100, indicatinguniform seam
thicknessacrossthe entire section. More-
over, reflections associatedwith the im-
mediateroof and floor rocks appearto be
constant,suggestingnearlyuniform lithol-
ogy with minimal lateral changes.Thus,
seismicdatacoupledwith the boreholedata
provided additional assuranceof satisfac-
tory mining conditions,
Example3. The highfrequencyvibrator
canalsobe an effective sourcefor shallow
targetsthat lie a few hundredfeet beneath
the surface.Figure 4 is a sectionof a sur-
vey conductedto aid an explorationdrill-
ing programby mappinga major fault sys-
temwhichcauseddramaticchangesin seam
elevation.Severalboreholesacrossthesur-
vey line revealedthattheaveragedepthsof
the mineablecoal seamlocated southand
northof the fault zoneare 48.4 m (160 ft)
and 91.4 m (300 II), respectively. The
averageseamthicknessis 3 m (10 ft). The
locations of several boreholes and the
depthsof thecoal seamare notedon topof
the section.The reflectionassociatedwith
thecoalseamis indicated.Two majorseam
disturbancesassociatedwith the fault sys-
tem were interpretednear SP-134 and SP-
16.5.A smallerfaultthatmay connectwith
the southernfault wasalsointerpretednear
SP-184. The robustand continuouscoal-
seamreflectionfrom SP-135 to SP-185in-
dicatesauniformly thickcoalseamof about
3 m (10 ft). Dramaticchangesin thearrival
time of the coal-seamreflection over this
interval suggestprobablesteepdips in the
seamnearinterpretedfaults. From SP-185
to SP-292,theseamreflectionis robustand
continuous,indicatinga uniformthickness.
A relativelysmallfaultwasinterpretednear
SP-214 where a noticeablechangein ar-
C;t.OPtiYSI~ 5: i HI I t AI)I’K; EDGE Ot- EXPL.OKAT‘ION DECEMBER 1991 27

5. 28 GEOPHYSICS: THE LEADING EDGE OF EXPLORATION DECEMBER 1991
rival time is evident(notesomedisruption that the seam dips to greaterdepthand
in reflectionsfrom thedeeperstrata).The remainsfairly level from SP-228 to SP- Data analysisand computermodeling.
databetweenSP-185and SP-227suggest 292. Major oil companies utilize computer
workstationsto significantlyimproveinter-
tz
pretationsof seismicdatathroughtheinter-
activeprocessof cross-correlatinggeolog-
ic and geophysicaldata sets. Computer-
generatedmodels are matchedwith the
recordeddata to strengthentheseinter-
pretations.Thissameusefultechnologycan
beusedtoenhancetheseismicprogramfor
coal-miningapplications.Improvementsin
microcomputertechnologyand software
developmentover the last 10 yearshave
made seismic interactive interpretation
workstationcapabilitiesbothaffordableand
accessible.With proper hardware and
softwareconfigurations,a microcomputer-
basedworkstationcanprocesshigh-resolu-
tion seismic data to conduct modeling
studiesforimprovedinterpretationandcor-
relationto common-depth-pointdata.Fig-
ure 5 is a schematicdiagramof themulti-
task capabilitiesof a workstationbeing
employedfor coalexploration.
Duringthelastfiveyears,thecollection
of soniclogsin explorationboreholeshas
beenregularlyconductedprior to seismic
surveys. Vertical seismic profiling and
checkshotdataalsoaregatheredin someof
theseboreholesfor subsequentmodeling
studies.A workstationisthenutilixedtoin-
tegratethe downholegeologicand geo-
physicalinformationin order to generate
syntheticseismogramsforcorrelationtothe
processedCDP seismicdata (Figure 6).
Key reflectorsare annotatedin the figure.
Recentconversionfrom analogto digital
recordingof well-log data has increased
productivityandimprovedthedefinitionof
rockformations.
Althoughcoalseamsareextremelythin
with respectto wavelength,theyoftenpro-
ducedistinctreflectionsbecauseof an ex-
ceptionallylarge acousticimpedancecon-
trastwith respectto roof and floor rocks.
In many cases,the seamthicknessis less
than the standardone-quarterwavelength
criterionneededto resolvethetopandbase
of a bed, but is closeto the tuningthick-
ness-one-eighthof a wavelength.There-
fore,it isimportanttousecomputermodel-
ing to studythe effectsof thin coal seams
onreflectionamplitudes(seeTting eficrs
and interjerencerejectionsfrom thin beak
and coal seams, GEOPHYSICSAugust
1991).
The interpretationof broadbandcoal
seismic data often requires advanced
methodsand techniques.The signatureof
the seismicwaveletprovidesa greatdeal
of valuable geologicinformation. Subtle
featuresandseamanomalies,noteasilyob-
servedin conventionalblackandwhiteXC-
t e~~qqlnnonn~n9%~n~~n~~~~~~~~~~
iii<
tions, can be enhancedthroughcolor at-
~si~i~i6i~~~idfiii~i~~~~~~~
r(_(r(rld_( ..~*~-Mfvnnnnnnn tributedisplays.Detectingandinterpreting
small faults (lessthan seamthickness)is
difficult, butthisis extremelyimportantin
coal-miningoperationsbecausea faultwith
Figure 7. Instautaueousfrequencyattribute displayof Figure 4. a vertical displacementof aboutone-half

6. the seamthicknessis enoughto stopthe the CDP seismicdatapresentedin Figure reflectiontunestofrequenciesrangingfrom
advanceof a longwallface.Figure7 isthe 4. Thisdisplayshowsthatthespectrumof 100-140Hz. Faultsproducefrequencyand
instantaneousfrequencyattributedisplayof thedataextendsto 200 Hz. The coal-seam amplitude variations due to interference
within the Fresnelzonethat straddlesthe
fault. Disturbancesin thetuningfrequency
of the coal-seamreflectionare evidentin
Figure 7 at SP-134, SP-165, SP-183, and
SP-214. The instantaneous reflection
strengthattributedisplay is presentedin
Figure 8 and showsa robust coal-seam
t
reflection.However,dimmingof reflection
amplitudesis observedat locationswhere
z
the reflectionfrequenciesare alsoaltered,
suggestingthelikelihoodof a geologicdis-
turbanceassociatedwith faulting.
cn
-t
Conclusion. Seismicreflectionprofiling
gainedincreasedutilizationinthe1980sbut
thistechniquewill achieveitsfull utility for
coal-mining applicationsonly if the in-
dustrycontinuesto utilize improvingtech-
nologydevelopedfor thepetroleumindus-
try. For example,in a milestoneeventfor
US coal exploration,ConsolidationCoal
Companyconductedits first 3-D seismic
surveyin 1989.The resultsprovidedmore
informationfor improvedcontrolin map-
pingtheseamstructurethandataavailable
from a grid of conventional2-D lines. In
the 199Os,it is likely that more3-D seis-
mic surveyswill be conductedto fully
evaluatecoal reservesfor improvedmine
planning.Collectionof suchdatasetsand
the requirementfor improvedaccuracyin
interpretationimpliesthatcomputerwork-
stationswill play a greaterrole in coalex-
ploration.IE
Acknowledgments:I am grateful to my
Jesuit mentors-Fathers Francis Heyden,
Sergio Su, Victor Badillo, Daniel McNa-
mara, and Miguel Bemad-for their unsel-
fish, inspiring, creative, and innovative
works.I alsothank Consolmanagementfor
permissionto publish thispaper.
LawrenceM. Goch-
iocoisa Conocore-
search geophysicist
assigned to direct
and enhance the
seismicprogram of
ConsolidationCoal
Company (Consol).
He receiveda BS in
physicsfromAteneo
de Manila University (1978) and spenttwo
years there teaching collegephysics and
working at the Manila Observatory mon-
31PFP,ZbTPT
?t~4~“~rrg:*nfznla, -_-*“““~‘r~Pe~p~~fff~S~~~
itoring solarflares and sunspots.G&hioco
7 c 5 n 7 $ ci 4 4 ; ; -’ - - - . . - _ _ ,’ $ & =: n’ & 2 n’ ; ;I, ,, ?=-*.-I
earnedan MS inphysicsffom Ohio Univer-
sity (1982), workedfor CitiesService,Geo-
source, and Explorer, thenjoined Consol
R&D in 1985 to adapt high-resolutionseis-
Figure 8. Instantaneousreflectionstrengthdisplayof Figure 4. mictechniquesto coal-miningapplications.
GEOPHYSICS: THE LEADING EDGE OF EXPLORATION DECEMBER 1991 29