REPAIR, EVALUATION, MAINTENANCE, ANDREHABILITATION RESEARCH PROGRAM

TECHNICAL REPORT REMR-GT-10

HIGH-RESOLUTION SEISMIC RE-FLECTIONco INVESTIGATIONS AT BEAVER DAM, ARKANSAS

by

~UIU Thomas L. Dobecki, Tanya L. NV ueller

Department of Geophysics< Colorado School of Mines

Golden, 'Colorado 80401

and

Monroe B. Savage

Instrumentation Services Division

DEPARTMENT OF TI-Z ARMYWaterways Experiment Station, IO orps of Engineers

P0 Box 631, Vicksburg, Mississippi 39181-0631

- - D'~I~ELi GTE

* o~AUG 41989

pla t d All ~ __I rwj) --

July 1989Final Rep )rt

Approved For F'-jblc Release; 1),sxrdbition Unlimited

P-repared for DEPARTMENT OF THE ARMYUS Army Corps of EngineersWashington, DC 20314-1000

Monitored by Geotechnical Le oratoryUS Army Engineer Waterways Ext e-riment StationP0 Box 631, Vicksburg, Missis;-ir. i 391F 1-0631

8

i

The following two letters used as part of the number designating technical

reports of research published under the Repair, Evaluation, Maintenance, and

Rehabilitation (REMR) Research Program identify the problem area under which

the report was prepared:

;roblem Area Problem Area

C7 Concrete and Steel Structures EM Electrical and Mechanical

G" Geotechnical E1 Environmental Impacts

H Hydraulics OM Operations Management

Cf Coastal

Destroy this report when no longer needed. Do not

return it to the originator.

The findings in this report are not to be construed as an

offizial Department of the Army position unless sodesignated by other authorized documents.

The contents of this report are not to be used for

advertising, publication, or promotional purposes.Citation of trade names does not constitute an

official endorsement or approval of the use of such

commercial products.

COVER PHOTOS:

TOP - Seismic Reflection Survey, Beaver Dam, Arkanisas.

I.. 1M iE -. Ccncept of Seismic Reflection.BO17 )M - Seismic Reflection Record, Beaver Dam, Arkalsas.

UnclassifiedSECURITY CLASSIFICATION OF THiS PAGE

Form Approved

REPORT DOCUMENTATION PAGE OMNo 07a4.0188

Ia. REPORT SECURITY CLASSJF CATJON "b RESTRICTIVE MARK:JGSUnclassified

2a SECURITY CLASSIFICATION AUTHORIT', 3 OiSTRi8UTION/ AVALABILITY OF REPORT2b DECLASSIFICATIONiDOWNGRADING SCHEOULE Approved for public release; distribution

unlimited.

4 PERFORMING ORGANIZATION REPORT NUMBER(S) S MON.TORiNG ORGANIZATION REPORT NUIMBER(S)

Technical Report REMR-GT-106a. NAME OF PERFORMING ORGANIZATiON 6b nFFCE SYMBOL 7a NAME OF MONITORNG ORGAN-ZAT!ONSee reverse 'if applhcabte) USAE1WES

Geotechnical Laboratory6c. ADDRESS (City, State, and ZIPCode) 7b 0, DRESS (City, Stmte and ZIP Code)

See reverse PO Box 631Vicksburg, MS 39181-0631

B& NAME OF FUNDING iSPONSORING So O;CCE SYMBOL 9 PROCL.,REMENT iNSTRUMENT DEN7IF.CATJON NUMBERORGANIZATIONI (if applicable)

US Army Corps of Engineer

Sc ADDRESS (Oty, State, and ZIP Code) 10 SO, RCE OF FUNDiNG NUMBERSPROGRAM I P=OECT 'TASK 'ORK UNITELEME1JT NO NO NO ACCESSO iC

W_h __to_, DC 234-_O__ WUI 3231511I TITLE (Inciude Security Classificati on)

High-Resolution Seismic Reflection investigations at Beaver Dam, Arkansas

12 PERSONAL AUTHOR(S)Dobecki, Thomas L., Mueller, Tanya L., Savage, Monroe B.

13a TYPE OF REPORT 13b TIME COVERED 14 DATE OF 3EPOPT (Year, Month, Day) I1S PAGE COUNTFinal report I ROMj.. j .TOJ 1uj A July 1989 9

5

16 SUPPLEMENTARY NOTATION

See reverse.

17 COSATI CODES 18 SI)BJECT TERMS (Continue on reverse if necessary and identfy by block number)FIELD GROUP SUB-GROUP Data processing Optimum offset Seismic reflection

Graben Seepage Seismic refraction

19. ABSTRACT (Continue on reverse if necessary and identify by block number)

Seismic refraction surveys require line lengths four to five times the desired depthof investigation. For many geotechnical applications, these are piysical ,nd gme..2...constraints which inhibit seismic refraction, such as narrow river valleys, changes indirection of the center line of a dam, or legal inaccessibility to land surrounding a site.The seismic reflection method offers an alternative to refraction for geotechnical siteinvestigations. Typical line lengths for reflection are approximately equal to the desireddepth of investigation. In the past, the required data processing, limitations of "engi-neering seismographs," and inapplicability of oil exploration seismic recording systems tovery shallow targets (< 30 m) precluded effective utilization of the seismic reflectionmethod for geotechnical applications. Rapid advances in microcomputer technology, develop-ment of digital engineering seismographs, development of high frequency seismic sources,

(Continued

20 DISTRIBUTION /AVAILABILITY OF ABSTRACT "21 ABSTRACT SECURITY CLASSIFICATION0 UNCLASSIFIED/UNLIMITED CM SAME AS RPT 0 DTIC USERS Unclassified

22a. NAME OF RESPONSIBLE INDIVIDUAL 22b TELEPHONE (Include Area Code) 22c OFFICE SYMBOL

00 Form 1473, JUN 86 Previous editions are obsolete. SECURITY CLASSIFICATION OF THIS PAGE

Unclassified

UnclassifiedSECURITY CLASSIFICATION OP TNIS PAGE

6A. NAME OF PERFORMING ORGANIZATION (Continued).

Department of GeophysicsColorado School of MinesUSAEWESInstrumentation Services Division

6C. ADDRESS (Continued).

Golden, CO 80401PO Box 631Vicksburg, MS 39181-0631

16. SUPPLEMENTARY NOTATION (Continued).

A report of the Geotechnical problem area of the Repair, Evaluation, Maintenance, andRehabilitation (REMR) Research Program. Available from National Technical InformationService, 5285 Port Royal Road, Springfield, Virginia 22161.

19. ABSTRACT (Continued).

and adaptation of oil exploration field procedures to account for inherent problems ofshallow depths of investigation now make shallow, high-resolution seismic reflection sur-veys a viable tool for geotechnical applications.

Field investigations of shallow, high-resolution seismic reflection methodology at anexisting structure site, Dike 1, Beaver Dam, AR are discussed. Dike I is built across agraben with vertical offset in excess of 70 m. The down dropped block consists of solu-tioned, cherty limestone. The top of rock is highly irregular, and solution cavities andsolution-widened joints exist in the rock. The seismic reflection results effectivelymapped the irregular top of rock (3 to 10 m dipth). detected previously unknown faultswithin the graben, and mapped sandstone and shale formatlon5 at the "base" of the graben(70 to 75 m depth). . -

Accesion For

NTIS CRA&I

DTIC TAB 0Unannounced 0

Justification

ByDistribution

:? Availability Codes

(\ ~dAv indior

Dist specill

Unci~qSECURITY CLASSIFICATION OF THIS PAGE

PREFACE

This work was performed from July 1985 to January 1986 by personnel of

the Department of Geophysics, Colorado School of Mines (CSM), Golden, CO, and

Instrumentation Services Division (ISD), US Army Engineer Waterways Experiment

Station (WES), Vicksburg, MS. The work was supported under an Intergovern-

mental Personnel Act Agreement between CSM and WES. This work was jointly

funded by the Repair, Evaluation, Maintenance, and Rehabilitation (REMR)

Research Program, Work Unit 32315, "Geophysical Techniques for Assessment of

Existing Structures and Structural Foundations," and the Little Rock District,

Corps of Engineers under Intra-Army Order No. 85-0045. The fieldwork was per-

formed during August 1985 and benefited greatly from the use of rental eauip-

ment and field personnel labor covered by Contract DACA 39-86-K-0008, "Active

Geophysical Methods Applied to the Tunnel Detection Problem." The Technical

Monitor was Mr. Ben Kelly.

CSM Principal Investigator for the work was Dr. Thomas L. Dobecki.

Mr. Monroe B. Savage, ISD, WES, and Messrs. Thomas Kertesz, Jeff Meis, and

Scott Stephens, CSM, were responsible for field data acquisition. Data

processing and interpretation were performed with the additional assistance of

Ms. Tanya L. Mueller, CSM. Technical monitoring was performed by Dr. Dwain K.

Butler and Mr. Jose L. Llopis, Geotechnical Laboratory (GL), Earthquake Engi-

neering and Geophysics Divison (EEGD), who were also Principal Investigators

of the REMR Work Unit. Mr. Jerry Huie, Engineering Geology and Rock Mechanics

Division, was the Geotechnical (Rock) Problem Area Leader, and Mr. William F.

McCleese is REMR Program Manager. Dr. Arley G. Franklin is Chief, EEGD, and

Dr. William F. Marcuson III is Chief, GL. The report was edited by

Ms. Odell F. Allen, Information Management Division, Information Technology

Laboratory.

Acting Commander and Director of WES during the preparation of this

report was LTC Jack R. Stephens, EN. Technical Director was Dr. Robert W.

Whalin.

L1

CONTENTS

Page

PREFACE ................................................................. I

PART I: INTRODUCTION ................................................. 3

PART II: GEOLOGIC SETTING ............................................. 5

PART III: SEISMIC REFLECTION COMPARED WITH STANDARD SURVEYS ............ 8

PART IV: DATA ACQUISITION ............................................. 11

Instrumentation ................................................... 11

Line C-i .......................................................... 12Line C-2 .......................................................... 15Line C-3 .......................................................... 16Vertical Seismic Profiling Data ................................... 16

PART V: DATA PROCESSING .............................................. 17

ES-2420 Processing ................................................ 17

C.G.G. (Lines C-i and C-3) Processing ................................. 23C.G.G. (Line C-2) Processing ...................................... 31Data Quality and Cost Comparison (ES-2420 versus C.G.G).............. 32

PART VI: GEOLOGICAL INTERPRETATION .................................... 35

Deep Basement Structure ........................................... 35

The Weathering Layer .............................................. 35Comparison of Field Methods Used .................................. 45

PART VII: SUMMARY AND CONCLUSIONS ...................................... 51

REFERENCES .............................................................. 53

PLATES 1-4

APPENDIX A: GLOSSARY ................................................... Al

2

HIGH-RESOLUTION SEISMIC REFLECTION INVESTIGATIONS

AT BEAVER DAM, ARKANSAS

PART I: INTRODUCTION

1. Shallow seismic stratigraphic surveying has been ignored typically

by the engineering industry as a tool for finding shallow anomalies and

delineating weathered zones that may adversely affect foundation conditions.

The Beaver Dam project attempts to show that the use of high-resolution seis-

mic data can be important to studies of fractures, cavities/voids, irregular

(weathered) top of rock, and other near-surface anomalies. Three seismic

reflection lines were shot downstream from the face of Dike No. 1, Beaver Dam,

AR, which is known to be leaking impounded water through subsurface foundation

formations. Although other geophysical methods have been used at the damsite

in hope of finding where and why leakage (Unomalous seepage) occurs, it was

hoped that high-resolution seismic reflection surveys could give an even more

accurate idea of where faulted zones are located and their relationship to

observed leakage. Further, with the Corps of Engineers program for certifica-

tion and rehabilitation of existing dams, it was felt that Beaver Dam would

provide an excellent test case to evaluate the utility of seismic reflection

for evaluating anomalies related to seepage.

2. The Beaver Dam seismic reflection project was undertaken by the

Colorado School of Mines (CSM), Geophysics Department, with several goals in

mind. Because of the known foundation structural problems mentioned, the

earthen Dike No. I has been under scrutiny for anomalous seepage since it was

built. It was hoped that high-resolution seismic data could resolve both

basement structures (more precisely define fault zones) and in addition,

resolve suspected large rugs*, voids, or zones of intense wtathering.

3. Dike No. I is situated just north of the main concrete dam and

embankment and occupies a low bedrock area which would otherwise be below pool

level. Geologically, the dike straddles a graben of estimated 305-m width and

76-m throw. This downdropped section has brought a vuggy carbonate formation

* Specialized and perhaps unfamiliar geophysical terms that are underlined

are defined in Appendix A.

3

(see Part II) down to a lowered elevation such that it is the foundation rock

for the dike. It is well above the top of the main embankment. This circum-

stance and the fracturing attending the graben boundary faults are suspected

to be the causes of seepage detected along the face of and downstream from

Dike No. 1.

4. The US Army Engineer Waterways Experiment Station (WES), as part of

its REMR research program, is investigating the use of seismic reflection for

the evaluation of existing dams and adjacent structures. At Beaver Dam, WES

has already employed seismic refraction, electrical resistivity, magnetic,

electromagnetic, ground penetrating radar, and self-potential (SP) surveys

along the dike. These surveys (US Army Corps of Engineers 1986; Llopis and

Butler 1988) have pointed out anomalous conditions at the soil-rock interface

(refraction) and localized fluid flow patterns (SP). WES felt, however, that

Beaver Dam would be an ideal site for testing the utility of seismic reflec-

tion, given the parallel refraction and extensive SP data, ground penetrating

radar, and borehole investigations already performed at this location. Given

the shallow penetration of radar and refraction (because of thick, relatively

uniform velocity stratigraphy and the physical limitations or refraction

survey line length) and the sensitivity of SP to flow and not structure,

seismic reflection surveys can provide the following information:

a. Lateral location and description of boundary and any otherfaults which affect the upper (approximately) 0 to 300 m.

b. Detection of localized cavities/voids associated with the vuggyBoone formation and/or faulted zones.

C. Correlation of the above with patterns of seepage determined by

other means.

5. An additional purpose in the seismic reflection investigation was to

evaluate different instrumentation and data acquisition methodologies to

determine optimal means for acquiring such data while retaining high data

quality.

4

PART II: GEOLOGIC SETTING

6. Beaver Dam is located on the White River, Ozark Uplift in Northern

Arkansas (Figure 1). The dam and other structures are founded on flat-lying

Paleozoic carbonate sedimentary rocks. The region is quite hilly as the White

River and its tributaries have steeply dissected these carbonates. Five of

these Paleozoic formations outcrop in the vicinity of the dam; these are named

and described in Figure 2. It is important to note that the Boone formation,

a cherty limestone, is the host rock for most of the springs and natural

caverns found throughout this region (US Army Corps of Engineers 1986).

7. The presence of this formation in the region of the dam might have

caused grave concern initially had it not been for the fact that the Boone

formation is situated at an elevation above the dam and reservoir on the south

abutment. The misfortune is that these flat-lying sedimentary rocks have been

offset by normal faulting such that Dike No. 1 occupies a graben where the

rocks have been downdropped in excess of 250 ft (76 m). As such, the founda-

tion rock for Dike No. i is indeed the Boone formation. Much of the leakage

observed is therefore felt attributable to the deteriorated character of this

soluble foundation rock and its deformation and fracturing resulting from the

close proximity to faulting.

8. In terms of candidate reflecting horizons in the Dike No. I area,

likely candidates, listed by increasing subsurface depth are:

a. Fill/weathered rock interface.

b. Weathered/solid rock interface.

c. Boone/Chatanooga/Sylamore contacts at -200 ft (61 m).

d. Dense or weak zones within the Cotter fm (Figure 2).

e. Crystalline basement (suspected depth >1,500 ft (457 m).

5

- "11

- -~ -- -- _ -- -

S 8 S 'T U

0 0 CP AT I E o-E s r N

SER

01 T E .

t;f E3

0 ' KE NO I2,

SO

H

KIRK( MOUNTAIN N, j~- EQuARRY SITE

R

f) T

A ___l __ __ _

I' NoD0

M N..s o. . . .... ...

Figure 2. Strat 4graphi'.column at Beaver Dam sit

G F 11 i I I -o

A'400 001059 W C-41- 0

C' "E ND 2, coo ~21t' .4 0 isl ..o xo4O,0 0 ,s50)DiKE NO ~ ~ ~ ~ ~ ~ ~ ., 3-- - I. _ t. T'.. o .t..oo "~40.K~qwi.

2.0*1 D A M -i ' ' 2 6 !0 5 14 0 *OC0.Z11 T4 c0 1 3 O h

D'K. NC:)SO 2'00 ... ,IS. so4 0

41.: 1. s t~f 4o 1-to 1 E. A04,.53

F 1 ~~ 4± 5 . o. ~'z o1t

D E 0 2± 7 ~ .qe..o ss, o~Oiso0 t.

30tj 9 1 ;71'l04 n4' ll! -- o4* 450ll .1 1-4 BI

%BAINN~ 0D4M A- 44. 004400% l04 LIwo wn.Rm.

CONCRETE - V T ~Wii

30 000. 00 O to..OO 0040 1".4 mt..0 "I "d5 -- I

R 20+1, 3)40 M3k OO t IC "114 00 1oIC-t10 *

40000W 11 .5 0[.M0 IaS t'll-.440.

Figure 2. Stratlgraphic column at Beaver Dam site

7

Ii|

PART III: SEISMIC REFLECTION COMPARED WITH STANDARD SURVEYS

9. Traditional geophysical surveys in support of geotechnical site

characterization have been dominated by (a) seismic refraction (b) cross-hole

seismic testing, and (c) electrical resistivity (Dobecki and Romig 1985).

Refraction in P- and S-wave modes provides fairly rapid areal coverage of a

site yielding the depths (thicknesses) and configuration of subsurface layers

as well as the compressional and/or the shear-wave velocity of each layer.

This is true provided that the velocity of each successively deeper layer

exceeds that of the layer above it. Even if this is satisfied, error can

arise if a given layer is thin ("thin" may be as thick as several tens of

meters); in such a circumstance, the thin layer escapes detection by refrac-

tion methods. Therefore, either low-velo-ity zones or thin, high-velocity

layers may be missed by refraction even if the data were obtained perfectly

(noise free). From an operational standpoint, refraction becomes difficult if

the targeted refraction horizon (e.g. bedrock) is deep. The required

shotpoint-to-geophone separation ("offset") to map the target effectively will

be roughly three to five times the depth of the target. This requires long

lines, which increases the chances for error due to lateral velocity varia-

tion. This also requires higher energy seismic sources such as dynamite to

accommodate for added attenuation of signal strength over the long travel path.

Oftentimes, logistical, topographic, or political influences can limit the

surface area available for data acquisition as well as the nature and strength

of seismic sources which could be used. Finally, in areas of complex subsur-

face geology, refraction Interpretation becomes extremely complicated (Palmer

1980).

10. Cross-hole seismic testing eliminates some of the problems asso-

ciated with refraction by providing very fine sampling of the subsurface P-

and S-wave velocity distribution with depth by making these as time-of-flight

measurements between boreholes. This is obtained however at the expense of

the areal coverage obtained from refraction and with the expense of drilling

two or more drillholes. As with refraction, some error may be associated with

thin beds, particularly low velocity thin beds, but this error may be cor-

rected if the presence of such thin beds is suspected (Butler and Curro 1981).

Still, even given perfect data, no information is available from depths

greater than that of the boreholes themselves.

i8

-I

11. Electrical resistivity in most geotechnical applications finds its

greatest use in determining depths and/or shapes of saturated zones (such as

water table or water-filled fracture zones) or in locating other conductive

horizons (clay seams and highly weathered rock). Unlike the seiRpiic methods,

resistivity is not based upon elasticity theory. The only relationships which

relate resistivity directly to rock strength parameters are empirical and

highly site specific (Keller 1974). Further, since resistivity is a potential

field technique (as opposed to a wave propagation technique), its resolution

is theoretically much less than seismic methods.

12. Under many circumstances, the combination of these three techni-

ques, tied to an exploratory borehole program, is totally adequate to support

a geotechnical site investigation. The obvious situations where these would

be inadequate include:

a. Sites where required depth of investigation exceeds(approximately) 50 to 100 m.

b. Sites characterized by the occurrence of low velocity layerssuch as clay layers or subhorizontal fracture zones at depth.

c. Sites with thin or small target zones in the subsurface whichmight otherwise require an extensive boring program.

d. Sites of limited access or of other logistical restriction(e.g. no explosives, presence of surface steel structures,etc.).

For such sites, the use of seismic reflection may be appropriate.

13. Seismic reflection uses shotpoint-to-geophone offsets comparable to

the depth of the target as opposed to three to five times the depth for

seismic refraction. This allows for shorter geophone spread lengths and

smaller source energies for a given depth of investigation as compared with

refraction. It is effective at detecting both high- and low-velocity beds at

depth, although determination of the bed's velocity is not done very

accurately. Reflection can also detect "thin" beds if the thickness of the

bed is somewhat greater than 1/8 the dominant seismic wavelength. Increased

resolution implies an increased bandwidth in the seismic wave/signal. Many of

the shortcomings of refraction, cross-hole, and resistivity can therefore be

eliminated or reduced with the addition of reflection to the geophysical

investigation program.

14. Seismic reflection 1--wever has inherent problems of its own.

Principally, obtaining good -rabiguous reflection data free from interfering

9

"noise" (e.g., ground roll, airwave, refractions) can be very difficult.

Obtaining good refraction data is generally quite easy; it may not be inter-

pretable, but it is easy to acquire. Great care in instrumentation, field

geometries, procedures, and subsequent data processing, however, are required

just to produce a visible and coherent reflection on a record section. This

requires greater skill on the part of the geophysicist. Once the reflection

has been recorded with fidelity, interpretation itself is much less cumbersome

than with refraction or resistivity.

10

PART IV: DATA ACQUISITION

Instrumentation

15. Seismic data were acquired using a 24-channel digital data acquisi-

tion system. The key specifications of this instrument are given below.

Acquisition Control Unit (ACU)

Sample interval menu selected from 1/4, 1/2, 1, 2, or

4 msec

Low cut filters menu selected from 5 to 320 Hz in 5 Hz

increments; 18 db/octave

Alias filters menu selectable; down 80 db at 128,256, 512, 1,024, 2,048, and 4,096 Hz

Notch filters menu selectable; 50 to 60 Hz (60 dbdown)

Acquisition mode write digital data direct to tape orinto mass storage memory, then to tape

Run time limited only by available memory,typically 4.096 sec at 1/4 msec;8.192 sec at 1/2 msec; etc.

Channels 24

Word size 15 bits plus 4-bits exponentrepresenting 6 db gain steps

Amplifiers Instantaneous floating point

Stacking No limit to summing of successivesignals

Display CRT with menu selected fixed gain,fixed gain normalized, or digitalAGC playback

Plotter

8.5 in. wide electrostatic paper;

print contents of full CRT memory

Digital Tape Recorder

Spec's 7.5 in. reel, 1,200 ft tape, 1,600 bpiFormat SEG D; menu selected multiplex or

demultiplex

11

Post Acquisition Processor (PAP) Programs Availabli

Bandpass digital filtering of raw or

processed data

Common offset gather

NMO correction

Transient suppression

Trace edit

Mute

FFT

Constant velocity CDP stack

16. Of utmost importance were the features of 15-bit recording (high

dynamic range) instantaneous floating point amplifiers (rapid, accurate data

acquisition) and post acquisition processing capability (in-field data

analysis). WES itself has 12- and 24-channel "engineering" seismographs

which it may wish to use for similar purposes on future projects. These

seismographs do not have many of the special features of the CSM system.

However, our procedures and analysis always kept the WES systems in mind to

determine how appropriate these might be in their current format and/or how

they might best be modified to allow for this usage.

Line C-i

17. Line C-i runs, for the most part, coincident with SP Line C

(Llopis and Butler 1988). It is roughly North-South, extending from north of

the northern fault zone, across the downstream berm of Dike No. 1, across the

southern fault zone, and part way down the main embankment (Figure 3). The

line was flagged at 5-ft (1.5 m) intervals with Flag C-i at the North and

Flag C-359 at the South for a total line length of 1,790 ft (546 m).

18. Line C-I was shot in common depth point (CDP) roll-along fashion as

follows:

a. An optimum offset (distance from shotpoint to first of24 geophones) of 12" ft (38 m) was determined by test shootingat several offsets.

b. Twenty-four geophone stations spaced at 5-ft (1.5 m) intervals

were used (Figure 4). Each station is an array of six single

geophones.

12

A.-, -a4

00

LLa)

-4

13 4

c. Each geophone station consisted of an array of six, series

connected, 24 Hz geophones at 5-ft (1.5-m) element spacing.

d. The seismic source employed was three stacks of a shotgun shell("Buffalo Gun") in a 2-ft (0.6-m) augered hole.

e. After recording, the entire arrangement was "rolled up"(advanced) two flag positions (10 ft or 3 m) and repeated.

SHOT POINT ~~0 0 0 0 0 0 C0 0 0 0 0 ,0 CC

SHOT POINT GEOPHONE GROUPS

5'Figure 4. Shot, geophone layout for one shot record, Line C-1

19. The choices of offsets, geophone groups versus single phones, and

Buffalo Gun source versus sledgehammer were based upon preliminary test

spreads in the center of the graben area. The target reflector for this line

was the Chatanooga/Sylamore beds at an estimated 250 ft (76 m) depth.

20. The resulting CDP coverage, given 24 geophone positions per shot

and advancing the shot two stations between shots, is a maximum of

6-fold stack. Subsurface reflection point spacing is one-half that of the

geophone spacing or 2.5 ft (0.8 m).

14

Line C-2

21. Line C-2 overlapped part of Line C-I by using geophone stations at

Flags C-171 through C-290 for a total line length of 595 ft (181 m). The

differences in Lines C-i and C-2 are how the data were acquired and what the

target reflector was. The target reflector was the top of sound foundation

rock estimated at 15 to 70 ft (4.6 to 21 m) from boring and WES refraction

data along Dike No. 1.

22. With a shallow target, the shot-geophone offset was changed, and a

switch was made from geophone groups to single geophones. The offset should,

by the rule of thumb, be on the order of the target depth. The use of single

phones for shallow reflections is because arrays tend to smear the high

frequency components of the reflected signal if the reflection path is not

near vertical. For a shallow reflection, this path is surely nonvertical.

Therefore, Line C-2 was shot in CDP roll along fashion as follows:

a. An optimum offset of 15 ft (4.6 m) was used.

b. Twelve geophone stations spaced at 5-ft (1.5 m) intervals were

used (Figure 5).

c. Single, 100 Hz vertical geophones were used.

d. The seismic source employed was four stacks of a sledgehammer

on a steel plate.

e. After recording, the entire arrangement was rolled up one flag

position (5 ft or 1.5 m) and repeated.

In addition, ground roll reduction was obtained by employing a 320 Hz low cut

filter (analog) prior to digitization. The resulting CDP coverage also was

six fold with a subsurface spacing of 2.5 ft (0.8 m) for reflection points.

515

Figure 5. Shot, geophone layout for one shot

record, Line C-2

15

Line C-3

23. Line C-3 was also an overlap of Line C-I but from Flags C-194

through C-337 for a total line length of 715 ft (218 m). Line C-3 had the

exact same target as Line C-i and was acquired as a practically redundant data

set. The principal difference was in the source and receivers used for the

two lines. Line C-i was a time-consuming and labor intensive effort with each

geophone station requiring the planting of six evenly spaced geophones (6 x 24

- 144 geophones per shot) and the shot requiring the augering of a hole and

reloading the Buffalo Gun for three shots. The difference in data quality can

be compared between such a painstaking approach with data acquisition and a

quick and inexpensive approach. Therefore, the same offset, spread lengths,

and instrumental parameters were used for Lines C-i and C-3. Single, 100 Hz

geophones were used in place of groups, however, and nine stacks of a sledge-

hammer on a steel plate replaced the Buffalo Gun. The two lines therefore had

the same CDP coverage over a common 715 ft (218 m) length. But Line C-i

required 2 days to cover this length; Line C-3 covered it in less than I day.

Vertical Seismic Profiling Data

24. The original plan for our survey was to tie the reflection data to

well data via vertical seismic profiling (VSP) in several of the test wells.

These tests were not successful as all holes had caved or were at least

bridged over at shallow depth such that the borehole seismometer could not be

advanced. Hole BE-2 was closed 90 ft (27 m) depth; no other hole (BE-1, BE-3,

or BE-4) was open as deep as 50 ft (15 m).

16

PART V: DATA PROCESSING

25. Lines C-i, C-2, and C-3 were initially processed in rough form on

the ES-2420 field acquisition unit. The lines were then processed again using

the state-of-the-art processing software available at Compagnie General de

Geophysique (C.G.G.) American Services in Denver. Due to the limited use of

seismic techniques for engineering projects, the two processing methods were

to be compared for feasibility and quality versus cost.

26. The ES-2420 processing was limited to full trace filtering, normal

moveout (NMO) at constant velocity, and CDP stack or gather. The processing

possibilities at C.G.G. were virtually unlimited except by cost and time

constraints.

ES-2420 Processing

27. Limited processing procedures are available on the ES-2420 includ-

ing editing, NMO correction (at constant velocity), common offset gather, and

CDP stack. All of the lines were initially processed on the ES-2420 as a pre-

liminary product for inspection of data quality and for comparison with other

processing systems. The total ES-2420 processing for the Beaver Dam data

required little more than I week. Elevation statics were not applied

(although the ES-2420 does have statics capabilities). The results were

encouraging but by no means the best possible from the data set.

28. Lines C-i and C-3 (see Figures 6 and 7) had obvious noise problems

which are discussed in detail in subsequent sections in addition to statics

problems. The ES-2420 cannot do deconvolution or time-variant filtering,

which seemed to be necessary. In addition, variations in velocity, both

lateral and with time, could not be accounted for by the ES-2420. Line C-2,

on the other hand, could be processed satisfactorily using a single velocity

function and thus was less affected by the restraints of the ES-2420. Gener-

ally, good results were obtained on Line C-2 by effectively eliminating traces

contaminated by noise within the zone of interest while the processing of

Lines C-i and C-3 was far from satisfactory. Little improvement could be

achieved on Line C-2 by further processing as explained later in the process-

ing section.

17

SHOT POINT 67 SHOT POINT 7

-0.0

_0.025

AUTOGORRELATION

.125.... -0.025

.- . . .......

o0.0-0.

0.100~ .... . .

-~---~- - ~ - 0.025

0.25

0.3500

0..025

- ,, JT~',j~\, 'A~)~- 0.350

Fiur 6.Baer.m in - ufleedrcr

OB NPLITUOE SPEC Tpljti0

-12

LATION -24

-- 4

- 0.025

-60 _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

....... 6 400 F~ HZ]J

- 0.0 0 _ _ _ _ _ _ _ _ ___ __ _

ISHOT POINT 1, TRACE 12

- ~~-12 ________1__________

0.025

- ~~-24 _____ _

- 0.05 __ __ __3_ 6___ _

-48 ______

DB 400FHZ

-,unfiltered records

0

SHOT POINT 40 SHOT POINT 16-12

0.0 -=0.0

0.025 -24

AUTOCORRELATION

0.050 -___ 3

-- 0.050.075 ---. .

- -46

-- 0.025

- -- DB 0j

0 .1 5 0j

z - -0.000. 175 0

0. 200u zzzz.' -12

- _ ___ - ____ - 0.025

- - . ~ -- 24

____ -0.050 1::0.300-4

0.325

-6 0

0.350 D

Figure 7. Beaver Damn, Line C-3, unfiltered records

RM1PLITUDE SPECTRUM0

1

SHOT POINT:16 TRACE:12

-24

TOCORRELATION

- -.- ---. 05

- - - -0.025

DB 0 400 Fi HZ'

-0.0

SHOT r' N' C 7PCEi 12

- 0.025 -2__ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

-24

-0.050 -35

-48 ___ _

-601 1 -

DB 0 400 F HZI

r Dam, Line C-3, unfiltered records

29. The ES-2420 was a good processing tool because velocities and trace

edits could be determined prior to the more expensive C.G.G. processing. This

type of processing could be done in the field immediately after acquisition to

determine any problems that could be corrected by changing field parameters.

C.G.G. (Lines C-i and C-3) Processing

30. Lines C-i and C-3 were intended to allow a direct comparison of the

best result from the two distinctly different approaches to acquisition. The

C.G.G. American Services Geomax software was used in an effort to give such

shallow high-resolution methods a fair opportunity to be tested with state-

of-the-art processing modules. The processing sequences for the two lines

were made as parallel as possible for comparison purposes.

31. An outline of the processing sequence for Beaver Dam, Lines C-i and

C-3 are listed below.

a. Demultiplex and reformat.

b. Gain compensation for transmission loss and sphericaldivergence.

c. CDP sort.

d. Geophone edit.

e. Shot and trace edits.

f. Statics to flat datum.

. Spiking deconvolution.

h. Operator length = 40 msec.

i. Window from 62.5 to 250 msec.

2. Automatic residual statics.

k. Velocity analysis.

1. NMO correction.

m. Mutes.

n. Automatic residual statics.

o. CDP stack.

p. Time variant band pass filtering, 72/88-240/280 Hz, TO -

TI0 msec, 72/88-180/220 Hz, T150 - T500 msec.

s. Dynamic trace equalization.

r. Statics to floating datum (end point for final stack).

s. Wave equation migration.

23

t. Band pass filtering.

u. Dynamic trace equalization.

v. Film display.

32. The sequence is similar to a simple sequence for normal deep

seismic processing but with some differences. The major problems with these

two lines were:

a. Optimizing filters.

b. Removing airwave noise and groundroll.

c. Choosing velocity functions near surface.

d. Resolving statics.

33. One difficulty with the data acquired in engineering geophysics

projects is that the acquisition methods differ radically from normal deep

seismic reflection data acquisition. The high-resolution data are often

recorded at fractional sample rates such as 0.25 msec. Also, the data lengths

are typically very short, such as 500 msec. Although this might not seem to

be a problem, it can be when processing with software tailored to normal seis-

mic reflection acquisition parameters. Many processing packages cannot handle

decimal sample rates as in this Arkansas project; therefore, the headers had

to be changed on an alternate computer (a Raytheon) to "fool" the system into

believing that the sample interval was I msec with a data length of 2,000 msec

instead of 500 msec. All subsequent processing was then done as if the data

were 2,000 msec. Velocities and frequencies, entered as one-quarter of their

real values, had to be converted back to real values before analysis. The

short data lengths were also a problem in some cases. Line C-2, which had a

data length of only 200 msec (800 msec in processing time), had problems in

filtering solely because the application window of the filters for optimum

efficiency was nearly as long as the whole data set and, in some cases, longer

for the lower frequencies.

34. Extensive testing was deemed necessary to resolve these probiem,

particularly because high--resolution shallow seismic data processing has been

relatively unexplored.

Filtering

35. Both of these lines had 25 Hz high pass field filters. Figures 6

and 7 show the amplitude spectrum, autocorrelation, and a pair of raw records

from each line. The data are dominated by low-frequency groundroll (Rayleigh

surface waves) and high-frequency airwave. More airwave was experienced where

24

• l I I I I i

shooting over hard ground occurred because deep holes could not be drilled. A

low cut of 36 Hz and a high cut of 280 Hz were used on the data, as suggested

by a harmonic filter test. Both lines were improved, but a bad data zone

still resulted from the airwave and groundroll, particularly on Line C-i. It

was hoped that an F-K velocity filter would eliminate the noise without remov-

ing reflection data within the upler frequency range. While the airwave was

obviously slower velocity than any data, this method proved to be fruitless

due to the fact that the frequency of the noise was very high. Thus, when

transformed to the F-K domain, the wave number over which the filter was

applied corresponded to a frequency which was far below that of the noise.

This type of filter works much better on noise that is low frequency as well

as low velocity such as groundroll. The only alternative was to lower the

high- and low-cut frequency and use that to reduce the frequency content of

the data. Figures 8 and 9 show the final filtered records and their amplitude

spectra. The bulk of the amplitude spectra lost at high-frequency noise.

These were the final filters used on these data.

Velocity analysis

36. Near surface reflections were very sensitive to velocity change

while deeper events (> than 100 msec) were completely insensitive to change

because their moveout is nearly flat for such short spread lengths. Unfortu-

nately, slight changes in the velocity of shallow reflections change their

character but do not identify necessarily which velocity is correct. The

lateral velocity variation reflects the boundaries of the graben and the

vuggier zone between them.

Deconvolution

37. A spiking deconvolution was used to collapse the wavelet using an

operator length of 120 msec. Several operator lengths and white noise levels

were tested. The data were rather steady state in quality once a length of

120 msec was reached; therefore, while deconvolution is necessary, little

testing is necessary. A medium length operator used in a spiking deconvolu-

tion routine with about I percent white noise should be sufficient as long as

an impulsive source is used.

Statics

38. Near surface problems, including those created by fault zones, had

to be resolved on these lines. Two passes of surface consistent of short

wavelength automatic statics were run to resolve problems created by

25

SHOT POINT 67 SHOT POINT 7

... 02...5 AUTOCORRELATION

0.050 --

0.075 ___-0.05

0.100 -__ -- 4

-0.125- _ _ -0.025

0.150 - - __-

OiE

0. 17 -0.0

0

0. 200w

0. 225 U. -0025

0.275 -- 0.05 -

0.300- _ _

0.325

0.350

Figure 8. Beaver Dam, Line C-1, time-variant filtered re,

05 RNtPLITUDE SPECTRUI

TOCORRELATION-2

-- 0.05 -

_____ 0.025

________F(HZ

OB 400

___ __ _ 0.0 0 - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

JSHOT POINT T. RACE 12-12 -

_______ __ _

-0.025

-0.05 -36 -

-48 ____

-60____ _

DB 400 Ft HZ)

;A1ne C-1, time-variant filtered records

OB

0

SHOT POINT 40 SHOT POINT 160.0 -12

0.025- AUTOCORRELATION-2

0.050- -0.05

-- 0.22-5

S-------60

0o.50- 0.0 O8

W -

0.20-L 0.025 -12

E0.225 - __

_ - ____-24

0.250 -- 0.05

E0.275-4

o.325-6

0-350 D

Figure 9. Beaver Dam, Line C-3, filtered record

08AMPLITUDE SPECTRUM0

SHOT POINT 16 TRACE 12

-12

OCORRELATION -24

_ 2 -- 0.05

- -36

-- 0.025 -48_ _ _ _ _ _ __ _ _ _ _ _ _ _ _

-0.0 06400 F eHZ)

0 _____________________________

-SHOT POINT 40 TRACE 12

-aA 0.025 -12F

- -~~24_ _ __ _

-0.05

-36 1 -----

DB 400 F (HZ)

er Dam, Line C-3, filtered records

perturbations in the weathered zone. This was particularly important since

elevations were not surveyed. Elevations were hand picked from topographic

maps and from elevations of piezometers spaced along the line. Future engi-

neering projects would likely benefit from accurately surveyed elevations and

station locations. Elevations in this case were probably within tolerances of

1/2 ft (0.15 m). In addition, weathering velocities are well known based on

refraction surveys; thus, datum statics are probably very good. Correct

weathering velocity is crucial for shallow reflection surveys to the extent

that, if necessary, reconnaissance drilling or seismic refraction surveys

should be commonplace before such work to avoid large errors in datum statics.

Migration

39. Although sedimentary dip was not really a problem at the Beaver Dam

site, migration was deemed necessary to collapse any diffractions from fault

planes. A wave equation migration routine was tested beginning with values

recommended for traditional seismic work and adjusted from that point. Several

combinations of downward continuation step size and velocity function were

tested. While the data were somewhat sensitive to velocity, the most critical

value was the increment for each downward continuation iteration. Due to the

frequencies and times involved in these data, the step had to be smaller than

normally used even in steeply dipping areas. The best migration obtainable

within the time frame available to use was achieved at 90 percent stacking

velocities with an increment of about 6 msec per downward continuation

iteration.

40. Further investigation of migration for shallow high-resolution

seismic data is needed. It is possible that other migration methods such as

finite difference migration or partial or full prestack migration might be a

better method for these data. These methods were not tried in this study due

to time constraints.

C.G.G. (Line C-2) Processing

41. Line C-2 was shot expressly for the purpose of examining the near

surface, or more specifically, the weathered layer/rock interface. Addi-

tional processing (over that done on ES-2420) was limited to filtering, very

minimal muting, constant velocity moveout, and stack. Deconvolution was

abandoned because the low signal to noise ratio would not allow the filter to

31

stabilize. Data were not corrected for statics other than datum statics. Inthis case, the ES-2420 processing was probably sufficien t, in fact better than

the more expensive option. This was mainly in light of the fact that the SEGD

tapes had header errors that would not allow C.G.G. to read all of the records

making this data set incomplete.

Data Quality and Cost Comparison (ES-2420 versus C.G.G.)

42. Lines C-I and C-3 were most obviously improved by the additional

processing at C.G.G. Figure 10 (with the ES-2420 section of Line C-i on the

left and the C.G.G. processed section on the right) illustrates the character

differences between the two methods. Time scales are similar; the C.G.G.

horizontal scale is about one-half the ES-2420. Most obvious differences are

frequency content (ES-2420 dominated by groundroll), C.G.G. migration elimina-

ting the large diffraction (at 75 on ES-2420), and better definition of faults

on the C.G.G. section. The more advanced processing techniques improved the

frequency content of the data by about 130 Hz. Most likely, the capability of

using multiple, time-variant velocity functions can be credited with the dras-

tically improved frequency of the data. On Line C-1, time-variant filtering

reduced airwave noise without diminishing the data quality.

43. Continuity was improved by static corrections, thus emphasizing the

fault zones. Time and space variant velocities also improved the continuity

of the data on both Lines C-I and C-3 to well beyond that achieved on the

ES-2420.

44. On Line C-2, the different processing methods made a minimal dif-

ference. It is possible that for weathered zone surveys, field processing may

be sufficient, mostly because constant velocity functions can be used. How-

ever, on these shallow surveys, correct elevations are most critical and

should be incorporated via datum statics application (available on ES-2420,

but not used in this example).

45. In terms of cost, processing on the ES-2420 involved about 2 weeks

rental cost or about 7,000 dollars, which is about 6 dollars per stacked trace

on Line C-i. The processing at C.G.G., which is probably representative of

computer time cost at any data processing service company, came to about

7,000 dollars including filming and copies. For the processing alone, this

amounts to about 7 dollars per stacked trace on Line C-i.

32

(S(NO03S) 3WII1

0 0 0

A..._A A i . A

-. AA I .. A. ,1 *.... * A ...*a, - a a, AA A

t i.A. A.

1 & & .*a& aL -i hA.".Z.- i- A.AA L

A-]] i AA i.Ai -A&A A .. ka. 1 CiA$.-~' -h ~ &£ A A

LAAA . A A. ' I A A A 0

A IA- A- - ', ' " - A... ,A A .AL-_A I - A A A

A.. A A..A*A . 44A. A .& - A- - t - - I a 4

a A..A.A r~AAA&.co A A AA ... A ~AA,

AA4 AAA * A1-A-A A.A A~ A A- A A A..

.A* A A....h L A A.

.. & A A * - A AA. A ~ALR ~ 4.&AA... A A .

..4

A 4"14. .4A&A AA k A AAA. A.A a4 A5 AA

a. a..h.- i*A . .A ~ A

e*e ol a Ii ii i • •

4-)

4-4 -

-4 ,

0 -H

-,I *.4

U 4-

00

4

0

C)

4

0

4-4

Ln OD

C! CC

33)

46. If the ES-2420 is owned and is required in any case for data

acquisition, its incremental processing cost is limited only to operator time.

If ultimate resolution of deeper structure is a survey requirement, then the

cost of outside, sophisticated processing is justified. For most surveys

involving structural interpretation within the upper 200 to 300 ft (61 to

91 m), the ES-2420 processing itself can give rapid and useful sections for

interpretation.

47. Relating this back to compatibility with WES seismograph systems,

it is apparent that the capability of doing shallow reflection analysis (as

with Line C-2) is possible because some microcomputer-based software can be

purchased or written which will mimic the ES-2420 capability. The deeper

investigations would be at a severe disadvantage unless data can be trans-

ferred to tape in a standard SEG format for advanced processing.

34

PART VI: GEOLOGICAL INTERPRETATION

Deep Basement Structure

48. Line C-I was surveyed to look specifically at deep structure

(faults) affecting the dike. Figure 11 presents reduced and interpreted CDP

stacked sections along Line C-I. The sections presented are the final CDP

stack and a migrated stack section. Full-sized and noninterpreted versions of

Figure 11 are included as Plates 1 and 2. The colored bands were chosen to

represent continuous reflections and, because of no available well control,

are not tied to specific stratigraphic horizons. The continuous strong

reflection along the central portion of the line represents the

Chattanooga/Sylamore sequence.

49. The key observation along this line is the evident and substantial

faulting. The known major north and south bounding faults are actually series

of faults of rather complex character. In addition, there are several

observed faults within the graben which had not been noted in the drilling

program. Using Line C-I, the positions of all faults were located and thereby

transferred onto the section from Line C-2 which only looked at shallow

horizons.

50. The large boundary faults of the graben were previously interpreted

from borings and other geophysical methods. Plan maps showing locations of

boreholes are in US Army Corps of Engineers (1986) and Llopis and Butler

(1988). Borehole locations along the C-line are indicated on the seismic

sections. These faults are located near boreholes B27.9 on the North, and

G29.3 on the South. Other faults exist in the graben at sta 301 to 321

(Flags 166 to 176) and between sta 351 to 371 (Flags 191 to 201). The faults

are indicated by disruptions in continuity of events, changes in wave form,

and offsets in events. The throw on these faults within the large graben is

probably small, but it is likely that these zones of weakness are prone to

leakage through the foundation.

The Weathering Layer

51. Figure 12 is a CDP stacked section along Line C-2. It is plotted

in reverse direction from Line C-i because it was processed only by the

35

400 FT

Mi L

I U if

~ .

I. ~ *2

M-5 gp.... ~ D-

~40 FT

to, ...... r!$'' r*

---... * *-9

b *

I ~ ~ M U g 9 1

____________________ ___________ _________________

-~ .. *8 _________

*~*8~8**~bbI bI~p8*itpb I, 1 sII, 1 ~'~~fs,~.~.:~I ~ ;:~ : .. 9b1511 ~llI89999,,9bh&.&IS iU~SZ~bIIIII ISIIIIIII~tIII'~up~4~ts- ,

21 T 9.l* - lb. Z I.

.~ 4.... .~ ~*~ * ~i!.lI~ 99.9 ~ ..h.N2.~9. R M .9 Mi __M

Fiur 11.CDsetos(peisfnlta

fro LinC1,BeaerDa

13 1 omaamm. cgSomtn iw

NO-Lc Jew FT. m ls 4" man

ato 0.0 RCSIGSQE EUATERUAYS EXPERIMlENT STATION

I. 84WNWt EAUER DAM ARKAN4SAS

~, ~ 010 419 to RS LINE C-1~ m m u. Lot

.L* swammc anma wai

*~ - .2 Z iff 5TWR3______

I::.L it Iol-.sa- .e. "mum su.FINAL EQUALIZED STACKw FIELD RECORING~

z.-n 0.3 C ,, 5

41 WILL: is mcit M R um u.l~t

plmm -. s .G.:: OI a easmg. M mmu us .v .ow S es-rn 10 inE tax 0C &

10- maw lift W s CID. D. Uil a. ow.nF. E.9W.l0.4 Ift" MIL10s- T . Jk I ami.6 flimm Lv s SLM Im s

6 LAU u8T r

st mumP ANm Co o% mu u.-., coitmusa s..Tu mGfu 00"S' bWCAM.i

~~. G~I~j EOMSTER

* - 0. ~ vaUATERUAYS EXPERIMENT STATION0-user,555 m

... ~ 0.1 u =1wI LIN C-"W-610 "ftmn n lo, IN41110.*11tim Tommspo LU.

lom I L4Ur . 1 OIL-5- mmm owUr meting m or

IF :@,:O-' 11111m STOm KS

, P0.2w -z in-emit r~mlbW - N%..4Mm SWIMD TOM .0m

.) ftNfo g a.l..z wi Wsu .11118,1111 W. OI L y~ Irf mor wUmn EQ AT O MIRA IO

5. ."-.~~* 0mu . fmg . o m us mwmw 5ms-u m u u104= FIELe URECORDinINGn..~ n~~~L mmum - UWt rn 5ftm s niimnSmiu

DISLA .u'.IIET5II ~ vaa"am .S n. p"k m

.of!- u4110.owt a mt , A.s 111cti va-

...... .sO.'. .b b...pW ,

* ~~~ tw-blI be.-. ~ ~ .f5

*5P5.

be-- **,~ . ' ~ *~~ *

6.1a

ral-

S-si

vfg. g :2

ZU

Fiurep. G

Iv I,.4(

461 491 urn i ~ fi all los ma m 4

ob.0 81 #po ..gS . ... ..........

IT *to,

gr -4V

agar

PS. I4

.NAP__ t C~~** t3ht

I *I 3.11. u

U * I

M, M. n. .; -

I (CHontin'111led)l

(23

ISO 190 200 022020

AV r7 I 25

j 12

z

UU

2 ...LUI~~~ , , , . -*u .

z

LuJ

01

Figure 12. Colmmon offseton E-40seismc

230 250 260 27o

230 250 260 6 270 260

Common of fset (upper) and CDP stacked sections, produced.ES-2420 seismograph, from Line C-2, Beaver Dam

ES-2420 system, and plotting direction could not be controlled. The ES-2420

was used because a large portion of Line C-2 data was destroyed after acquisi-

tion and not available for further processing. The key reflections on this

section are the bottom of fill/original land surface interface and the top of

competent rock. Superimposed on this section are the faults determined from

Line C-I.

52. Figure 13 is the result of shallow drilling (US Army Corps of

Engineers 1986) in this area which shows striking correlation to Line C-2.

The reflection data faithfully reproduce the deep fill section and level com-

petent rock interface near boring E29.5. Borings indicate that the weathered

layer is much thicker when the faults reach the surface. These zones are

typically where the majority of leakage through the foundation occurs. It is

desirable then to locate other faults along the foundation and their corre-

sponding weathered zones near the surface.

53. The very shallow data on Line C-2 does not give a good indication

of faults directly, but the data show a good profile of the boundary between

the weathered zone and solid rock. Several zones of deep weathering are also

seen over fairly narrow lateral limits. Blue lines in Figure 12 indicate

stratified rock. The green lines indicate the interface between weathered and

solid rock. Deep weathering zones correlate with fault zones. The mechanical

fracturing due to faulting has increased the susceptibility of the bedrock to

dissolution, cavities/void formation, and enhanced weathering.

54. Seismic reflection therefore has been able to detect shallow zones

of deep weathering (to approximately 25 m depth) in the bedrock. When coupled

with deep structural data, the weathered zones are seen to be fault con-

trolled. Subsequent drilling for piezometer installation and exploratory

boreholes revealed brecciated rock core (gouge) in the interior of the grdben

in the vicinity of the 'faults' mapped by seismic reflection. No mappable

offsets were detected by the drilling interior to the graben, however.

55. Engineering seismographs could have produced equivalent results

along this line with a few modification. High pass analog filtering (pre-A/D)

on the order of 200 to 300 Hz would be necessary. Data would have to be taken

in CDP fashion which would require specialized cabling and a roll-along

switch. Finally, software for correcting NMO, CDP sorting, and CDP stack

would need to be either purchased or written.

43

CM'

Top Dike w

Top Sound Rock f.

30 mFigure 13. Results of shallow drilling program at Dike No. 1,

Beaver Dam, in area of seismic Line C-2

44

Comparison of Field Methods Used

56. One result of this project was a comparison of field acquisition

methods for high-resolution seismic. The final stack and Migration on

Line C-3 are given in Figure 14 (also, Plates 3 and 4). This can be compared

with Line C-I directly for quality (Figure ii). While some variations in pro-

cessing were necessary to optimize the results from both field methods, pro-

cessing was as parallel as possible. Figure 15 shows Flags 181 to 256 on both

Lines C-i and C-3. The most obvious difference between them is energy pene-

tration and signal to noise ratio. Figure 15 is over the southern edge or

Dike No. I where the most fracturing occurred at the graben's edge. The con-

tinuity of events is much better on Line C-i. While neither line allows dif-

ferentiation of all the individual faults over the zone, at least major

sedimentary features and faults can be discerned on Line C-i. On Line C-3

between Flags 214 and 244, little interpretation can be made except to call it

a "disturbed zone". On Line C-i over the same interval, several large struc-

tural features can be interpreted. It seems that the extra time and money

involved to collect the data by the method used on Line C-i is worthy for the

additional resolution. In addition, Line C-3 is really a better case since

velocities picked on Line C-i were applied on Line C-3. Normally, the proces-

sor would not have been able to pick velocities as accurately for Line C-3 if

it alone was available for analysis.

57. Therefore, Line C-3 would probably have been adequate to locate

major faults in Dike No. I area in this sense, rapid and inexpensive seismic

methods would have been sufficient. However, in seeing the improvement solely

due to extra care and the use of geophone arrays gained on Line C-i, Line C-3

procedure cannot be recommended. For future surveys, easier to manage arrays

could be used; also, the sledgehammer would probably be a very good source.

Time savings could be realized while retaining high-quality data.

58. Rapid seismic profiling for shallow, foundation description, as per

Line C-2, is the preferred method. Using geophone arrays instead of single

geophones smears the resolution of shallow reflections by combining the higher

frequency components of the signal in an out of phase manner, thus reducing

ultimate resolving power. Further, the sledgehammer source was certainly ade-

quate. Care must be taken to minimize airwave energy, reduce groundroll via

high pass filtering and high frequency geophones, and predetermine proper

45

384 if144

8I s u .1 Ja u s e sit I61 in1 t1 qm It wZ III af III a %

N I I N 1 mil

-.- .._...'... .. '. .. :.... , 1i !, ..,..,.. .... r . ! ..j.:, _8.

si "*Lts! .. 7Nb* --

0. 0 **L .s

.... ft.E5 :~ Pa !L .0 5

* ~;.. ~84 U35.% a I

•**3 ~ 5 _, .

RA 9 K"-, '*

*N33- ~ .**.

'3~ * .?*

':" . . ... ,,. ..... •..... ............ . .3,,, ;.:..... .,, ::, ...... . . .. '-,- .. . ,, *: ..3 ...... ,.. . .

*,. .. 8'.,., 39,. ... 9:3 ' ... b g*, .* 3 *,

,. - . ... . * . .:...s *s 359 ~ *. .. :.,.'.,,. . . .*-.3.. : , *... ,. 8I ,,. fl* '...:.....

.... , _i.- "% L $t-. 3 *

* . * 3 . . g " g 8 8 3 p l 3 3 .. ... . *.8

•_ _ __" .. 5 S '3',. . .:.. .,.I -

"" - ****.., .,.- r ,... 1.,Lt%-- '1 L8 ' 1 *.9 * J**"

Figure 14.

300 FTS F LAOSCMPTN

I 4 I a" 00 St. Gotag ii

[m& illi U* HfIRE Htifflftflf*i+It114111fIIr TEK. -a ago,

.... .. *,;.f b~ ~0.1 0 MR".? onflI'4

'Lf .- Pl- raw61. is m MU

b F0*ii.4

.,-awl TO TS SfRVTR

~~~~~~ ft _ V.~ia .M; 4oa .m .

~~~~~'~$- 'or, my.m ~ si m;4454 4m

.1*- ~ ~ ~ ~ IU S 4 fss .,M w amI I UW1. .C~&.I P.. ~ __ -'ik~e - - - ornS.0B:?.% ;:_______________________MUM

>-11WEVtA it," 1011 MO a a i tmm

. . . . .. .-.. -. P~r .ll PG MalOI<A# 4m

I......'~~~~~1 gotiiw UmsU@rjt

-2 M_ 0.

.,4 !LA. .olTN

M.,.'4 W mu ~

.58 PLhS CDMPLA UPTG

__1__ ST..REM I CILCIM L0.40 PL": Tell,~1 Ff onuttmm 'WL aU- . no t.L.U-

TWO~~~a Lt 2-9auPI

S ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ M stlot ecins(peri fnlstcMIoe i igaeHsakfrom~~G*o LieanBevr a

9 6C40 sa

I"* owAN St.. SOUN iRN.M

0.0 PROCESSING Se.EIC!

Of-WaluquUATERUAYS EXPERIMIENT STATION

w UER DAMAKNSAS

W14111115S MI AIMs

L) Gl- lwmwa MLINE C-3wPMI Leave , FEW LOG

0~Iwx IIM S I WKS 2ws

0 2 13 IIiua sMM. Siam IBMO-z s.-w $%I=

IS 111M WLWWMFINAL EEXIALIZED STACK

- FIELD3 RECORDING

o 3 '""la wts u : tovoFew TS% 0111 nill sau am<~a eel:v ,*S lamin m PA.6 WIU 11501= S at WEWS

3:AN to~a Sa. w1s 6n

oQUALI TY CONTROL 12111Wilt Cog oup.06 FINKS y. Too N

0o i.1I 1= coA A.~ L:O .iinSKNi. woeno I fy. SAIVI~Ma enlA.l.tO 4 Ooci n 12 r WI . T.a.. Ia mebsl mImIN LC. a IS 11.0: To o~

41. 11 I -or FI 21,0: is miniAw Go FI

COMPUTING

as" in An cog eta 011:11l .l Comm1611 IATm ST.. 1113111 GOWnGm. Pip: urn I. mue. ommml,. Mw"I

-LAY..D.Tyu GEOflRSTER

o 0 PROCESSING SEQUENCE

414-41341.111,16"usi UATERUAYS EXPERIMIENT STATION

I. BEN"m DAM mAR

0.1 0C-ow -0 111Ilmm

ugo. i-a eraw LINE C-3

WOR - ama M.LO

oj -111111 STTCIV MM ITEEUTON1IRTO

m.. E mm M eS _ __ __ __ __ _ __ __ __ __ __ _

0,2 zc ".3-W NKIA F 190 REKII.-IFWUA $101111nwma a

OISLA Tof'fj3 ovn FIL mmS FIELD RECORDINGso apaom m :ftem miel.

0,3u swim : oncn ciops flw me. I~~~rM mm..en .n

3- mmm fi is mmM1:11 INim as min~ IS"SW

0 4 ULIT OTOL' o Mm"a

SA $8. 111a

s migrated stack)

AN10I M. OS n

liii IIl I'I I6:I

._. ........._ _- .--.- .-. _: ., . . 0

.......... p..*.*

.. , , .. .. .,,.A 'i ' ,;.. .. .•*,..,... ,;.,..... .:: ...... .

* •0

. , *..: ... . . .

Y.;,.;; .LU

L.L

rggZ,~ AA 0

A -Z ...-- -. ,. .

S-~. ~ ~ 0, 4

300 FT

.7 .. ..... ... .;* ' . .,l S ., S' . 2' ; , " . ...,

. . .. ,m z

U

:: i - 0 .2

.-. :LL

.,..-v, '.,;: --0.3

.° .o*;.9 <:

.... ,4~~.,~~ 6' ~ j:.' .*:. 0.2"

0.4

Figure 15. Direct comparison for identical processing of seismicsection from Line C-i (top) and Line C-3 (bottom) for the same

subsurface coverage

49

shot-geophone spread geometries. Shooting in CDP fashion dramatically aided

interpretation and resolution on smaller anomalous features. This would how-

ever require some specialized equipment (CDP cables and roll-along switch) and

some minimal processing capability for CDP sort, NMO, and stack. These are

readily available from equipment/cable manufacturers and can be purchased/

written for IBM-type microcomputers. Such measurements can therefore be

easily made with even a 12-channel engineering seismograph with minor

modifications.

50

PART VII: SUMMARY AND CONCLUSIONS

59. Seismic reflection data taken at Dike No. 1, Beaver Dam, AR, have

shown that the dike area has been affected greatly by faulting whose influence

extends all the way up to the bedrock surface. More faulting than suspected

ay early drilling programs has been detected. Each fault is also the locus of

deep weathering of the dike foundation rock. The soluble vuggy Boone forma-

tion not only has been dropped down to form the dike foundation which causes

seepage, but also has been fractured by faulting to increase the dissolution

which adds to the problem.

60. Common depth point data acquisition techniques were generally

proven superior to the simple common offset methods of Hunter et al. (1984).

*Geophone arrays were preferred for deep structural reflection data, and single

geophones with high pass filtering were preferred for shallow foundation

investigation. The high-resolution seismic reflection surveys were successful

in delineating a weathered, irregular top of rock beneath Dike 1 at depths of

20 to 80 ft (6 to 24 m). The deeper structure (\200 ft) of the graben was

also delineated. Also, evidence of previously unsuspected faults in the inte-

rior of the downfaulted block was observed in the seismic sections.

61. For the deeper data, advanced data processing methods proved supe-

rior to using very simple filtering and stacking techniques. More simple

techniques however were actually preferred for the very shallow data. In

either case, deep or shallow, some processing capability coupled with adding

CDP acquisition capability will be required for the use of engineering seismo-

graphs to acquire similar reflection data on future projects.

62. High-resolution seismic surveying is a rapidly advancing methodol-

ogy in all areas: engineering seismographs, field procedures, seismic

sources, data processing procedures, and office and field-portable microcompu-

ters. Even during the time since the fieldwork at Beaver Dam was performed,

significant advances have been made. These advances have resulted not only in

lower cost but also in improved data product. A high-resolution seismic

reflection survey, covering about 1,000 ft (305 m) with 12-fold CDP,

4 ft (1.2 m) geophone intervals, and 4 ft (1.2 m) shot points, would cost

about $12,000. This estimated cost is for a site within a day's mobilization

distance from the contractor and for a simple site and objective, i.e., easy

site access, no elevation surveying required (site either flat or survey

51

II

data supplied to contractor), and a well-defined mapping objective (e.g.,

irregular top of rock at 30 to 50 ft (9 to 15 m) depth and shallower anoma-

lies). The cost of the work conducted at Beaver Dam was approximately $35,000

and included multiple mapping objectives as well as 'research' objectives

(seismic source evaluation, geophone array optimization, data processing

requirements assessment).

52

REFERENCES

Butler, D. K., and Curro, J. R. 1981. "Crosshole Seismic Testing Procedures

and Pitfalls," Geophysics, Vol 46, pp 23-29.

Dobecki, T. L., and Romig, P. R. 1985. "Geotechnical and Ground-WaterGeophysics," Geophysics, Vol 50, pp 2621-2636.

Hunter, J. A., Pullen, S. E., Burns, R. A., Gagne, R. M., and Good, R. L.

1984. "Shallow Seismic Reflection Mapping of the Overburden-Bedrock Interfacewith the Engineering Seismograph-Some Simple Techniques," Geophysics, Vol 49,~pp 1381-1385.

Keller, G. V. 1974. "Engineering Applications of Electrical GeophysicalMethods: In Subsurface Exploration for Underground and Heavy Construction,"

American Society of Civil Engineers, NY.

Llopis, J. L., and Butler, D. K. 1988. "Geophysical Investigation in Supportof Beaver Dam Comprehensive Seepage Investigation," Technical Report GL-88-6,US Army Engineer Waterways Experiment Station, Vicksburg, MS.

Palmer, D. 1980. "The Generalized Reciprocal Method of Seismic Refraction

Interpretation: Society of Exploration," Geophysics, Tulsa, OK.

US Army Corps of Engineers. 1986. "Beaver Dam, White River Arkansas, Seepage

Investigations and Remedial Plan to Control Major Seepage at Dike No. 1:Little Rock District," Little Rock, AR.

53

131 1111101 918 161514

~b L~ 'I ~O l3,10 Po *jow',Po b-01o l" .4 joo

p~b F lo0 Poo 0 pbok000 o

P0oto optb ,0PU 1

p1

;

,b~p .1100.0WO, P"

I'llh$b Ii .hi I uht P I bi.o PIP

', " -orbI b I 0 w 14

O ppollJ 10-.00?

"I. PP*.T~~wj I

16 FLAG S COMIPUT ING

I STATION~S COMIPUTED AT CGG DATA PROCESSING CENTERI6IS CI4AfPA ST.. DENVER COLORADOJ

FLOATINGDATUM PLANE: 1080 FT. CORRECTIONAL VELOCITY: 2800 FT.'SEC.

- ~ 0.0 SEC.

01 -DEMULT IPLEX1,1 potlP 0 0002000 ,lbO-GAIN COMPENSATION FOR TRANSMISSION LOSS10 ft.il 0,)Op00 AND SPHERICAL DIVERGENCE

v, - 0 0 0301-COP SORT0 pot 0 NoQ4-GEOPHONE

EDITS~ ~ 05-SHOT AND TRACE EDITSwo to 0.1 07-STATICS TO FLAT DATUM

r li 08-SPIKING DECONUOLUTIONOPERATOR LENGTH 40 MSEC.lop po __ ~ t 0' PltINOOU FROM 62.5 TO 250 MSEC.

09-AUTOMATIC RESIDUAL STATICSwit lk W10-VELOCITY ANALYSIS

'4. ,~r a.?11-NMO CORRECTION

lob 1 12-MUTES

~ 0.2 13-AUTOMATIC RESIDUAL STAT ICS~ 14-COP STACK

It, lot101 )w15-TIME VARIENT BAND PASS FILTER1 1 t . 72-88-2401280 HZ. TO-TIOD MSEC.

.'Peo's.16-DYNAMIC TRACE EQUALIZATION

b4,RI~ b b 0CP it 1 p*' * 17-STATICS TO FLOATING DATUM.1180 18-FILM DISPLAY

,o pl''wi I DISPLAY PARAMETERSto 1 V.ww"' :0orbrt bHORIZONTAL SCALE: 12 TRACES PER INCH

* .' )p~~J~~, W~~bP~~,'VERTICAL SCALE: 14 INCHES PER SECOND* j'4 po POLARITY: FIELD POLARITY

r lop1 QUALITY CONTROL;o. P, DATE NOVEMBER 1986CGACTND 4230*l~w 04 PROCESSED BY: TANYA MUELLER CHECKED BY* DR. T.L. DOBECKI

,"woo.b~ MONROE B. SAVAGE

It. ---------------------------------------------------------------------------------------------------------------

33 TP. = 1;4 MILE

PLATE 1nPUT I NG

ilSING CENTER

LORADO,ECTIONAL VELOCITY: 2800 FT./SEC.

)MASTER!NG SEQUENCE UATERUAYS EXPERIMENT STATION

COMPANY

ON FOR TRANSMISSION LOSSVIUERGENCE BEAVER DAM ARKANSAS

AREAEDITS

DATUM LINE C-iUT ION L INE40 MSEC.

5 TO 250 MSEC. 644UAL STAITICSS STAT IONS

UAL STATICS SOUTH NORTH

fNO PASS FILTER

I HZ. TO-TIOO MSEC.

Hz. T15O-T500 MSEC. FINAL EQUALIZED STACKUAL IZAT ION

ING DATUM FIELD RECORDING

PARAMETERS RECORDED BY: CSM CREU DATE:AUGUST 1986

FIELD SYSTEM: ES-2420 FOR IAT:SEGD12 TRACES PER INCH ENERGY SOURCE: BUFFALO GUN SAMPLE INTERVAL: .25 MSEC.

14 INCHES PER SECOND NO. OF SHOTS/HOLE: 3 RECORD LENGTH: .5 SEC.>OLARITY GROUP INTERVAL: 5 FT. FOLD: 6

NUMBER OF GROUPS: 24 SHOTPOINT INTERVAL: 10 FT.Y CONTROL GEOPHONE TYPE: GSC-200 GEOPHONE FREQUENCY: 8 HZ.

CGG ACCT. NO.: 4223101 GEPHONE ARRAY: 25 FT. GEOPHONE SPACING: 5 FT.

CHECKED BY: DR. T.L. DOBECKI RECORDING FILTERS: LC: 25 HZ SLOPE: 18 DB/OCT

MONROE B. SAVAGE NC: 720 HZ SLOPE: 18 DB/OCT

NOTCH: 60 HZ

----------- ----------------------

sp,CALE CH'4 CHI

1/64 MILI 240 125'

-/

I I I I III I

176

1

391 381 371 361 351 341 331 321 311 301

.ipp,

pop- 0 L I p 0 0 1 0 0 p I lop

ItPOP 104 Will 111ijilift 11=7 L It 00*bi"ob 0010"Woo " 0 0 0 0 1 -- I P OP 00 P)'.W4 "I" 1, 6 it It 0 0 oftommWarommuo

polo

p ]PPI".0.100 , "Op P"NW10100p 0 1 BPI WO1, Ilot p W 0 It 1010 No 0 10 0 k o plopbol." "o, loop; i t 0 pool., It", I pool.) pop j

loop,1 10,

) 1 1.0 -0 b 0 It spOiOsop b Ito

to PIP ONO 61rp kk! : , I I.

Nov. 0 volobb& 7t I It -plo" Do,Zi= - , : .r it "!P- I - "

" , ; .' p - 60f Wit1... 0-0.6.61bb-6

MIT 6'." t lo 0 'owlpop I'm P Wo I- I I W"WiPw

III, It, P orIPA I It 11 P-1, to -P -P -P W 0:

below $.)I Lipop, 09WlPI, It to" m-,-R

Mob 0 It P

Ib *-f itW -- ' , P I-,P It -410 so"I I. It 4 1 1 DoIp - I I 1-b-pow lb;-0 r i.4 D. Pt It "I. two &IIt 0 0 p N.J. 0 ONO. 'Pol'bill"ll"ImL., 10 pipat-lo.

otbb. obom-p-j". k 0 ? I I ;t " p k lopp-p 0 P IPWLjpjWP'Pn

IbW>lW' taw two,101011b 11,10, 'O'b" two I, rot W oop".0, pop Itop lot 01 1'Lio-P No 1. 10 op jo P pp w wapmo," P WP0. 10 111

1 0, 1 1 1"Ir ' , .

p, 10imp;0 .0 0. 1 Pilot -Pjo ; imp, "t 1b 0 pop lopp,

two' Posz , Ito Plopo now b-b I.; 1 8.1poke, I - 0 b 10, rp . 11 W* - :W 0otPWWPP0 low bo it, W 01

hL;4 :.' Wp t;p'-v-'Wj I-&I. op lot 10 1It I, I lop, -ibv -pOp 0 0 Pit It1.0 1 lotoMOP. It lbDo r lb" low*plop, Pp.

OJ11I i : I rWO b - 10 m4p Op, 00-1- bv- 10 b abo* Pp P Pop r

00 10 W W Plot 0,ro Wn wpb: .11.0job, .0-bo *lot r P &P, 1. mpoBobo DOI. P P.

P&6 L I 1. 1. 1 1 bo 1. P, . I P) 1, biwo 1. "l0jow PIP W" . k1. 100 --. P) pollWoo.

. -*.; ft I b P o."'Ilk ' 2-0 0 : It - "tIN, I '!b ;c - W4 .*No :4wpio I It lim

Spit so No Orr !p it W. PIMbow Itto b:1WP' ti, bow I. D.

wb ski, No. jo P p It PO" 10104!0 1040. 1 1So Doi iwo lop 11111 10 110 10,11, 101,

Jor I . , b , . M...loplo. - P I Mot ft 0 - P. r : ip : 0 V-*lot I lopip -o :P I jw ;. W UJ 1" SOrr go. Pok P ;o

mokopoo;Ir - b.10% ot IOa., b !I I b OWP'fi "10 " ' - r 111D. * ,r r r.. limp I,100 t 6,& 10), 1

IPo6wim. A* Ao- bI I ", ,

311 301 291 281 271 261 251 241 231 221 211

111101 -hh bli,00*10"7 ' ~? I~I~ 1 ~~0,00 "Im"-.-

po 10 1 6 ioi bP"WIfi0Pmp00O&

0 p j R ow mrP I:r O ka 11

b bs 10 01 . , P !p o

W. 0.0 oV I

-irv Wo,,bk

96

211 201 191 1ei 171 161 151 141 131 121

____ __

O. 100 '0 p b.t~ M-1000 0 16 IO NIA I

*.opboo . II I I 1" 1 b'0 0 . 0 1P

1. p 'p l ,I ; o o po ,& *mo I.~~bb I.~ I.F~ .0 Poop1 hN" to~, b p b * b . , ;r1 ,Wt .O

io I 0 * o "II

PLATElo I Contnued

641 631 621 611 601 591 581 571 56644 1 1 1

SEC. 0.0 1I I ! I{ii{Iiwo,, lot 00"""

Pool ~~.~plo moopp lo

bo~

)W 1b m oij 14 ?"'?

I. op p* ,~ Lur~mi~.~ ,'~' ~ho

pop~ PIPI

ip, -vo p 5.J'r1i'

Wo LL631bf- e ,mw OS 1

256

A 571 561 551, 541 531 521 511 Sol 491 481 471

INIP

bo~ I 0 ,: P

W)b -W b~mh I *,

bPj 1DP bob P

4. -; '' V -r.;!pr ;

:0' W Ir 1 POM 0 P tP

~~~~~~~: 10.1.~P? ii~~P.b~PP J14 10 "r I' 'IPj t 6 P 3'P~' jb .1

Iwo 1 1 .

lb 0. ,b

.6 -v:~p rA *0 1

W I PPI A~btP ppIr &&I~ Ip P -V 0 I*WO

,~~P fP~ W 1t;b0b1 W )

j w*_ 011p !

1471 461 451 441 431 421 411 401 391 381

op 6;, b- 00 1

tbI ~ ' 4 ,' I I )Jb I 01

polo

10 m-9~ , Ot

bbb kL~'' ~ IF:'I II

VVV

11121 91 81 71 61 51 41 31

I 6I0-0 I-v-bI

bI'bbj bIlp bI "0

gU~lI~uLIP*"

NO up I . P I ]h, ?ppr

-Pb. P *,

P m 4 Pb,. P Mrp ~ ~ ~ ~ ~ ~ J BooPPi3~~i~~ ~___ __

tIPA W,

job,~lot-

VIN ba 0!0bP~ 14Pb~'loo ~p 3PbIPJ I, D Pb' "rlo 0' popmjb~I~~.* Pb~p~bhpP,

16 FLAGS COMIPUT ING

111 STATIONS COMIPUTED AT CGG DATA PROCESSING CENTER

- FLOATING 1616 CHAIIPA ST.. DENVER COLORADO~ ~DATUM DATUMI PLANE: 1080 FT. CORRECTINA VELOCITY: 2800 FT.'SEC.

i LL~~L~L~1LLGEOfiASTER0.0 SEC.PROCESSI NG SEQUENCE

0O1-OEflUL T IPLEX

lorllotL~ 02-GAIN COMlPENSATION FOP TRANSMIISSION LOSS0 b NoAND SPHERICAL DIVERGENCE~s b * ~ *~ '~,03-COP SORT~ *1 04-GEOPHjONE EDITS

91 05-SHOT AND TRACE EDITS0.1 07-STATICS TO FLAT DATUMi

~p 08-SPIKING DECO4'.OLUT IONb 1 .p~'Im'bbOPERATOR LENGTH 40 FiSEC.~'i ~ 3 ~ ~ b~b~bP~UINDOU FROMi 62.5 TO 250 FSEC.

.~ ~ 09-AUTOIATIC RESIDUAL STATICS~"

t - 1O-VELOCI TY ANALYSISINWIP ftw, jo -0 l-NflO CORRECTION

13-AUTOtlATIC RESIDUAL STATICS~.** *14-COP STACKoml ISl"ol r".o;1-TIME VARIENT BAND PASS FILTER6IL" a. ,~ 0 o 0 1 72188-2401280 HZ. TO-TIOO IlSEC.00~**b;..I;~III

72188-1801220 HZ. T150-T500 MSEC.oor W pppr 1-DYNAMlIC TRACE EQUALIZATIONbo~ os 0 oP.P 17-STATICS TO FLOATING DATUMl

18-UA.E EQUATION MIGRATION* ~ *.19-BAHD PASS FILTER

0.3 20-DYNWfIC TRACE EQUALIZATION~Z 21-FILMi DISPLAY

~'~' DISPLAY PARAME TERS,in~ 3.I""~~~'HORIZONTAL SCALE: 12 TRACES PER INCH

Nip I wbr. o. PWVERTICAL SCALE: 14 INCHES PER SECOND0 II~"~ " POLARITY: FIELD POLARITY

.&j ~ abL0.4~h QUALITY CONTROLr~b ~DATE NOVEMIBER 196 CGG ACCT. NO.: 4223101~ ~ IPROCESSED BY: TANYA MUIELLER CHECKED BY: DR. T.L. OECKI

I t o o #tMO N R O E . S A L4A G E

SCALE

33 TR. = 1/4 MILE

PLATE 2

TINGCENTER

ONAL UELOCITY: 2800 FT./SEC.

STERSEOUENCE UATERUAYS EXPERIMENT STATION

COMPANY

fR TRANSIISSION LOSS'GENCE BEAVER DAM ARKANSAS

AKREA

,Um LINE C-I9 ON LINE'c.644STATICS64

STATIONS

I STATICS SOUTH NORTH

'ASS FILTER•TO-TIO0 rMSEC.T1o-T,50 SEC. WAVE EQUATION MIGRATION

.IZATIONDATUMDATIUM FIELD RECORDING

IZFION RECORDED By: CSI CREU DATEAUGUIT 1986FIELD SYSTEM: ES-2420 FORtAT:SEGOENERGY SOURCE: BUFFALO GUN SAMPLE INTERUAL: .25 MSEC.

ORMETERS NO. OF SHOTS/HOLE: 3 RECORD LENGTH: .5 SEC.TRACE TERS IGROUP INTERVAL: 5 FT. FOLO: 6

TRCES PER INCH NUMBER OF GROUPS: 24 SHOTPOINT INTERVAL: 10 FT.

GEOPH.NE TYPE: GSC-20 GEOPHONE FREOUENCY: 8 HZ.-WITY GEOPHONE ARRAY: 25 FT. GEOPHONE SPACING: 5 FT.

RECORDING FILTERS: LC: 25 HZ SLOPE: 18 DB'OCT

'CONTROL HC: 720 HZ SLOPE: 18 DB/OCT

tG ACCT. NO.: 4223101 NOTCH: 60 HZ

NECKED BY: DR. T.L. DOSECKIMONROE B. SAVAGE s

CH24 CHI

I I I --- 2 0' 125'

t'64 NILE

176

391 381 371 361 351 341 331 1 321 311 301

tIBJIIJ~ ~ ~ ~~~~~0 it It Or4LLIU I~} ft~~[~1III~t!-'Pop-

1+~'

~1M ,' *

iboob,

0; .m 4bht imp ',m 3

__________ ~?blop,

bwwJ 6 p

~ ~ IP vbhiII

.7 ~bb w 0 1

boo.. w"wor

io ipm o.o m, noa, - 0.-

J-S~

'a. _bD 1. 1Plp'mgor mo 0 .r l~ o b

281 271 261 51241 231 221 211 201 191 181

IY~ I oi o

1o I. tri I, 1. I

* Opp LL

~ bbbb~Not

loot I p!'

lo wrPo

;vor - -, -6L6

Pr ;; ,b 0 pn s ,

lo: op loo

Pilo.~ ~*P0' Z-f 6

96

211 201 191 18 t! 161 151 141 131 121 1

AM i WNI

)piht

§0 ib b14 I-P 0 p lg

1b~ 1I b t W r '

4-

1=I -or: L'I

.0. 0 610 0 Iv. Jll _ _ _ _li

g., 1110 toAp~ *~ IPpp* ~ I~ ;~

PLAT 2tCotnud

rZ-

641 631 621 611 601 591 581 51 56

6441 1 1

0S E C. 0.

J L OP P1-0 b. 1- 0b )*bhp.7P

0~WillIlkb

Not w"00" * 0 bo0 o P -oo *0.0 00 0 0 , "oj~bbb.

b w O-W) bO. W*O4 10,

ml ~P *pOpb1h14Al

IFI, bon 'Il

Op Ofj *"omlp

256

51150471 4 461 451

1 bo

~ ~ ~i ~ I , w :1 *

K.o I fO* I'-

wo'$Jm 1

0 1 * b P108I P h b ~)

r b~ b

off IWP p loo, r !

pipINAM 00,1pugj~~~~:bl 0 o)t i

bb lt*hrI Opp

potaIr..* r p~bp~

481 471 46a1 451 441 431 421 411 401 391 381

PIPI).- ,, ?Z

IP

b P- p I . p I ,

b~bP If N! I Mimi ~,g~h"*I

Lp *p j _'bl

t kvbbI

PLTI2(ocldd

104 FLAGS

5 1 4 13 1 2 1 S TA T ION S c

Itttlttttltthlbi ,IbttttL.UttttPDATUM 1

0.0 SEC.

~LjutE 00 .1

0.2

~ '" 0.5

0.0;

1b1 0.461 O

ap lb oIb .

-~blip

F

COMPUT I NG

D AT CGG DATA PROCESSING CENTER

,, rr ST.. OENVER COLORADOLAlE: 1060 FT. CORRECTIONAL VELOCITY: 280 FT./SEC.

GEOMASTERPROCESS ING SEQUENCE UATERUAR

01 -DEIULT IPLEX

02-GAIN COMPENSATION FOR TRANSMlISSION LOSSAND SPHERICAL DIVERGENCE

03-CDP SORT04-GEOPHONE EDITS

05-SHOT AND TRACE EDITS07-STATICS TO FLAT DATUI08-SP IK ING DECONVOLUT I ON

OPERATOR LENGTH 40 MSEC.UINDOU FROM 62.5 TO M50 nSEC. 28E

09-AIJTOM TIC RESIDUAL STATICS1O-VELOCITY ANALYSIS1I-NM CORRECTION1 2-flUTES13-AUTOtTIC RESIDUAL STATICS14-COP STACKIS-BAND PASS FILTER

7218-180220 HZ FINA16-DYNAMIC TRACE EQUALIZATION17-STATICS TO FLOATING DATUM

18-FILM DISPLAY

DISPLAY PARAMETERS RECORDED BY: CS

HORIZONTAL SCALE: 12 TRACES PER INCH FIELD SYSTEM: E'

UERTICAL SCALE: 14 INCHES PER SECOND ENERGY SOURCE: IPOLARITY: FIELD POLARITY NO. OF HITS/HOLI

24 SINGLE GEOPH

QUALITY CONTROL GEOPHoNE TYPE:GEOPHNE SPC I N

IER 1986 CGG ACCT. NO.: 4223101 RECORDING FILTEiy: TANYA MUELLER CHECKED BY: OR. T.L. DOBECKI

MONROE B. SAVAGE

SC LE CH24

33 YR. = 1/64 MIlLE 240'

PLATE 3

WATERWAYS EXPERIMENT STATIONCOMPANY

-a BEAVER DAM ARKANSASAREA

Ll LINE C-3LINE

288STATIONS

SOUTH NORTH

Ec FINAL EQUALIZED STACK

FIE, FIELD RECORDING

?EU RECORDOE BY: CSI CREU DATE:AUGUST 1986

120 FIELD SYSTEM: ES-2420 FORMAT:SEGDR ENERGY SOURCE: AMER SAMPLE INTERVAL: .25 MSEC

9O. OF HITS/HOLE: 9 RECORD LENGTH: .5 SEC.

24 SINGLE GEOPHONES FOLD: 6

STGEOPHONE TYPE: CGS GEOPHONE FREQUENCY: 100 HZ5 FT. GEOPHOE SPACING: 5 FT. SHOTPOINT INTERVAL: 10 FT.LC: RECORDING FILTERS: LC: 25 HZ SLOPE: 18 DB/OCT

C: I, HC: 720 HZ SLOPE: 18 DB/OCT

NOTC - NOTCH: 60 HZ

CH24 CHI

240' 125,

I .. . ..... .. . . - - - - - •III

314

271 261 251 241 231 221 211 201

SEC. 0.0 It Hht[tt::tP# 6~, bk1i.

1 1 l o o p. . k"osl 0,

0. OF bk

j'p' jobil'.pk jo o

~W b- Mokt

4v io

wI .Ib3 J .IIJI ' * Wo

0.4 ~ frb.F~rV-~IUk~)lop p ~ ~le

F . e

244

119le 171 161 151 141 131 121 11101 9

I~ ~~ l ioi iI

0 1 It

DI~ .P~ - )i

14

244

131 1 111 101 9181 71 61 51 411 1 I

ti v ilPooi 1p1-1p. 1, 0i -bi S 'o6

I o loti

ii ,,

- m ,6, N% M b& .. h h m- oA

Paw MI " a

~IMP I

0P Io) 9* i wr An_

I Lep oL

9.. 1,w 0

lk to p

-tt

"-h.I 1 -~pop ,0- ',

PLAT ItCocuddlow 0,

184 FLAGS

514 121 11 1 SrAiii ION

.1 - FLOATING31 DATUMi1~ii 1LLL.0.0 SEC.

-or .2on P ;* S

L ~0.31

0 7 11r

__ ~ bb~1h~b 0.42D

lowR'~'Ir'~l5polo blm

~COMPUT ING

~COMPUTED AiT CGG DATAl PROCESSING CENTER

JTED 1616 CHArMPA ST., DENVER COLORADO1CH .._.DATUM1 PLANE: 1080 FT. CORRECTiONAA VELOCITY: 2800 FT./SEC.

GEOMASTER

PROCESSING SEQUENCE

01-DEMULTIPLEX02-GAIN COMPENSATION FOR TRANSMISSION LOSS

aT AND SPHERICAL DIVERGENCE

03-CDP SORT04-GEOPHONE EDITS05-SHOT AND TRACE EDITS

01 07-STATICS TO FLAT DATUM

Oi 08-SPIKING DECONUOLUT ION

OPERATOR LENGTH 40 MSEC.UINOOU FROM 62.5 TO 250 MSEC.

09-AUTOMATIC RESIDUAL STATICS

10-VELOCITY ANALYSIS

I1-NMO CORRECTION

12-MUTES

13-AUTOMATIC RESIDUAL STATICS

14-CDP STACKI, 15-BAND PASS FILTER

72188-1801220 HZ

16-DYNAMIC TRACE EQUALIZATION

17-STATICS TO FLOATING DATUMI 18-UAVE EQUATION MIGRATION

19-BAND PASS FILTER

20-DYNAMIC TRACE EQUALIZATION

21-FILM DISPLAY2

DISPLAY PARAMETERSHORIZONTAL SCALE: 12 TRACES PER INC;iVERTICAL SCALE: 14 INCHES PER SECOND

POLARITY: FIELD POLARITY

QUALITY CONTROLDATE NOUEMBER 1986 CGG ACCT. NO.: 4223101

OVEIUEl; PROCESSED BY: TANYA MUELLER CHECKED BY: DR. T.L. DOBECKISED BY' MONROE B. SAUAGE

SCALE

33 TR. = 1/64 MILE

j ,eliu . m~mlll~lllllIl••Illlllmm a

PLATE 4

UATERUAYS EXPERIMENT STATIONCOIPANY

BEAUER DAM ARKANSASAREA

LINE C-3L INE

288 1STAT IONS

SOUTH NORTH

WAUE EQUATION MIGRATION

FIELD RECORDING

RECORDED BY: CSI CREU DATE:AUGUST 1986

FIELD SYSTEM: ES-2420 FORMAT:SEGD

ENERGY SOURCE: HAMER SAMPLE INTERVAL: .25 MSEC

NO. OF HITS/HOLE: 9 RECORD LENGTH: .5 SEC.

24 SINGLE GEOPHONES FOLD: 6

GEOPHONE TYPE: CGS GEOPHONE FREQUENCY: 100 HZ.

GEOPHONE SPACING: 5 FT. SHOTPOINT INTERVAL: 10 FT.

RECOROING FILTERS: LC. 25 HZ SLOPE: 18 OB/OCTC: 720 HZ SLOPE: 18 D8/OCT

NOTCH: S0 HZ

sP

CHZ4 CHI

Z40' 125'

314(271 261 251 241 231 221 211 201276 1

S E C. 0. LIL.01p771:71 ~ ~oo pto1 F1 b[- Pwo

ob" popwo -

'I mt m~b,. mmb.b ION,4 h m b n m ~ ,~.

Pop,, Ob p

0.1 W b 0 ,'I 1. b W

IN-4 i. - W a -

-c lplpihrt 0, fpp Im -pi 1pb AW b o I

0.4 Or 0aat:11' lop, jk: 4 1op 0i m,%W

I. .6Ib0 lk r1 1~ ~ . -I r m-w ' 1 "0 lop~

244

16-. -~P . - _ .' 0",w.pp1

h. W .:i P u " 0 .. -it~ 7 N 0foo'' IP.0-

Pooli. 'D 'Nobb0 W.PP " Ito 00.0It w~. a Oo l o-

Moo~ 0.

00. - j -A

ti ~ I 6

0 , poo" Z3.b , P0pul N'b It L .'~ 'I Itp wS~L .II ~ Isi b l

0, b . -0 1 P- . t w, b ml, I , , Ibb si p

Pi * P, 1, P ,p 1, 1, 0 11 o. o- 0 P: ~ ~ * I ~ I ~ I*

244

K41 131 111 ltl 91 81 71 1 51 41

P 0 Do P bo 0 o, rKo oo01 .0,1 o01001o Kb01 I IpprK P P0

wo 'SP 00-1 OroHI01o&"P

ob b p 0, P

P,

two P 0 PP L

AV '; ; ; r~. ~ " L " .0 0'.' i " -, NI' ." -* I.

I-. POO(r lC u '

APPENDIX A: GLOSSARY*

Amplitude spectrum - amplitude-versus-frequency relationship such as computedin a Fourier analysis.

Autocorrelation - correlation of a waveform with itself... is equivalent topassing a waveform through its matched filter.

Common depth point (CDP) - having the same midpoint between souice anddetector.

CDP gather - the set of traces which have a common midpoint (see Figure Al).

CDP stack - a sum of the traces corresponding to the same midpoint.. .cor-rected for statics and normal moveout (NMO) and then summed (stacked).

CDP method of data acquisition - shooting seismic data in a manner such as togenerate CDP gathers (see Figure A2).

Common offset gather - a side-by-side display of traces which have the same

shot-to-geophone distance (offset).

Common offset method of data acquisition - see Figure A2.

Deconvolution - a process designed to restore a waveform to the form it hadbefore it underwent a linear filtering action (convolution).

Demultiplex - to separate the individual component channels which have beenmultiplexed (mixed together).

Elevation statics - see "statics."

F-K velocity filter - discrimination on the basis of apparent velocity.

Floating datum - a variable reference surface used in areas of variable

topography.

Fold - the multiplicity of common depth point data.

Gain compensation for transmission loss and spherical divergence - gainapplied to a recorded seismic trace to account for decrease in wavestrength due to energy absorption by the medium and due to geometricspreading of the wavefront.

Graben - a down-dropped block bounded by normal faults.

* Taken from Sheriff, R. E., 1984. Encyclopedic )ictionary of Exploration

Geophysics: Society Exploration Geophysicists, Tulsa, OK.

Al

Migration - rearrangement of interpreted seismic data so that reflections anddiffractions are plotted at the locations of the reflectors and diffract-ing points rather than with respect to observation points.

Mute - to change the relative contribution of components of a stack withrecord time.

Normal moveout (NMO) - the variation of reflection arrival time because ofshotpoint-to-geophone distance (offset).

Operator length - the time-domain length of the impulse response of a con-volution operator.

Optimum offset - the best shotpoint-to-geophone distance for observing aparticular reflection in a noise-free window.

Sort - rearrange seismic traces with respect to some attribute shared by each.

Spiking deconvolution - deconvolution in which the desired wavelet is a spikeor impulse.

Stack - a composite record made by combining traces from different records.

Statics - corrections applied to seismic data to compensate for the effects ofvariation in elevation, weathering thickness, weathering velocity, orreference to a datum.

Throw - the vertical component of separation of a rock unit (bed) by a fault.

Time-variant filtering - varying the frequency band-pass with record time.

Vertical seismic profile (VSP) - measurements of the response of a geophone atvarious depths in a borehole to shots on the surface.

Vugs - large pores in a rock mass from a few millimeters to a few centimetersin size; often caused by dissolution of carbonate minerals.

A2

Geophone group

Spread of 24-group locations

A ; B 5 C 7 D 9 E 1' F 13 15 17 19 21 23

Shotpoint

Reflector

Figure Al. Illustration of a set of seismic

raypaths sharing a common midpoint or common

depth point

A3

1 Rcords 10 Traces in The Ronge Shon

I x. I I

? .3 4 5 6 7- 8 9 Ito0--- -- eo s-•o

-50

SP 0SP f-*---.... C The*e Dots Represent TheSP 2 o .- .9-4 a Horicontol Posttions of TheSP 3 ** o*o* o -RefleCtirg Poits ForSP 4 Horionfol Horizons of AnySP 5 Depth.SP 6

R1/S5

S1 S2 S3 S4 R2R3R4R5

GROUND SURFACE

REFLECTOR

Figure A2. Illustrations of the procedures involved inCDP data acquisition (upper) and comnon offset data

acquisition (lower)

A4