Exh

ibit

INT3

57

June

26,

201

2

Uni

ted

Stat

es N

ucle

ar R

egul

ator

y C

omm

issi

on O

ffici

al H

earin

g Ex

hibi

t

In th

e M

atte

r of:

Pro

gres

s E

nerg

y Fl

orid

a, In

c.

(Lev

y C

ount

y N

ucle

ar P

ower

Pla

nt, U

nits

1 a

nd 2

)

ASL

BP

#:09

-879

-04-

CO

L-B

D01

D

ocke

t #:0

5200

029

| 052

0003

0 Ex

hibi

t #:

Iden

tifie

d:

Adm

itted

: W

ithdr

awn:

R

ejec

ted:

St

ricke

n:

Oth

er: IN

T357

-00-

BD

0110

/31/

2012

10/3

1/20

12

Uni

ted

Stat

es N

ucle

ar R

egul

ator

y C

omm

issi

on O

ffici

al H

earin

g Ex

hibi

t

In th

e M

atte

r of:

Pro

gres

s E

nerg

y Fl

orid

a, In

c.

(Lev

y C

ount

y N

ucle

ar P

ower

Pla

nt, U

nits

1 a

nd 2

)

ASL

BP

#:09

-879

-04-

CO

L-B

D01

D

ocke

t #:0

5200

029

| 052

0003

0 Ex

hibi

t #:

Iden

tifie

d:

Adm

itted

: W

ithdr

awn:

R

ejec

ted:

St

ricke

n:

Oth

er: IN

T357

-00-

BD

0110

/31/

2012

12/3

/201

2

Environmental Karst

dolines, uvalas, semi-blinn and blind valleys, sinking streams, and sp:rings.. C)osed 'depressions have developed by subsitle.nce and/or suffosion of residuum into cavities in the' underlying Ocala limestone.

Subsidence sinkhol€'s develop throug.h the progressive collapse of arches or domes which span air-filled voids in the surface residuum. h void in the Ocala limestone, roofed by residuum, moves. towards the surface by the cot lapse! of roof material until it finally breaks through to the surface to form a sinkhole. Collapse of the roof temporarily produces a cylindrical hole which .. .rapidly weathers into a gentler conical or bowl-s\~ped depression. suffosion sinkholes are 'funnEl-sh~)ed depressions, which develop in residuum through snasmodic minor subsidences and more continuous piping of llnconsolid3.ted mat.erials into widened joints and solution pipes in the ocala limestone beneath.

As a result of sev~re droughts during the 1954 an~ 1977 growing seasons, agriculture in the Dougherty Plain has become increasingly dependent upon groun{lwa ter from the ocala aquifer for irrigation. In 1970 less than 8 million m3 of water wer~ withdrawn for irriga tion, in 1977 more ~ han 150 mi'llion m3 were withdrawn. Almost unhf!ard of prior to 1970, center Fivot irrigation systems increased to 316 in 1976 and to more than 1,000 in 1979 (Pollard et a1., 1979; Kunaell, 1980).

Incceased use of the ocala aquifer for irrigation could,. in the lon'g term, result in a lowering of the regional piezometric surface. In the short term cones of depression will be developed around center pivot and ether irrigation wells. In either case one possible cause for c'oncern could be accelera ted development of subsidence sinkholes in the region. ~t cones of depression, surface residuum loses hydrostatic support and steep hydraulic gradi€nts are produced. .Increased groundwater flow velocities cause erosion of subsurface residuum into bedrock cavities.. Natur-al and artificial recharge (in the form of irrigation water) percolates freely through the newly established vadose zone eroding the roofs of air-filled voids.

In Alabama an estimate~ 4,000 man-induced sinkholes a['e thought to have formed since t'900, most o.f them inc1uced by a decline in ·the vater table due to groundwater withdrawals (Newton, 1977).. In some cases man-induced sinkholes near discharging wells have resulted in groun1lvate'r contamination (Spiqner and Graves, 1977) _ A further problem is that some sink.hole subsidences arc ca tas·trophic. On the Far west Band, South Africa, on December 12, 1962, a tbree-storyed crusher pla.nt belonging ·to the west nriefontein Gold

92

~'

i'~:<

,<;'

"1:1

~i

~,Ii

iii. I I'

1,11 1, ~ t

<.J.in

.... !i

~.

:i1. ,

, ,!

,. ,~

"

.;

Brook & Allison: Fracture Mapping/Subsidence

Mine was engulfed by a huge sinkhole.. The disaster took 29 lives. On Augtlst 3, 1964, another sinkhole formed sudnenly in the same general area swallowing a house at the rnyvoor'litzicht Mine, 5 per-sons were killed (J~nnings, 1966). Accelera tell sinkhole develorment in the Far- t~est Rand is in respon'se to a lower water table.. Pumping of water from the gold mines has produced cones of 1eFression 15-305 m below the for-mer piezometric surface. In acldition, many sinkhole collapses were trigger~d by large volumes of artificial recharge when water was (lischarged at the surface by the gold mines in the course of underqrol1nd fumping (BLink and Par triage, 1965).

Because of the threat of accelerated sinkhole development on the Dougl1erty Plain, as a result of increased irrigation, there is a need for ground subsidence susceptibility maps, which can be used by land use and water,resource plannersa An attempt has heen made to develop such maps in a sample area--Dougherty Coanty--using easily acquired sinkhole and bedrock fracture nata.

'rHE STUDY AREA: DOUGHERTY COUNTY

Dougherty County is approximately 43 km E-W, 21 km N-S, and covers SQS km2.. It is underlain by Cretaceous to Recent sedimentary rocks,lJhidh dip souteastwards at 2 m/kma These rest on crystalline basement rocks and older paleozoic sedime.nts ... \·l. Only the extreme souteastern corner of the county lies outside the Dougherty plain topographic province.. This area is a part of the 'rifton Upliind ana is capped by clays of the Miocene Hawthorne Formation. It is separated from the rest of the county by the ?elham Escarpment. Elevations on the Tifton Upland reach 100 m , on the Dougherty Plain they range from 50-75 m (Fig. 1).

The topography of the tIpper surface of the Ocala limestone in Dougherty County is highly irregUlar because it has been diffecentially weatbered... The limestone. may be less than 15 m thick in the vest, where it occ~sionally outcrops, but increa5es to more than 75 m in the east. The Ocala is covered almost everywhere by surface residuum averaging 13 m in thickness in the no.rthwest. and 19 m in the southeast. Besiduum thickness is hiqhly variable and may increase by more than 30 m over a distance of less t~an 3 km (Wilson and Pickering, 1973). The residuum is sandy, silty clay and contains boulders of weathe-red siliceous limestone up t.O 2 m in diameter; it is thought to be primarily derived from tbe weathering of the ocala limestone~

The Flint River and other surface streams in the county occupy channels t~at cut into the Ocala aquifer.

93

x;nvtronmental Karst

For most of the y~ar these are effluent strp.ams but at tides of peak flow they may become influent. The Ocala aquifer is recharged through sinkhole~ and blind valleys. Sinkholes vary from 6-8 m d~ep and frow 150-300 m in diamet'er. Many are alluviated, some containing perched water bodies. other depressions with open ponors are sites of rapid recharge and therefore sites of potentially rapid groundwater pollution.

In recent years man-induced . ground subsidences have became more freguent in Dougherty Gaunty" particularly in the count'y seat Albany.. In one incident, Hilsman Park, located in ' a sinkhole, vas selected by lltany city officials as an ideal location for a recreational laKe. Clay vas hauled in to make an impervious floor and logs and tree stumps were placed in the center of the sinkhole in an attempt to plug it. A well vas drilled immediately adjacent to the area and water was pumpe~ into the depression for several days forming a small lake. The lake lasted only a short time, the vater e~ntually drained underground when part of the depression lfloor, including filling material, logs, and tree stumps, subsided into a subsurface cavity (Hait, 1q63). A second subsidence occurred near ·Sanks Halley Art Gallery in Albany. During heavy rains, storm runoff is funnelled into a sinkhole on the grounds of the gallery and is then pUliped out through sewer lines. On June 6, 197.3, during an especially heavy rain, onE of the pU~p5 failed and 2-3 m of storm water vas ponded in the deFression. This water triggered a subsidencp. only 15 m from tbe art gallety (Wilson and Pickering, 1973).

KODELLING PBOCEDUFE

The relative susceptibility of an area in Dougherty county to ground subsiaence was considered to depend on ,the numcer of subsurface cavities in the OCala limestone, and on the liklihood of subsidence or suffosioD of residuum into theM. Ogden and Reger (1977) concluded from stUrlies in Monroe County, West Virginia, that areas underlain by the most cavernous rock display the most dolines. They found that the percent of the limestone area in dolines and the doline density were useful indicators of areas of potential suhsidence. Pord (1964) has ~emonstra ted that in th" central Kendip Hills of En~land the formation of one doline (the "mother") ten'tis to promote subsurface conditions that r are conducive to the formation of additional dolines (the "daughters") in the same area. Data on sinkhole density and on the ~ercent of area in si.nkboles were ther~fore used in modelling as being indicative of both the ",umber of cavitip.s in limestone and of the liklihood o~ further subsidence or suffosion of residuum occurring. In addition, as there is

94

, /,

I I I !

i I \ ! I I \

.l \

! ,l

'Brook & Allison: Fracture Mapping/Subsidence

potential development of solution voids along zones of high seccndary permeability because these concentrate qro'Und ."ater 'flo!ol, data on fracture d~nsity, fracture intersection density, and th~ total length of fractures in an area were also \lsed in modelling the presence of solution cavities in the limestone.

A geographiC inforoation system DBMANG/CONGRTD was used in !:inkhole anll bedrocr.. f!'a,~ture aata analy.sis (Bol<ans, 1971). The prograro DEMANG (Data Base Manaqer Grid} was used to Quild and maintain a grid-format data base. The program acco~odates up to 30 variables and any number of cells. CONGFID (conversational Grid) was used to d isp la y gr id-foT:ma t da ta in gre y-s~ale choropleth map form via a 'line print.er.. It is presently dimensionerl for 29 variables a .nd a maximum o'f 20,0 00 cells.

CONGRID has four ~3.p output options: (1) simple variable display (a data file map), (2) intersections of variables, (3) unions of variables, and (14) linear combinations of variables.. Cells in the grid-format data base ~ith 3-5 sinkholes form a set, cetls with 6-8 fractures form another set. The intersection of these tvo sets (map option 2) includes all cells with 3-5 sinkboles AND 6-8 fractures. The union of these tvo sets (map option 3) includes cells with J-5 sinkholes on 6-8 fractures. The linf!ar combinatio'n option is used when weighting of variables and values is nee~ed in data analysis. For example, if the decline in the gronnd water table is considered to be twice as important in triggering sinl<hole collapse as the depth of surface residuum, these two variahles can be ~eighted 10 and 5 respectively i~ ~odelling ground subs idence susceptlbili ty. Map options (1), (2), and (~) were used in t .his study.

In order to ,levelope sinkhole and b~drock. data files in OBMAHG, oougherty county was partitioned into 855 cells in 19 'rows and 45 columns. C~ll size vas 1.0

X 1. 1 km.

SINKHOLE ~ND F~~CtOFE DATA COLLECTIO~

The n.s. Geological survey 7.5 minut" quadrangles shoW approximately 40% of the sinkholes in Dougberty County, the remainller are_ too shallow to be dep icted on these roa~s which have only a 10 foot (3.05 ml contoU"t: interval. Sinkholes were thereforo mappea from February, 1973, 1~24,OOO scale, color infrared imaq~s (NASA Proje ct 1473). Color infrared transparencies ~erp. viewed stereoscopically at 2X magnification. Sinkholes were identified by tbe presence of surface .ater bodies, from ve!:Jetation - ano soil moisture patterns, and from topoqraphic expression. Sinkhole boundaries were drawn at the breilK of slope with the

95

Environmental Karst

OOU13HI!.,ny _COUNTY

PHYSIOGRAPHIC SUB-PROVINCES

[=:J Dougherty Plain

~TI1ton Upland

,t N

Mile s

0 ' 3

o ~ Kilometers

Fig. I. The physlc.l and bltilt enyj l'O:ll:l!!nl of 00~9l1erty f.ounty. Georgi a .

.r-._.]

n:.,...-,'_ . .,..._. J' 1· . u ...... !~~ .:~ ' ~ \:_._.,-,.~_._. .--.. . t . ". ... ' • Q t v- .. ·;-.-c-·-·--·-·-) ~r1 ... ' '.; . I:l~ t;7 ~:".. ,~ ... "\.. > ._.-" .,c._.

I . .. ( ••. , .. ".~ _. . ... . • ., ..• : :. 1 ~ .~. \"''O~''.'''. ~ '~~'.Q "~"' I tJ .. ~ •• '~ .1 • • : • • 0 • '" • •• • , .••..•... _.,... o~ . '\:.I. ...... 0 ~.~ •• ~: ••• ,. '"" •• : ~ .1 . ~~" ' • • :0';' . "\>;A CI~ .'," , '" ........ ~ .. ' D,i ••••• \(I I·

f ... ~ _....) ""_.',, 'Jo.~ •• ' ~,."j, ?.h 9 ~ ci n. '- :;.<~. ,_.,:;. ,~ . •• '. ....,. . """,'" ~"(/' i \ .,.,1l' •• ;. ,,';'1,1'.;,' .," .~' ~o/P lo "" " c>'" .~ • .1', i \ ~.::': :,: .. ,_ ~., ;., . ".' 17;'" AI....., ~ '" "'. , • • O. .-'

• ''I> .,~.... ", '0" D t:?. 'O.JI.I'!l () ... • ..... . " 1 \ .. " . ••• /0.. , ,1 • "', "'- .... " ." ,.,.. .. ,",', ' <;:0 ." I. N1tA'oj'(/O".!. .. ..~_o to 6 " ~ I· • .. ,_ .'l.S', ~ CSJ II ()~.. • 00" ~O'

\

• .. ~ • .: •• " , "0 , Q • Q ... "f •• 0 '111 l) '"". .~ • .,...~ _ Q, .B \l • ~. I:) ~ • • )~" \:~_ 1'. . '.; I ... tl'o· .. , .,(,f?c "~.",.,, ;"::.f'~."'. " -',! ~ . " ., , .. ' ... 'U. ,' .. ,.!)!' _ ..... ,,:" .. J \ • g(' •• ~' • '_.;. \ • , ,, <'>:, • ' • ",0 : •..• .

. "'''~ " .:. n. ~V' "q, .. ~.~.. '.. 1 . .,) ;0,4 ,,·,r-" -0::"" . r. ~.. ~ -"8' 0 a ',,· ~ . L._._~::r....' k.~. 'Q'. ';' , .. • ' .• ' • .J

. . __ . ___ •. ......!..-J.d2 . ..r.~r...!-·._~-: _ . ..!....'--!..._._._ . ..:....J Mil ••

~ ~ o •

Kllom"t"r" N

Fig. 2. The sinkholes of Dougherty County. Geor'gia .

96

1.J ~I .~,

i lOr \~

L':_ -~'.

~v

> • .;'

Brook & Allison : Fracture MappIng/SubsIdence

Eu.rrounu·ing f 1'1 t tl1r t'a iG . . Plan im~t[ic control . was estal:lished by als~ lJIapp5.ng roads and. railway lines .. "Pbotographic distortion was relOove.:J. ani! sinkhole boundaries wer e traD~{erred to 1:24, 000 sca le topographic maps 11sin9 a Bausch and Lom b zoom tr3n sfe r scope~ In total 1,011 sinknoles werp. mapped, the mean density for the county lleing 1_1 sinkholes/kll 2 (.Fi g ure 2). DRMANG data files we.e developed for the nu~ber of s1nkholes in each c~ll ( ~ ickhales lying in mar@. than one cell were counted in each cell they occupied), and for the percent ~area of ·each : cell occupied hy sinkholes. The maximum Dumbe"I:' of sinkholes in any cell was 15, 32 cells contained more than 10 sinkholes (Figure 3). Five cells han more than JOX of their area covered by ~inkholes. 26 , cells had m?re than 20'" covered.

The ' rela·tionship between the number of sinkholes in a cell and the area occupied by them provides an insight into the evolution of_ sinkhole topography in Dougherty County (FiquI'e 4). It sug'lests that there are two .distinct stages in sinkhole development. Tn the first stage there is a qradual increase in the number of sinkholes in a cell and in the area they occupy_ ahen the number of sinkholes reaches lij-1S, a threshold is reached beyond which the topography enters a second stage of evolution... In this' stage the lateral grovth of sinkholes and their coalescence to form uvalas becomes ~ore important than the formation of new sink.holes. The area of the cell occupied by sinkholes continues to increase but the number of separate depression s decreases •

i\ most important char::lcteristic of the s inkholes in Dougherty County is that they have pI'onounced linear shape elements.. To· test whether these show statist.ically significant. pr-eferred orientations, the azimuths of prominent long axes or other linear shape elements in 205 Sinkholes , in randomly selected cells , were m~asured and grouped into l O-degree classes (Figure 5).. Chi-sguared analysis vas used to test for non-randomness in the distribution (Pincus, 1953).. Six classes, their midpoints i\t. 325 , 315, 305 , 5 , 25 , and 35'., were found to be ' significa ntly non-random at the 0 .. 05 level. wh~n adjace-nt significant =lasses were grouped, thr~c prpferred orientations emerg~d at 315, 5.aI!rl30'.

Wilson (Doug Hilson, personal communication) .reForts joint directions at 300, 10, an" 30' in ocala limestone -visible ~long the banks of thO' Plint River in Dougberty county dueing periods of low flow. The close agreement bp.t.ween !I1easuc·ed ioint directions and the orientations of linefir sinkhole shape elements suggests that sinkholes in OoulJherty Count.y have developed above, a nel are p.lonljated parallel to fract.ur es in the

97

·'2 -'5 111·-11

"'[.%9 4 ~ 7

00-3 0

0

~ N

Mil ••

KIIQlT'ot •• ",

Fig. 3. Hllllber of sinkholes by cell. Dougherty County. Georgia .

14-15

.J 12-13

..J .., o ~ Ii-IO

'" .., ..J 8· o . - 9 :J: >< Z u; 6- 7 "-o <to .., 4- 5 00 ::Ii :) Z

0- .,. ,

0

.,. en .,. en .,. en C\J C\J '" , , , , , . , ." 0 " 0 " 0 C\J C\J '" SINKHOLE AREA BY CELL (PERCENT)

" '" A

Fig. 4. Relationship between the numbtr of sfnkholes in /I cell and the area

'of t~e cell occupied by sinkholes. Dougherty. County, Georgfa.

98

3

5

l. i1" '?

Brook & Allison: Fracture Mapping/Subsidence

underlying ocala limestonp.. As the sinkhol~s have f01:med hy subsidence-'or suffosion of surface residuum into subsurface cavities, this implies that joints and faults are also the major avenues of grounn water movement and solution. ' ,- '.

~RAC!URE DATA

The distrihutio'n and shapes of sinkholes in Dougherty count:y 'Were .use'd to map possible fractures in the Oea la l"imestone. ' Mapping \~a 's completed in three stages. In the first sta'ge,all pz::onounced sinkhole long axes and other l',inear shapp. eleme~t,s were identifieQ aDd marked. In the seconli stage, lil)ear shape elements were connected "where :tlicse appe,a,red 'to lie along a single fractu,re. in addition, fractur,es were 'drawn where several sinkholes fell 1\lon'g ' a straight line.. In ths final stage of :mapping the color infra_red images of the 'county were examined for additional evidence of fractures in the underiying belfrock ' anu for evidence which might suggest modifications to the fracture map prepared from sinkhole data. In total, 1,298 possible fractures were mapped, the mean length being 1.9 ltm/k.2 (Figllre 6). DBMANG data files were developed for the number of fractures, the number of fracture intersections, and the total length of fractures in .each cell. Thirty cells had more t .han 9 fI"acturp.s and 276 cells more than 5 f~actu~es; 155 of the 855 cells had no fractures.. 'th.ree cells had morP. than '15 fracture intersections and 20 cells more than 4.5 km of fractures... (Fra'ctures lying in more than one cell weX:f! connted in each cell they occupied.)

In an attempt to expla.Ln the fracture pat.tern in Dougherty County, fracture end point coordinates were digitized and lengths and orien·tations calculated .. Fractures \lere then grouperl in 10-degree classes and the number and total length of fractures in each class estimated (p'igure 7). C.hi-square analysis vas userl- to test the non-.ranaomness of these distributions (Pincus, 1953). In both data sets ·the same six classes, their midpoints . at 31St>, 32511-" 335 00

, 5°, 3511-, and 40°, were significantly different f~om random at the 0.05 level. When ' adjacent classes were grouped three major preferred fracture orientations elllergeil at 325·, sc:;· , and 40'·.

Preferred fracture orientations in Dougherty County agree well with th03e measured. in nearby areas o£ Florida and Georgia.. Vernon (1 <:151) recognizes a fundamental regional pattern of two systems of fractures trending NW-SE and ~lE-SW in northern "Florida and Farts of southwE'st 'Geprqia. Wor.k by Ellwooo: in Climax Blowing and Glory Hole Caves beneath the Pelham Escarpment, near Cairo, Georgia, 75 km south of Albany, has revealed two preferred passage orientations at 317~

99

~5011 --------------------,--------------------

~40

9 ILl ..J 30 o :z: :I<: Z in 20 lL. o a:: ILl III ~ ;:) Z

o· 360'

LONG AXIS ORIENTATION

Fig . 5. IlUflber of sinkhole linear shape eleclenl s by lO~ orientation c llss

for a randofl s.tlP! " of 20S sinkholes . Dougherty County, Geor'9f1 .

o 3

o • K,lomet.r . N

F19· 6 . Possible f r ac tures tn the Oula l1r'lestone /\lapped fo r si nkho le data ,

Ooughp.rly County. Georgia.

)\1(,

90'

1-~':"

~~t

-

A

B

6 PERCENT

PERCENT

NUMBER

Fig . 7. Number (A) and length (B) of mapped f ractures by 10' orientation class, Dougherty County, Georgia.

Ju le t Values o( VUhbln Spectfttd for Jntersedlons 1-4

SPlCIFI(D VAlU£S OF VARiABlES YAA IAaI.[

IKTER5£CTIOH IIIfT(II5[( TI0II IMT£RS[CTIOI UITUS[CTIOM 1 , ) •

Ihlllbf!1" of Sl nktlolu 12~15 1~ 1 5 4- 1S 2-15

Percent Aru of Ce11 Co~ • .-.d by 1s..24 10..29 0-)5 0-)5

Sinkhole,

Nt,JIber of fr"ct~ru J.' 5-' 3-' 1-'

MUfIlber of fnctu,. l nt erncl tonl \2·'9 .. " 4· 19 2-19

le~tf\ of Fractuns (kill) l , ' -1. 2: 2.8-1.2 1. ,..7.2 1.0· '.2

101

and qO-. Fault planes have been identified in Climax Ca ve a t both orien ta tions suggesting t.hat passages have formed along a conjugate set of shear faults.. Ell~oo~ has tentatively inte.rpre.~ed the fracture directions in terms of a stres·s dist:ribution \lith . the axes of greatest and least. stress: ~orizontal . and oriented at approximately 90 11 and 361)~ · respectively, and the axis o·f intermediate stress vertical. · .

The fracture set.s in 'Dougherty County \lere most likely produced by ·a · maximum s·tress E.rom the west. acting against the Chattahoochee ·A.nticline, trending 350

0 in haseme.nt rocks vest of Oo.ugherty County. By this interpretation, f.ractures a.;t' ··:t5·~ are extensio.n fractures; those at 3251> and 40!o>, .a<..:p·on·;ugate set of shear fractu ·res. The broad range of the 325~ and 40 G

fracture directions suggests that they may have been produced by a residual s·tress field in basement rocks, which ca used up~a·rd migr?-:tion of structural !eatures and their impressment on the younger sediments ..

SUBSIDENCE SUSCE~TIBILITY ' MODELS :

The sinkho.le: aD ~d fracture data files in DBMANG were used to . model . via . .. CCNGB.ID the relative susceptibility of cells in Douqnerty County to qround subsidence. separate ~ · morlels were produced by intersections· and by linear combination of the five variables. The susceptibility of a cell was assumed to increase with and increase in all variables except sinkhole area. For this variable, susceptibility is assuned to reach a ~aximum when 15-2q% of the c~ll is occupied by sinkholes. "his assumption is based on the observation that when 20~ of tbe cell area is occupied by sinkholes, fu.rther development is dominated by lateral growth and coalescence of sinkholes rather than by the development of new sinkholes.

INTERSECTION

Intersection modelling of susceptibility involved the use of CONGRID to . identify and rna p cells wi th specified values of the five variables~ Four intersections were mapped. A broader range of values for each of the five variables vas specified for successive intersections (Table 1). This m~ant that each intersection identified cells included in the previous intersection pIns a number of additional cells. These add.ttional cells were considered to be less susceptible to sinkhole development than the cells already identified. In the first intersection all cells having 12-15 sinkholes, 7-9 fractures, 12-19 fracture i~tersections, 3.7-7.2 km of fractures~ and 15-24% of the area occupied by sinkholes were identified. Only one cell had all of these characterisics and by the criteria established, it is

102

t~::" '

.. ~

", r;';

~~

; ~.

..... .r

f

tlrook 6< Allison: fracture Mapping/Subsidence

the cell most sn~ceptible to fl1ture sinkhole devel"l'ment... rntersect. i on 2 i~entifil=:!d cells \o'ith 10-15 sif'kholes, 5-1) fractures, ~-1q intersections, 2 .. 8-7.2 km 0: fractut"es, .:lnd 10-29~·t of . the area occupied bv sinkholes. fourteen cells wer~ identified, 13 more than were itlent.ified by intersection 1. Intersection 3 added lQ2 cells to those ident:iiied by intersection 2, and inte rsection tl addetl 127 cells to those identified by int~rsection 3: specified values f6r intersections 3 and 4 are qiven in Table 1_ Of the total 855 cells, 577 were not identified by

· intersection 4. Thesp. c~lls are c ons idered to he the least susceptible to futur e ground subsidencp. ("Figure

"8).

LINEAR COMBI~ATIO~

In linear combination modelling the variables an~ · the valu€s for each variable werp- weighted according to their judged influence or. the susceptibility of an area to future ground suhsidenr.e. Each cell ~as assignea a mar value based on the e1uation:

map value = W r + Ii " r + W r + ... W r xl 1<1 1<2 1<2 k3 1<3 kn

where w variable weight r va llle weight k = index of the variable. l-n

kn

The linear combination m0~el was generated using the variable and value weights liste .l in Table I1.. The number of sinkholes and the number of fractures in a c~ll were considere(l to be tbe most important measures of susceptibili ty to future sink·hole d€l'v€l'lopment and were assigned the highest weight 20.. The number of fracture intersections was "thought to be the next most significant variable, follove1 by the t.otal length of fractures in a cell; these variables were assigned weights of 15 and 12 respectively. The area of the a~ll covered by sinkholes vas considered the least useful in predicting future sinkhole development and was assigned the lowest weight, 6. Value weights for all variables ranged from 1 to 9.

For· the variable anrl value weig!lts shown in Table II, a cell with values of 5, :i, 7, 6, anc] 6 for the number of sinkholes, area occupied hy sinkholes, number of fractures, number of fracture intersections and total f racture length respectively, would have a map value = (20) (4) + (6) (1) + (20 ) P) + (15) (6) + (12) (3) .= 464. A sp.conrl cell with values of 2,1,3,4, and 3 for the same variables would have a map value = (20) (4) + (6) (1) + (20) (6) + (15) (4) + (12)" (6) = 338. ~ap values for each cell were calculated by CO:-lGRIO and then WE're clas si fied into fi ve qronps each cover ing a n equal

103

Er'

-- ' .1:

[j', o ' ; , . , ~ '

-'

,

LllV1I"Unrnent81 l\.arst

SUSCEPTIBILITY TO SINKHOLE DEVELOPMENT

High" Cells Included in Intersection 1.

MOder~telY high ~ Cells added by intersection 2.

Moderate B1 Cells added by intersection 3.

Moderalely low m~wl Cells added by intersection 4.

I N

M . ~ ~

0

• low £:=I Cells nol included in inlerseclion ,_ 4.

K ·<: ... . I "'~

fig . 8. Intersection ~C!I of 9'01llld subsidence suscepllbillty. Ooughert)' County , Georgia.

SUSCEPTIBIliTY TO SINKHOLE DEVELOPMENT

_H;9h

~ Moderately high

12m Moderate

I ii·'::. I Moderately low

DlOW

M.le.

0

• K.l o mltl l! "

Fi ~. 9. l i neilr cu",!llna tl on model of ground sulls i denee suscept f bill ty. Oolughert,)' County. Geor~ia.

104

3

• I N

, ,

Brook &' Allison: Fracture Mapping/Subsidence

portion of the total range of. map values assigne~. In a relative sense, these grour-s of cells were considered

. to have high, moderately high, mo~erate, moderately lo"V, and low sllscer-tibiliy ·to ground 5uhsidencp. (Figure 9). ' It should be strEssed that these terms are relative, cells oesignated as highly susceptible in the lillear comhina tion model 'may have a different susceptibility to cells designated highly susceptible fn the intersection· model.

DISCUSSION

Wor~ in Dougherty County has shown that fractures in the ocala limestone can be mapped from sinkhole data through >50 m of surface residuum. The fractUre map of

~ the county ' (Figure 6) should be useful in locating high yield, h.lgh specific capacity wel.ls. If all irrigation wells were of high specific capacity this would lII.inimize drawtlown and reduce the possibilit'y of qround subsidence. The most likely sites for such wells are a·t fracture intersections or along single fractUres (Parizek, 1976). The fracture map of the county may a.lso prove useful in selection of suitable sites for sanitary landfills. Improperly located landfills have al£eady resulted in contamination of the ocala aquifer in the Albanv' area·.. 'the most suitable locations for a landfill ~re taose falling in interfracture areas where there is Dot rapid recba.rge to the aquifer via underground solution cavities located along fract.ures. As Figures 2 and 6 show, however, Dougherty County is a poor "Waste· environIilen·t because a relatively permeable residuum with numerous si.nkholes overlies a heavily fractured bedrock. The most suitable w~ste disposal sites are located on the Tifton Upland, 15-20 km southeast of Albany. This area is underlain by impermeahle clays of the Hawthorne Formation.

The intersection and linear combination models o.f relative ground suhsid€nce susceptibility are in broad agreement (Figs. 8 and 9). "Furthermore, t.he.ir accuracy is supported by independent data. In both models, c~lls considered most susceptible to qrounll subsidence correlate with: (1) areas of sba.l1ov residuum {particularly le:Ss then 10m {Doug Wilson, unpublished data on residuum thickness)}, where subsidence may he m·ore rapid; (2) tvo troughs in the piezometric surfaCE of the Ocala aqUifer to the west of the l'lin t Si ver that W ai t (1963) feels are areas in which the l"imesto.ne is cavernous (Figure 10A); or (3) regions where the difference betwe~n the lowest piezometric surface on record (December, 1977) and the highest piezometric surface on recor~ (March 1978) exceeds 3m (Figure 10B). In thes~ areas there is a greater loss of hydrostatic support for the residuum during d["ouqh t periods. These relationships sugges·t that easily acquiren sinkhole and bedrock fracture data

lOS

:, I rori ! i -,

, "--'

\ .

J

....;

~ ~

" Ie: - ~

, ,

ii

I A

I ..... ~,. i._._ . ..

V.rt.~le

W~I'J~t V.llIt' I¥l .,;, v., •• V.'\1hU (\'IiI)

I S" ..... I(. lO \"IlL' 4 6 t I I t

[::-ot-rdo, (, tv I 4 6 , , 4 , I • SI.lholu

I"~ .. ,, ~ ~ I ':~l---;; .. , .-. ., "-H ,,_., "_,,

re«~t;;; - -- , 10.; , \., 10. " 15." It\·lt lS-n :!C.l4 .n

l;.;:~-;-;;- .. _- ---- -;. ~. - -"1-- ._.;- ... . 4 I" • I G

I,,,,,,,,, .~ 'j~I' .. . ~,:,;.~-;; -- ; _. ';' ~:; I l~l H '·1 I.t 10.11 II.Il"'""il.ii "." ".19 1""''''° n.... 1 L . , , to t I I , r~:::2111"O~ 'J~ --_.- ; I ~:o,· I.D.'~B·-' N~l' z.i.u -;;:,':s " ."S .• : ~.5-'i. ) 6.T.7:", ~ i:~~:~~: _.'1___ 1: .. ' _L_'-L'_ ~ to 9 I _. _ •

2\0.,..--·

~?oo 180

\op

[:::::::::::1 Greater than 20 feet Mn ••

0 , ~ Greater than J 0 feet

0 • r<lIom"I.'~ N

F-ig. 10. Fonn and variabllHy of the plezollll:!tric surface 1!l t he Ocala aquifer,

Dougherty County, Georgia . The piezometric surface of August 1957

is shown in (A), the difference between the lowest piezometl"1c surface

on record (December, 1977) and the hi9hest surface on record

(Ha rch , 1978) is showt"l i ll (8). Elevations arc 9iven in feet.

Diagram (A) ic; after :~djt (1963), (8) is aflp.r K~ader and Wagl!er (1980).

f

Brook & Allison: Fracture Mapping/Subsidence

·car. he used to develop relatively accurate ground subsiil€'nce susceptibility maps for covered ka·rst areas ..

BEFERENCES

Brink, .1\ ... B .. A. and Partridge, T.C.. (1965).. Transvaal karst: some considerations of development and morpholog·y, with special referenCE to sinkholes and . SUbsidences on the Far West ·Rand. South African Geographical Journal, 47: 11-34

Pard, D.C. (1964). Origin of closed depressions in the central Mendip !Iills. Abstracts of Pa~ers, 20th International qeo.graphic Congress, London:: 105-106.

Ffoka·ns, F .. H .. Georgia r Georgia ...

Jennings, J .. E ... Afr ica. q 1- 62.

(1977) DBMANG/CONGRID. The University of school .of Yorest Besollrces, . A.thens,

(1966). Building on dolomite in South Civil Engineer in South Africa, 8(2):

Kvader, T. and wagner, J .. E. (1980).. Ground water resou·rces o ·f the Dougherty Plain, southwest Georgia.. Unpublished report prepared for Decatur County, GeoLgia: 110 p.

Nevton .. J.G. (1977). Induced sinkholes - a continuing problem along Alabama highways. In .J .. s. Tolson and F. L. Doyle (eds.), Karst l1ydrogeoloqy, UniverSity of Alabama press, Huntsville: )0)-304 ..

Ogden, A.E. and Reger. J.P. (1977). Morphometric analysis of dolines for predicting ground subsidence, Monroe County, gest Virginia .. Ifl R.n. Dilamarter and S. C. Csallany (eds.), Hydroloqic Problems in Karst Regions, Western Kent'ucky Uni ve rsi t y Press: 1 )0-139.

Parizek r F~~.. (1976).. On the ~ature and significance of fracture trac~s and lineaments in carbonate and other terranes. In Karst ffydroloqy and Water Resources, Proceedings of the u.s. Yugoslavian Symposinm, Dubrovnik, 1975, VolUme 1: 47-100 ..

Pincus, R. J. (1953). Th e analysis o·f aggregates of orientation data in the earth sciences. .10 urn a1 of Geology, 61 : 48 2-~O9_

Pollard, L.O .. , Grantham r R.G. a nd Blanchard, H.F.. (1979). pre liminary appraisal of the impact of agriculture on ground water availability in southv·est Georgia. u.s .. Geological Survey, "Rater Besourc~s Investigation, 79-7.

107

I " . 'I' Er :' ,

"

I 1 '

~': "

~ ' : ~ . . . . ' \ 1 .

;

' ..

~ ( I' " , ii'

1:' r: f " ."

I',

f ~ I· 1';

I!

i I' I i

'';

:J "

Spigner, A.C. and Gra ves 5. L. (1 971). Ground-vater deve l opment problems associated with folded carbonate-rock aquifers io the Ironiale a~ea, Alabama. In R.R. Dilama rter an~ S. C. ,Csa llaDY (eds.), Rydrolo'lic Prol:lems in Kar.s t n"gions,

W€stern Kentu cky University Press: 2 41-248.

Vernon, R. O. ( 1951) . Geology o f Citrus and Levy Counties, Plori~a. Geo logical Bulletin 3 3~ Plorida Geological Survey ..

Wait , f? L.. (1 963).. Geolo'1Y and groun d water resources of Dougherty C'Junty, Georgia. rr. s. Sur vey Water Supply Paper, 1539-P.

Geological

Wi Ison, J '. D. an d analysis of Unpublished Planning and

PiCkering, S. M. (1973). Geological the DO!lgherty Co unty arp.a,. Georgia ..

report to the Southwest Georqia Development Commission: 40 p ..

108

'-'

~s an Indicator of Origin of Karst Landscapes in Indiana

Ardith Hansel

DepartJ!lent of Geograph y University of I llino i s Urlana, T.llinois 6180 1

hES!RACT

Morphometric analysis of surface form is . a viable approach to accounting fo r distinctions cetween different subtypes of karst. , This approach is tested on sinkholp. .foems exhibited by tWQ types of karst (i.p. .. , normal fluviokarst and exhuming fluviokarst) deveioped on the Mitchell Plain of south ct'ntral Indiana.. By utilizing surveyed data and sophisticated morphoBet ric techniques, it was possible t o · identify classes of sinkhol e ~orm types. The two karst landscapes . vere .

. fonnn to be form specific on the basis of a significa nt difference in the f.reque ncies with wh ich the representative form types occur in the tvo landscapes.

U7aODUCUON

An important object.ive of geomorphic research is th~t of ,determining the relationship between for m and or,191.0 in landscapes. This i s the morpllogenE! t ic tradition in .geomorphology.. Common goals of morphogenetic studi es inclucle a·ccounting for varia·tlon in surface forms and a'ttempting to identify lin.ks between the landforms an~ the conditions and events whic h give ri~e to them. T.wo basic proble ms have hindered the accomplis~ment of thest::t goals. First, inadequate descri ption of surface morphology has often made it impossi ble to undertake meaningful cOAparisons betwee n landscapes. Second , poor understanding of th~

geomorphic processes . and how· they \lork together to produce specif ic l anaf orms has made it very difficult to account for variation in r ea l \lorid relief forms.

Unlike many otheL processes in q~omocpholo9Y , the lime.stont:' solution process in ka r st is fa irl y well understood . Unfortunately , this understanding of limestone solution does not explain the variation in surface morphology in karst landscapes . . Aga in, the problem is twofo l d . Pirst, it is becoming increasinqly appaLent that limestone solutiop is but one of severa l interacting processes important in the karstification of a 1a ndscape.. Second, in many cases tho karst forms hav~ not been adequateljT defined. Recent ~-iorke rs

109

ENVIRONMENTAL KARST

Edited by Percy H. Dougherty

,.

I

GeoSpeleo Publications Cincinnati, Ohio

COVER PHOTO

COVER PHOTO: The cover photo shows the Pinnacles in the Gunong Mulu National Forest, Sarawak, Southeast Asia and is part of the illustrative material from the article by Michael J. Day on management problems in the park. The deeply-etched limestone spires rise up to 35 m through the forest covering the upper slopes of the mountainsides. The Pinnacles represent a well "de.yltl.o~ed, large scale spitzkarren which is a function of', ~'t"" i/lteraction be­tween environmental conditions and geologic structure.

Copyri ght 1983 by , GeoSpeleo Publications, Box 389022, Cincinnati, OH 45238

Library of Congress Catalog Number: 84-80188

ISBN ~ 0.9613107.0.7

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or trans­mitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written'permission of the publisher.

Printed in the United States of America by Diversified , Graphics Corporation, Wooster, OH 44691.

CONTENTS

Introduction

Contributors

Part I Environmental Case Studies EFFECTS OF HIGHWAYS ON KARST SPRINGS--AN EXAMPLE FROM POCAHONTAS COUNTY, ,lEST VIRGINIA by Eberhard Werner

2 SUBURBAN EXPANSION AND SINKHOLE FLODDING--A CASE STUDY FROM PUTNA~1 COUNTY, TENNESSEE by George Huppert, Larry Knox and Betty Wheeler

3 VALLEY TIDES--LAND USE RESPONSE FLOODS IN A KARST REGION, SINKING VALLEY, KENTUCKY by Percy H. Dougherty

4 LOCATION OF SALTPETRE CAVES IN TENNESSEE, ALABAMA, AND GEORGIA by Merilyn Osterlund

5 PROBLEt4S ASSOCIATED WITH URBANIZATION IN THE INNER BLUEGRASS KARST REGION by John Thrailkill, William Hopper, Jr., Michael McCann, and Joseph Troester (Abstract)

6 GROUNDWATER POLLUTION BY SEWAGE, CREAI-IERY WASTE, AND HEAVY METALS IN THE HORSE CAVE AREA, CENTRAL KENTUCKY by James F. Quinlin (Abstract)

Part II Management Problems~

7 KARST-RELATED t4ANAGEMENT PROBLEMS IN THE GUNONG MULU NATIONAL PARK, SARAWAK, EAST t4ALAYSIA by Michael J. Day

8 INTERPRETATION OF KARST FEATURES--A MANAGEt4ENT PROBW1 AT CLARK RESERVATION STATE PARK, ONONDAGA COUNTY, NEW YORK by John Mylroie

Part III Methods of Study

9 ,FRACTURE ~IAPPING AND GROUND SUBSIDENCE. SUSCEPTIBILITY ~10DELLING 'IN COVERED KARST TERRAIN--THE EXAMPLE OF DOUGHERTY

Page

vii

vi ii

3

15

25

37

51

51

55

77

COUNTY, GEORGIA by George A. Brook and Terry L. Allison 91 10 FORII AS AN INDICATOR OF ORIGIN OF KARST LANDSCAPES IN

INDIANA by Ardith Hansel ID9

v