ML20235T621
| ML20235T621 | |
| Person / Time | |
|---|---|
| Site: | Perry  |
| Issue date: | 02/28/1988 |
| From: | SEISMOLOGICAL SOCIETY OF AMERICA |
| To: | |
| References | |
| CON-#189-8106 2.206, NUDOCS 8903080390 | |
| Download: ML20235T621 (30) | |
Text
.
N iS - CC N-b DOCKET NUMBER N bkh
~
e Pron & UTIL FAC"W.wa .w.N . _
~ . -
- 7. my~., w' " .c.'W :2 s . ,.t,,
,fh0 y< m. = :' ~ ~ ~ n y -31, . , .v;-z,' s x v. a- p.
{
~ ' "
' ISSN 00371106 6' NUMBER 1 VOLUME 78 - . ..
BULLETIN OF THE J
(!.
- .. u . .
SEISMOLOGICAL SOCIE
~
OF AMERICA by_rmw M w'D . :
BOARD OF EDITORS f .J ~
DAVID M. BooRE, Editor, Menlo Park, Califonda f.C Q .
W
. AL17.R J. ARABASE, Salt Lake City, Uf ah M]. U O V ,[ p 7 bh b ,
n
( gg _
C,B.Caoust,I.ong Beach, California ~ ~' . r-- Docrns, ;
I I CHART.Es A. LANGSToN, University Park, Pennsylvania 818@tieh y
g
( c < L1
< .o i O (lj Ground Motions from Subductaon Zone Earthquakes -
FEBRUARY 1988 C. B. Crouse, Y v...
% e /
h K Vyas, and Bruce A. Schell 1 l .O ( Statistical Properties of Peak Ground Accelerations beor#by the SMART 1 Artsy Norman A. Abrahamson 26 The Effect of Topography on Earthquake Ground Motion: A Review and New Raoulta s '
Louie Geli, Pierre-Yves Bani, and Beatrice Jullien g2 An Analysis of the FEocts of Site Geology on the Characteristics of Near Field byleigh64 Wavee . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . J. R Murphy and H. K. Shah 83 Seismic beponse of a Class of Alluvia! Valleye for Incident SH WavneFrancisco J. Sdn Shallow Structure Effects on Bra =<hnA Teleasiamic P WaveformsThomas J. Owens Modeling Strong Motions Produwl by Earthquakes with Two. Dimensional Numerical Codes Donald V. Helmberger and John E. Vidale 109 Elastic Finite. Difference Modeling of the 1971 SanJohn Fernando, E. Vidate Califorma.
and DonaldEarthquake V. Heimberger 122 --
The 1978 Tahaa, tran, Earthquakes An Interpretation of the Stron_g Motion beords Jdar S
- Taheriand John G. Anderson 142.
e
- 'ne Foreebock Sequence of the 1966 Chalfant, California, Kenneth . Smith and Keith F. Priestley 172 The Northeastern Ohio Earthquake of 31 January 1986: Was It laduced?C. Nicholson, E. Roeloffs .
An Experiment in Earthquake Prediction and the 7 May 1986 AndreanofCarl Islande Earthquake Kisslinger 216 Subduction Zone Earth.
I bgional Variation in the Number of Aftershoeks (m, E 6) of tarp,ingh and Gerardo Sudrez 230 quak es ( M. E 1.0) . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . S. K S 243 The Seismicity of the Ade! aide Geosyncline, South Australia . . S. A. Greenhalgh and R Singh .
Geometry of the Juan De Fuca Plate beneath Washington and Northern Oregon from. .
Seismicity . . . . . . . . . . . , ........
Stress Fields due to Faulting in the Presence of a High Riswiity Barrier (Asperity)-Plane. . Kacpe Strain Models . . . . .
f*I Seismic Survey Over a Topographic Scarp in the Snake River PlainK D.299Miller, D. W. Ste t
& Event location at any Distance Using Seismic Data from a Single.Three. Component Station B. O. Ruud, E. S. Huseby S. F. Ingate, and A. Christoffersson 306 Frequency Domain Coberent Processing of Regional Seismic St'anals at Small ArreyZolta
- Application of Frequency Variable Filters to Surface. Wave Amplitude AnalysisDavid 339 R
- D U An Air Powered Impulsive Shear. Wave Source withand Repeatable SignalsHsi.hng Li Gary L Maxwell S55
{ qgj .....
e stt Short Notes 370 g g$ A Moment. Magnitude Relation for Hawaii . . . . . . . F. R Z4nica, M. Wyse, and F. Scher6aum
- ( ([ j Magnitude. Intensity Relations for Australian Earthquakes. A. Greenhalgh, D. Denham, R
. .Nition Rabmowier 380 I II) Microcarthquake location by Means of Nonu~near Simples Procedure
' (,) ( [ g Parkfield, California, Liquefaction Predictions. L Holzer, M. J. Dennett, T. L Youd, and 885 A. T. F. Chen LS() Seismological Notse . . .
390 Announcements . . 399 Chinese Pubbestions . . . . . . 399 Medal Nominations Invited ............ 400 Berkeley Seismographic Station Centennial .
e~ s 50g 8903000 90 880228 .
PDR ADOCK 05000440 PDR i $gl%R G .
t . m_
- g. _ _ _ - _ . _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ - _ - _ _ _ _ _ _ _ _ _ - - _ - .-
m -.-
4 e
Balletin of the Seismological Society of America. vol. 7s. No.1. pp.188-217, February 1988 THE NORTHEASTERN OHIO EARTHQUAKE OF 31 JANUARY 1986:
WAS IT INDUCED?
BY C. NICHOLSON, E. ROELOFFS, AND R. L. WESSON ABSTRACT On 31 January 1986, at 11:46 EST, an earthquake of m. = 5.0 occurred about 40 km east of Cleveland, Ohio, and about 17 km south of the Perry Nuclear Power Plant. The earthquake was felt over a broad area, including 11 states, the District of Columbia, and parts of Ontario, Canada, caused intensity VI-Vil at distances i of 15 km, and generated relatively high accelerations (0.18 g) of short duration !
&(fJ st the Perry plant. Thirteen aftershocks were detected as of 15 April, with six f 4 $t ( occurring within the first 8 days. Two of the aftershocks were felt. Magnitudes l S.: e ,3 for the aftershocks ranged from about 0.5 to 2.5. Focal depths for all of the i 5 gg g earthquakes ranged from 2 to 6 km. Except for one small earthquake, all of the I
g gg g aftershocks occurred in a very tight cluster with a north northeast orientation.
N tg,$t g( Focal mechanisms of the aftershocks exhibit predominantly oblique right elip motion on nearty vertical nodal planes oriented N15' to 45'E with a nearfy
, g horizontal P axis north of east, i
Three deep waste disposal wells are currently operating within 15 km of the
[
r' epicentral region and have been responsible for the injection of nearly 1.2 billion liters of fluid at pressures reaching 112 bars above ambient at a nominal depth of 1.8 km. Estimates of stress inferred from commercial hydrotracturing mess-
! j urements suggest that the state of stress in northeaster Ohio is close to the
! theoretical threshold for failure along favorably oriented, preexisting fractures. 1 f
i
! This implies that effective stress conditions near the bottom of the two most
! active wells may be at or near the critical level for incip;ent failure. Two and, f possibly, three earthquakes have occurred within less than 5 km from the wells since 1983.The relative distance to the main shock epicenter and its aftershocks (about 12 km), the lack of large numbers of small earthquakes typical of many l
induced sequences, the history of small to moderate earthquakes in the region I f prior to the initiation of injection, and the attenuation of the pressure field with
( f distance from the injection wells, however, all argue for a " natural" origin for ths
- 1986 earthquakes, in contrast, the proximity to failure conditions at the bottom of ,
the well and the probable spatial e6sociation of at least one earthquake suggest that triggering by well activttles cannot be precluded.
(
' INTRODUCTION t Earthquake occurrence in the Eastern United States is still a far from fully
- explained phenornenon. Although stresses within plate interiors are now known to be as large, if not larger, than those found at plate margins (e.g., Sbar and Sykes, 1973, Z,oback and Zoback,1980), and sneasured strain rates are sufficient to produce
'* 4 J large damaging earthquakes over long periods of time (Musman and Schmidt,1986),
b it is still unclear as to why certain earthquakes occur when and where they do. A k
- major handicap is the long repeat times between moderate to large earthquakes in
+ the east, and the fact that only a few have occurred since extensive monitoring Lt
I '
- capabilities have become operational. With so little information available, every
-O earthquake in the east becomes a valuable opportunity for furthel insight into the nature of intraplate seismicity. The magnitude m3 = 5.0 earthquake in northeastern
, , Ohio on 31 January 1986 is a case in' point. Because of its comparatively large size, f i the potential for aftershock activity, and because of its proximity to a major critical 188 4
4 I
b THE NORTHEASTERN OHIO EARTHQUAKE OF189 1/31/86 UARY 1986:
facility, the Perry Nuclear Power Plant, a substantial response by I comtnunity was initiated. Analog portable seismographs were operating w !
br of the main shock, and broadband, wide-dynamic range, digital in ON recording seven agenciesdata within 27 hr. The net result was that 49 stations we or institutions.
Several issues were raised by the occurrence of the earthquake. Of
- urred cbout was whether the main shock indicated a level of seismic h previously believed to exist in the region. The 31 January earthquake wa!
{
- cle:r Pawer
, the District at distances however, approximately 30 earthquakes of sm iort dur. tion recorded in the area (Figure 1). The largest of these prior earthquakes wa comparable magnitude (m3 = 4.5 to 4.7) and occurred in 1943. Neither the{ '
, earthquake nor the recent one in 1986 exhibited any manifestation of sur
/ Tcf the the earthquakes. faulting, leaving open the question as to the structure
! ,1 cf the
\dntation, Another aspect of concern was that the 31 January earthquake and its a ue right slip sequence may have been induced by adjacent high pressure fluid injection of ith e nearly hazardous and nonhazardous waste. Three wells that penetrate into ba 5 km cf the currently operating within 15 km of the earthquake epicenters. The observati iy 1.2 bpon
. min:l depth and the knowledge that under certain conditions uring meas- have triggered small to moderate size earthquakes led to the speculation clos) to the gnjection
- ,, wells may have played a significant role in triggering the recen ig fr
- ctures.
ie two most fvg 4 tr
- e. Two and, Vl v3' S
>m th3 cells a.
tttershocks j 82*
. cal of many Q 81
- BU*
n th2 r:gion , v:',. J re fi;ld with v^' .g,
)rigin for the ~ " , LAKEERIE
,, -y2*
/*tt:m of i et* e3* a:' 79*
s" *""
LOCATION ass o INSTRUMENTAL ggwy sEana q o NSW l' O O O 5 inTeustry far from fully o3
( now known to D Q
, 3,o.
bar and Sykes, h a 2.0+
E ient to produce yi.J g g Qo E - 41
- g (yo. c.
schmidt,1956),
iere they do. A O o s.o.
pg gy N earthquakes in -
- 1 ive tuonitoring ( (
tvailable, every the 31 January 1986 eanhvare), uake and(largeFic.1.
si Map of northeastern Ohio showing the lo Nuclear Power Plant (PNPP)
Ub Eh t into the according to intensity). s tid circles i ificant historica seismicity open symbols scaled n northeastern tvely large size, 1900 to 1940; solid triangles are deep waste inpetion wells drilled seismicity precedes ininanon ofinjection actmties. Diamonds are poor .
e between 19 a major critical based on felt reports; squares are instrumentally located earthquakes. M located earthquakes. twically Recent regional eanhquakes (M E 4.5) are shown in the inset. fied from Stov,er et al (1979). 1
(
b _
>rR
'1 4
,, 190 d. Nicuot. SON, E. ROELOFFS, AND R. L. WESSON This paper consists of two parts resulting frotn various lines of investigation carried out by the U.S. Geological Survey and incorporates cornpilations of data tude (Fi re '
municati n l
- from a number of different sources. The first is a basic study of the main shock and depth control its aftershocks and includes locations, focal mechanisms, and information on * .
'gati previous historical seismicity. The second involves an investigation of the deep fluid wi hi 0* 9f injection wells and an assessment of the degree to which the wells may have sity) at 0.32 e influenced the local pattern of earthquake activity. velocity model earthquake tra ,
EARTHQUAKE Ac rTVITY IN NORTHEASTERN OHlo horizontal stan l Historical seismicity Within the t i Cornpilations of historical earthquakes in northeastern Ohio based on felt reports estimated to be extend back to at ler the mid 1820's. Instrumental recordings oflocal and regional I"V'#8'0" f '0K earthquakes began in northeastern Ohio when John Carroll University, located in and a focal mec .
the outskirts of eastern Cleveland, started operation of its observatory in 1904. A nearly vertical n seistnicity map for this section of Ohio (Figure 1, Stover et al,1979) indicates a 1986). Figure 2 s sustained level of activity, most of which occurred prior to the initiation of major a lution (small e injection operations in the late 1960's and early 1970's. Since 1823, the repeat tirne The focal mech planes oriented for felt earthquakes is about 5 to 6 yr, although earthquakes large enough to cause ,
damage (intensity VI) are relatively rare. The tnagnitude 4.5 to 4.7 earthquake that cl ekwise from tl occurred on 9 March 1943 was recently relocated using a regional velocity model a small compom n dal planes is tl appropriate for the Central United States (Dewey and Gordon,1984). Its revised ,
location (Figure 1, 41.628'N 14 km,81.309'W 10 km) is only about 13 km west wave inversion c.
period characters of the 19S6 event. Thus, neither the size nor the location of the 1986 earthquake could be considered unusual. . Both the U.S. (
Other earthquakes of even larger magnitude have occurred regionally within and intensity studies 1986; Weston Ge(
around the state of Ohio (Figure 1, inset). The largest earthquake within the state was part of a swarm near Anna in 1937 and had a magnitude between 5.0 and 5.5.
Of particular interest, however, were two small earthquakes that occurred in NORTH.
northeastern Ohio in 1983. The first occurred on 22 January and had a magnitude STI m.o = 2.7 (NEIC). It was reported by John Carroll University, as well as by stations near Anna and in western Ontario. Its location is rather uncertain (Figure 1);
nevertheless, the best estimate ofits epicentral position (41.77'N,81.11'W; Weston Geophysical,1986) places the earthquake less than 5 km from one of the major i injection wells and within 5 km of the Perry Nuclear Power Plant. On 19 Novernber 1983, another earthquake of about magnitude 2.5 was also observed by stations operated by the University of Western Ontario (Weston Geophysical,1986). Its position is also poorly known, however, because its seismogram is similar in many respects to the January 19S3 event; its location is believed to be nearly the same. w_
lts absence from various earthquake catalogs in the United States implies a detection threshold for this part of Ohio prior to the 31 January earthquake of at least magnitude 2.5 or greater.
Main shock The earthquake of 31 January 1986 occurred at 16:46 UTC. There was no immediate foreshock sufficiently large to record on the instruteents at John Carroll University.The tnain shock was felt over a wide area and as far away as Washington,
$M%
D.C. The magnitude of the event was m. = 5.0 (NElC,1986) based primarily on data from Europe, or m.a = 5.0 (SLU) from surface waves. Fic. 2. Focal mechanis The main shock epicenter was located at 41.650*N latitude and 81.162'W longi-c%74%*[sN'NM
- n. w '. = ; q' -W'
~
z, - . - - - .
-..,. 2 7 T.~ W *,. v -
THE NORTHEASTERN OHIO EARTHQUAKE OF 1/31/86 191 tvestigation tude (Figure 1), using P-wave arrivals from 41 stations (J. Dewey, written com-ons of data Inunication,1986). The focal depth was held fixed at 2 km, owing to the lack of n shock and depth control and because hypocentral solutions from earlier iterations tended rmation on toward negative focal depths. All of the stations used in the location procedure were se deep fluid within 10' of the earthquake, the closest station being CLE (John Carroll Univer-a may have sity) at 0.32*, and the farthest was POW (Powhatan, Arkansas, SLU) ht 9.55*. The velocity model was the same as that developed by Nuttli et al. (1969) from earthquake travel times in the Central United States and resulted in a maximum horizontal standard error in location of i4.6 km at the 90 per cent confidence level.
Within the resolution of the data, the scalar moment of the main shock is estimated to be between 1 and 3 x 1022 dyne-cm (M = 4.6 to 4.9), based on an a felt reports inversion of regional surface waves, with a centroid focal depth between 4 to 6 km, and regional and a focal mechanism that is either right-slip (N22*E) or left-slip (N115'E) on vJocated in s'
\ nearly vertical nodal planes (Dziewonski,1986; NEIC,1986; Herrmann and Nguyen, f 1904. A 1986). Figure 2 shows two possible double-couple cornponents of the moment tensor #
[. ion Qofcatesmajor a solution (small circles), as well as the teleseismic first motion data (large circle), fl The focal mechanism solution using only the first-motion date indicates nodal '
e repeat time planes oriented at N55'E and N32*W (R. Needham, NEIC,1986) or about 35*
> ugh to cause clockwise from those of the moment tensor solutions. Both results, however, exhibit thquake that a small component of reverse-slip. Whether this discrepancy in the strike of the elocity model nodal planes is the result of incorrect first motions, poor resolution of the surface-
,). Its revised wave inversion or possible fundamental differences beween the long and short-it 13 km west period characters of the earthquake's seismogram is uncertain.
6 earthquake Both the U.S. Geological Survey and Weston Geophysical Corporation conducted intensity studies immediately following the main shock (Wesson and Nicholson, ly within and 1986; Weston Geophysical,1986). Some of the highest intensities found (Modified thin the state n 5.0 and 5.5.
NORTHEASTERN OHIO EARTHQUAKE 01/31/86 16:46:42.3 5.0 m t occurred in STRIKE: 55: DIP. 73: RAKE: 171*
3 a magnitude as by stations N T in (Figure 1);
yW; Weston T pe major
.aovember ed by stations
. cal,1986). Its ,% 9 milar in many 4, a ST. Louis arly the same. W- ' '
- -E lier, a detection (, o a T the of at least a N as #
There was no 4 ovmos at John Carroll e COMESSION =
ss Washington, S HARVARD d primarily on ,
motions carge circle. R Needham. NEIC,1986) and from inversion of near regional surface waves (small
,1,162*W longi. cirices)(SLU = Herrmann and Nguyen,1986. HARVARD = NElC,1986).
I
i 192 c. NICHOLSON, E. ROELOFFS, AND R. L. WESSON l
Mercalli intensity VI-VII) occurred up to 15 km away from the instrumental coseismic che l ' the predicted
! epicenter. Damage consisted primarily of broken windows, damaged chimneys, cracks in the walls and foundations, fallen ceiling tiles, items thrown off shelves, In Figure :
decreases are :
I and broken gas and water mains. The city sewage lagoon in Chardon sustained considerable damage, and a large area of disturbed water wells was found to extend . fault consiste from southwest of Chardon northeast to Thompson (Figure 3). Seventeen people main shock.1 l 1.5 km long, l were treated for minor injuries. Isolated intensities reached VII, although in general beneath the e :
the maximum intensity was VI. The intensity at the Perry Nuclear Power Plant m water level ;
was V.
Response of shallow water wells. Reports were obtained from 12 wells, indicating
! Note that an increase or decrease in water level, water pressure, or flow rate following the contams a em main shock (Geauga County Disaster Services Agency, written communication, of this reverst 1986; Calhio Chemical Company, written communication,1986; interviews with the top of the residents,1986). In one well close to the epicenter (A, Figure 3), the water level rose the sign of th:
by nearly 1.5 m and was sustained at that height for nearly 48 hr. Interestingly, the shock (A and level changes pattern of these changes is consistent with the pattern of cornpressions and gated to withi
/ dilatations predicted for the coseismic volume strain change associated with the Despite the main shock. Similar agreement between changes in wells and coseismic volume the wells, the strain change associated with fault slip has been reported by Wakita (1975) and Roeloffs and Bredehoeft (1985), although many examples exist in which observed such artesian volume strain level rise (ah) x ........ ... is the bulk mo 4i'4s -
3 i sis the accelerJ LAKE COUNTY : predicts a mal
- ranges from 1
- of Conglomerate l b,, : sites where the i!
5 wometers ,, i water level ros
- region where t
- ) THo son : Corresponding e
- difficult to ass
~
,' i taken qualitati-
- 4 major quadre o :
~
- : s Aftershock datc The analysis 8O '>
- 1986 and inch-e analysis of arri
} GEAUGA COUNTY {
-: O : groups, includi:
4i'33 + ,
,, .p Tennessee Ean CHARDg O physical Corpo 0 tation deployec
- 6 :
- !,,,,,,,,,,,,,%,,,,,: (e.g., MEQ - 80 GEOS instrums s t
- 15' st
- 05' ni2ed to radio ti.
were still in og Fic. 3. Volume strain produced by 22 cro of right sli over a 1.5-km long vertical fadt, oriented amphtude spect N55'E, extending from 1 to 3 km in depth, and centered be$ow the main shock epicenter (solid hexagon).
Contour interval is 0.2 ustrain; extension is positive, and compression is negative. Diamonds are sites of bilities of the G1 ete vel rise, and triangles are sites of water level fall. Sites indicated by A. B, and C are discussed in of station nam I+ * .M m %
+
4 -
O
- : n..n .' ' ~_ - --
~~*~~'"
}m
(( ]
l THE NORTHEASTERN OHIO EARTHQUAKE OF 1/31/86 193 coseismic changes in water level are too large, or of incorrect sign, to reconcile with e instrumental the predicted volume strains produced by earthquake faulting (e.g., Vorhis,1968),
sged chimneys' In Figure 3, the locations of water level (or pressure or flow) increases, and -
iwn off shelves, decreases are superimposed on a plot of the volume strain produced by slip on a trdon sustained fault consistent with the first-motion focal mechanism and seismic moment of the f und to extend main shock. The calculation assumes 22 cm of slip on a vertical, right lateral fault 3venteen people 1.5 km long, extending from 1 to 3 km in depth, oriented N55'E, and centered ugh m general beneath the epicenter of the 31 January earthquake. All 10 observations of changes ar Power Plant , g g ,; gg g ; gg Note that for faulting that does not reach the surface, each major quadrant 1 wells, indicatm.g contains a small region in which the volume strain is of opposite sign. The radius te followmg the ]
of this reversed region increases with the depth to the top of the ruptured area. If n e mmunicat,gn the top of the fault that slipped is placed at a depth of 2 km,instead of 1 km, then mteniews witf the sign of the observed water level changes at two wells just south of the main
[ DT I'V'I ' shock ( A and B, Figure 3) would be in disagreement. Therefore, the observed water
( jstingly, the level changes would suggest that faulting during the main shock may have propa- .
unpressions an I gated to within 1 km of the surface.
.ociated with the Despite the agreement between the sign of the observed and predicted changes in l I
oseismie volume the wells, the observed amplitudes are much larger than would be expected from akita (1975) and such artesian water wells and the size of the earthquake. A given comprehensive 2 which observed volume strain (ae), applied to a confined saturated aquifer, will produce a water level rise (ah) given by: Ah = (BK,/p.g)oe, where B in Skempton's coefficient, K, is the bulk modulus of the undrained reservoir rock, p,is the density of water, and
' ' ' ' ( g is the acceleration of gravity (Rice and Cleary,1976; Roeloffs,1987).This equation
- predicts a maximum water level change of 10 to 50 cm/ustrain, if B is 1.0 and K,
- rages from 10 to 100 kbars, typical of the type of rock forming the Sharon Conglomerate into which most of the shallow surface wells penetrate. At the two
/g :
i sites where the size of the water level change was reported (A and C, Figure 3), the water level rose by 1.5 and 0.6 m, respectively. However, both wells are located in a region where the inferred volume strain was on the order of 0.2 getrain or less, o oN s n pon% M u ewW waw M 6g d d 2 2 M m. h b, hMee, ;
i ; difficult to assign much quantitative weight to the water well observations; but if
- i taken qualitatively, then the spatial pattern of water level changes identify 3 of the
. i 4 major quadrants of volume strain produced by a strike slip earthquake.
t
[ Aftershock data and analysis i The analysis of the aftershock sequence covers the period 31 January to 15 April
- 1986 and includes data collected by the U.S. Geological Survey, as well as the
,COUNTYi analysis of arrival time and first motion data obtained from the other cooperating
- groups, including: Lamont Doherty Geological Observatory, St. Louis University,
- Tennessee Earthquake Information Center, University of Michigan, Weston Geo-i physical Corporation, and Woodward Clyde Consultants. Most of the instrumen-
- tation deployed consisted of single-component, high frequency analog recorders a (e.g., MEQ - 800's). There were, however,10 broadband, wide dynamic range digital ;
o- J GEOS instruments (Borcherdt et al.,1985) deployed with internal clocks synchro-nized to radio time code. These stations started operation on 1 February, and several i
were still in operation as of 3 April. Station locations, time histories, Fourier g vertical fauh, oriented amplitude spectra, as well as discussions of the deployment and instrument capa- l bilities of the GEOS stations, are given in Borcherdt (1986). A more complete listing l
'f, $ 'n$ d) ," .((,, A P, and C are 6 cussed in of station names, affiliations, and locations for the sites occupied during the !
l
194 C. NICHOLSON, E. ROELOFFS, AND R. L. WESSON aftershock study is given in Table 1. A map of most of the 64 sites occupied by the various stations deployed is shown in Figure 4.
The velocity model used to locate the earthquakes is given in Table 2. It is a 8g composite from several different sources and consista of five sedimentary layers ' U over crystalline basement at a depth of 2.1 km. The velocity interfaces chosen are CFD CLD !
HTG TABLE 1 STAnoN sean LocAnoNs tame DEPLOYED To McNrrOR Anr.RSHoCKs THAotlCH 15 APRIL 1986 tensma MIN com % auni a sme a. p ,,, g- _ PAT CON 41N42.06 81W12.55 PER GAR LDGO 1 Feb.-28 Feb.
41N47.30 81W10.64 LDGO TOM HLH 1 Feb.-2 Feb.
WEL 41N41.20 81WO7.01 LDGO HPV 41N44.41 1 Feb.-28 Feb.
81WO3.08 LDGO HSE 41N33.77 1 Feb.-2 Feb. WC01 81WO6.76 LDGO POP 41N37.23 2 Feb.-28 Feb.
TTR 41N35.25 81WO7.05 81W11.69 LDGO 3 Feb.-28 Feb. h LDGO WC04 WKR 41N36.06 81WO3.13 LDCO 2 Feb.-28 Feb.
2 Feb.-2 Feb. WC06 HSOH 41N35.66 81WO7.84 Michigan WC07 MTOH 41N36.68 81WO3.07 Michigan 1 Feb.-2 Feb. WC06 1 Feb.-2 Feb. WC09 CH0H 41N35.56 81W11.84 SLU HAOH 41N36.46 31 Jan.-3 Feb.
81WO8.51 SLU PAOH 41N45.41 31 Jan.-3 Feb.
81W11.95 SLU 31 Jan.-3 Feb.
CALM 41N34.1 81W10.3 TEIC 2 Feb.-7 Feb. V ELFM 41N36.8 81W10.9 TEIC FARM HOWM 41N38.3 41N35.0 81W10.4 81WO7.9 TEIC TEIC 3 Feb.-7 Feb.
2 Feb.-7 Feb. U i MONM 41N36.7 1 Feb.-7 Feb. 0.0 81WO2.9 TEIC C:
1 Feb.-7 Feb. 0.05 BUR 0 41N39.24 81 WO4.94 USGS (Denver) 2 Feb.-11 Feb. 0.50 0 CAL 41N41.21 81WO8.89 COT USGS (Denver) 2 Feb.-11 Feb. 1.00 0-41N34.73 81WO5.93 1.75 CUY USGS (Denver) 2 Feb.-11 Feb. O.
41N33.56 81W10.15 2.10 ERJ USGS (Denver) 3 Feb.-11 Feb. 17.
41N39.44 81WO5.00 20.00 FOT USGS (Denver) B Feb.-11 Feb. 20.
41N38.90 80W59.69 40.00 USGS (Denver) 4 Feb.-11 Feb. 99..
HAM 41N36.18 81WO8.48 HAR USGS (Denver) 2 Feb.-11 Feb.
41N36.67 80W59.62
- Cleveland Elee '
HWK USGS (Denver) 2 Feb.-4 Feb.
41N41.83 80W59.03 LOX USGS (Denver) 2 Feb.-11 Feb.
41N44.58 81 WO2.60 based on extens MON USGS (Denver) 2 Feb.-11 Feb.
41N35.52 81WO2.39 top of the Prect USGS (Denver) 2 Feb.-11 Feb.
WSH 41N37.61 81W13.30 USGS (Denver) 2 Feb.-11 Feb. average of down '
GS01 41N48.27 GS02 81WO8.52 USGS (Menlo Park) 1 Feb.-S Apr. S wave velocitie 41N43.75 81WO9.47 mantle are base, GS03 USGS (Menlo Park) 1 Feb.-3 Ape 41N39.45 81W10.07 USGS (Menlo Park) 1 Feb.-3 Apr.'
GSO4 41N36.85 81W17.55 Inain shock (i.e" USGS (Menlo Park) 1 Feb.-11 Feb.
GS05 41N35.64 81WO8.19 USGS (Mente Park) from refnction GS06 1 Feb.-4 Feb.
41N37.75 81WO3.77 USGS (Menlo Park) 1 Feb.-3 Apr.
well constrained.
GS07 41N32 40 81WO4.26 USGS (Menlo Park) velocities used at GS08 41N32.38 1 Feb.-11 Feb.
81W12.93 USGS (Menlo Park) Furthermore, this GSO9 41N24.81 2 Feb.-10 Feb.
81W11.91 GS11 41N09.20 USGS (Menlo Park) 2 Feb.-10 Feb. top of the Precan 81WO4.42 USGS (Menlo Park) 2 Feb.-10 Feb. m (6,000 feet) dev Michigan; SLU = St. Louis University; TEIC = Tenneuee Earthquake
- As a test ofInforma th incIuding a simpl U.S. Geological Woodward.Clyde Consultants. Survey; Weston Geophysical Weston Geophysical Corporation; Woodward.Clyde section over grani =
l lent station covers
-; y ,
_& ^a
,,, 1.
- A_% C abk R A '
-- d
195 THE NORTHEASTERN OHIO EARTHQUAKE OF 1/31/86 TABLE 1-Coruumed na .c ~ -n ,
ied by the
, m, n c.w
, w USGS (Menlo Park) 4 Feb.-10 Feb.
41N37.10 81WO7.18 4 Feb.-15 Apr.
. It is a GS55 41N40.45 81W13.41 Weston Geophysical CFD Weston Geophysical 1 Feb.-20 Feb.
ary layers 41N31.44 81W20.19 1 Feb.-8 Apr.
CLD Weston Geophysical
- hosen are 41N37.17 80W57.27 20 Feb.-15 Apr.
HTG Weston Geophysical 41N32.82 81WO6.12 1 Feb.-14 Feb.
KEL Weston Geophysical 41N27,77 81WO4.41 1 Feb.-1 Mar.
MFD Weston Geophysical 41N33.56 81W15.41 1 Mar.-15 Apr.
MtN Weeton Geophysical IL 1986 41N33.63 81W21.91 1 Feb.-15 Apr.
PAT Weston Geophysical 41N48.06 81WO8.81 2 Feb.-15 Apr.
"'"' PER Weston Geopbyeical 41N4 4.29 81WO3.09 24 Feb.-15 Apr.
TOM 81WO9.31 Weston Geophysical 3Feb. WEL 41N45.00 Woodward Clyde 31 Jan.-15 Apr.
2 Feb. 41!.36.90 81W18.08 1 Feb.-16 Apr.
WCol Woodward Clyde 28 Feb. 41N40.05 81WO9.53 1 Feb.-14 Apr.
WCO2 Woodward-Clyde 41N43.87 81WO4.46
} WCO3 Woodward Clyde 1 Feb.-22 Feb.
t WC04 41N35.10 81WO9.36 Woodward.Clyde 1 Feb.-14 Apr.
%.A WC06 41N32.40 81WOL75 Woodward.Clyde 3 Feb.-24 Feb.
-28 Feb- 41N48.00 81WOS.58 8 Feb.-25 Mar.
WC07 Woodward.Clyde
-2 Feie. 41N40.24 81W14.48 23 Feb.-14 Apr.
WC08 Woodward-Clyde 41N35.45 81WO9.36 27 Mar.-14 Apr.
-2 Feb WC09 Woodward-Clyde
-2 Feg' 41N40.04 81W14.45 WC10
.- 3 Feb. TABLE 2
..-3 Feb.
VtLocrry MooEL Usto to Locm EvzNTs Listro IN TAat.s 3 t-3 Feb.
3,,, , , , , , , y,,,nm , y,,,,g Vdv.
nW 1-7 Feb. <w %f% %fe 1-7 Feb. 0.60 3.00 Glacial till 0.0 0.05 1.80 1.90 Devonian shale gj y g, 0.05 0.45 3.00 1.58 2.33 L80 Silurian t!olomite 0.50 0.50 4.20 b.-7 Feb. 2.53 1.78 Ordovmu ilmestone and dolomite 0.75 4.50 1.00 2.70 L76 Cambritti sandstone and dolomite b.-11 Feb. 0.35 4.75 Precastrian granite 1.75 L74 6.15 3 54 eb.-11 Feb. 2.10 17.90 lower crust 6.70 3.87 1.73 eb.-11 Feb. 20.00 20.00 Mantle 8.15 4.63 1.75 eb.-11 Feb. 40.00 99.00 eb.-11 Feb. I Cleveland Electric Illuminating Co. (1982).
{(7b E
~
based on extensive regional compilations of well data drilled top of the Precambrian basement (Cleveland Electric Illuminating An d
at leas reb.-11 Feb. average of down hole and cross hole velocity logs is used to determ Feb.-11 Feb.
Feb.-11 Feb. S wave velocities in the upper 0.5 km. Veloc d Feb -3 Apr.
Feb.-3 Apr.
Feb.-3 Apr.
main shock (i.e., Nuttli et al,1969). Velocities i h i well constrained. With the exception of the near surface l Pi and S veloci b F Feb -3 Apr. velocities used are not based on actual b 830in situ m Feb.-11 Feb. top of the Precambrian interface, which near the shore) of Lake Eri
- Feb -10 Feb.
? Feb -10 Feb- m (6,000 feet) deep but near the epicentral i reg 2 Feb 10 Feb. ~
including a simple layer over a half space to accommodate l the slow Ugan = University of section over granitic basement (Wesson g and ggg Nicholson g;1986). Owing tion Center, USGS = ,
m Woodward.Clyde =
f -.-
(
196 C. NICHOLSON, E. ROELOFFS, AND R. L. WESSON .
, , , , , i ,
s: - so -
LAKEERIE c$ci nn '4a ' 4 LAKEE
- A WEL g
&C502 A WCO3 PAINESYtLLE
} A MNK CALA A HyH & IDH "8'~
gg . . WCC8 A A CFO
~
A Fame
'02 e . ,0, A g a+5++ . A C300 j p p -CS55 A HTG WCOI AA C504 A gt HAM A wnm "
CHeH US A Meu ~
W OY COT A f.R T. . i a MgN T O HIE
& C508 bC307AWCOs A CLC w- ,
~
3C' ~ , , , , ,
23* 10' ele a
Fic. 4. Seismograph stations deployed by all cooperating institutions to record aftershocks of the January main shock (large square). Solid symbols are digital recording sites. Shaded creas on this and '
subsequent mape represent areas of dense population. Da4 bed lines are county boundaries.
O_
tried, the earthquake epicenters did not vary by more than about 0.5 km. Only the focal depths were significantly affected, with systernatic biases of up to I km observed between the various velocity models used. .--
The earthquakes were located using HYP0 ELLIPSE (Lahr,1985) and as many [ tt:
of the available arrival times as were internally consistent. Arrivals based on the z digitally recorded GEOS instruments were given preferential weight because of the E higher precision of timing, the greater resolution in picking the arrivals, and the $
greater confidence in identifying the shear wave arrival on the three-component instruments. A serious complication was that many of the single component stations reported secondary arrivals that were often a converted phase (e.g., S to P). Thus, -'o in order not to mix both converted and direct shear arrivals, only S arrivals from
- the GEOS three-component stations were used. However, all of the available P.
wave data were included, permitting better station coverage and therefore greater precision and azimuthal control. Fic. 5. (Top)Lo as of 15 April 1986. I epictaters. Local ne Aftershock locations the deep waste dap '
nactmtion. nlons i As of 15 April,13 aftershocks were located (Figure 5). Most of these events depth.
occurred within the first 8 days; two were felt. Coda magnitudes ranged from 0.5 to 2.5, based on a formula developed for earthquakes in the Central United States believed to be q (Stauder et al.,1981). These values match other magnitudes derived from scalar etwork of stati<
moments determined from the GEOS digital recordings, assuming the largest on the order of a aftershock is equivalent to a magnitude 2A (see Table 3; Glassmoyer et al.,1986; event being reco Borcherdt and Glassmoyer,1987). In addition to the aftershocks, several events Figure 6 show
,p -
,b - - A"X6 *-
e T = _' x.--,7?*-% C.
~
-Q . . .
- d
. . . e w**~
~
1 THE NORTHEASTERN OHIO EARTHQUAKE OF 1/31/86 197 l i ,
h an ,
otPfw5 LAKEERIE Q * **
" )
- Q1983 -
saar f,g t -
c 40 f
j o asna ts o s.o*
"" ~
4 .
- 2 to
- ..o-
_ nacussuct LAKE * ~
- i.o*
a c.c.
6 Mio 1943 1
, O t.c.
k]7 l
. j O .o-GEAUGA {
s.o.
)
%.o*
C* no un O
'- ' 5.o*
g ,
a i
_ ac- to-l -
si.
aftershocks of the a orca.s on this and a-2nes. '
r i i km. Only the o # "" I if up to I km ; _.
i Paleozone
. g.. .o
.) and as many !, 1983 ,
s based on the : Basernent because of the g O
- -%, and the $ l
,bponent --6.3
,o .at stations S to P). Thus, i arrivals from
e available P- '
erefore greater DISTANCE (KM) as of 15 Apnl 1986. Imre uncertainties in location are associated epicenters. Iecal quarry blasta are shown as crosses. CHfI and CHf 2 (Calhio fI and Calhio f 2) are f these events the deep maste disposal wells. SALT in the Painesville brine well. (Bottom) Vertical crosa.sect .
g' ration, al ng the line A-A 'shown above. Main shock focal dep'n orresponds with snean centroid ed from 0.5 to Un.ited States
.d from scalar believed to be quarry blasts were also recorded (Figure 5). Because of the dense ig the larges! network of stations (Figure 4), location accuracy for all of the eventa detected was "d"" '
on the order of 0.5 km at the 90 per cent confidence level. with even the smallest
- e. ral events event being recorded by at least six stations.
Figure 6 shows the aftershock locations as well as station coversge within the I
.. 4
F I
1%
C. NICHOLSON, E. ROELOFFS, AND R. L. WESSON TABLE 3 Data LOCATIONS oF RECENT EARTHQLlAKEs AND Bt.ASTS IN NoatutAstr.AN OH onen Latitude Longitude Depth (YrMoDy) tHrMn Sec) (des ann) (der min) (km) g No.
raw ERH ERZ Amauth PHA (see) (km) (km) Cap Main Shock and Prior Eventa 1 430309 03:25 25.00 41N37.80 81W18.60 6.50 4.7 j ..
830122 07:46 57.80 41N45.90 81W 6.60 2.00* 2.7 14 1.45 IL9 18.8 206 I 860131 16:46 42.30 41N39.00 81W 9.72 2.00* 18 0.38 4.20 5.80 92 5.0 41 0.75 4.60 ,
58 i Aftershocka 860201 18:54 49.20 41N38.82 81W 9.42 4.971.5 860202 3:22 48.53 41N38.75 81W 9.53 4.99 21 0.13 0.80 1.66 100 1.2 24 0.06 0.25 0.57 l 860203 19 47 19.61 41N38.84 81W 9.50 6.10 72 e 860205 1.9 6.34 2.40 41N38.94 81W 9.64 2.07 44 0.10 0.33 0.68 73 0.9 860206 18:36 22.24 41N38.56 81W 9.64 5.922.4 20 0.21 0.83 0.98 49 860207 43 0.12 0.39 0.82 48 15:20 20.20 41N39.06 81W 9.24 4.59 1.4 s .,,,
860210 20:0613.49 41N39.16 81W 9.27 4.97 1.3 27 0.08 0.29 1.06 42 860223 26 0.09 0.42 1.16 69 3:29 48.41 41N39.10 81W 9.30 5.13 0.9 i 860224 16.55 6.37 41N38.96 81W 9.81 3.72 23 0.09 0.39 1.17 73 860228 0.8 1 1:39 34.07 41N39.11 81W 9.59 4.31 10 0.09 0.71 3.32 126 0.8 12 0.06 0.69 1.66 ]
860308 20 42 49.48 41N38.71 81W 9.31 4.42 92 !
0.9 860312 8.55 26.59 41N43.64 81W10.25 2.00 20 0.10 0.46 1.37 103 I Fic. 6. (L 860324 0.7 (Right) Vertic 13.42 41.22 41N38.10 81W 9.95 4.66 10 0.06 1.35 0.78 216 !
1.5 860410 6:58 5.59 41N38.74 81W 9.67 4.69 16 0.10 0.53 1.56 88 ;
0.8 epicenters, 22 0.09 0.36 0.86 65 Blaats to one of tj 860205 15:39 6.45 41N40.08 81W 2.28 0.90 1.1 i shock. A v -
860205 17:57 3.85 41N40.02 81W 2.46 0.01 1.0 13 0.09 1.30 1.24 74 '
12 0.07 0.46 0.74 75 Figure 6) is
{ with a nortl Magnitudes. Moments, and Focal Mechases Date origin j in additic (YrMoDy) (HrMn Sec) Magmtude N.E.t Marmeude Moment Marmtudel N/NE Slip Plane P Axis N.M.3 (dyne-em) (6) i was detecte(
Strde Dip Shp Anmuth Plunge 860131 16:46 42.30 (centroid depth: 4-4 km) (ftrst. motion data) 55 73 171 280.4 5.8 at GSO2 an, 1-3E+23 4.6 22 81 860201 18:54 49.20 1.5 1.5 160 69.6 7.3 depth of 2.(
860202 3:22 48.53 0.9 1.2 2.0E+ 18 1,1 30 70 169 72.0 6.6 zote se 860203 19:47 19.61 2.0 1.9 21.3E+18 70 75 160 293.3 2.9 860205 6.34 2.40 0.1 2.1 45 82 140 273.3 21.1 Lake Count) 0.9 1.3E+18 0.9 by adjacent i 860206 18:36 22.24 2.5 2.4 38.8E+18 195 90 165 61.0 10.5 2.4 860207 15:20 20.20 1.1 1.4 8.0E+18 30 90 165 256.0 10.5 size would r 860210 20: 6 13 49 1.0 1.3 4.6E+18 1.6 5 70 169 47.0 6.6 earthquake s 1.5 180 80 -170 .
060223 3:29 48.41 0.1 0.9 45.9 14.1 in e ep cen '
1.0E+ 18 860224 16.55 6.37 -0.1 0.8 0.8 7 80 170 50.9 0.1 860228 1:39 34.07 -0.1 0.8 "#
860308 20 42 49.48 0.1 0.9 860312 8:55 26.59 -0.3 0.7 0.8E+18 2 70 160 43.4 0.8 I Single-ever 0.7 860324 13:42 41.22 14 1.5 tions) were cc 860410 6.58 5.59 -0.1 0.8 195 85 -180 59.9 3.5 deployed (Fig
. Fbed depth. 25 81 150 250.9 14.3 FPFIT (Reas4 t New England coda. magnitude formula (M, = 2.2 tog D - 1.7) (Chaplin et al.1980). nism solutiom 40 see) (Stauder et al,1981).t New Madrid coda. magnitude formula (M, = 2.7planes .,
log Doriente - 2.7, D R 4 from tast nor i Moment magnitude (M. = tos Me - 17.2,. M. < 2.5; eM- 10.7, .
- 0 67 log M M, t 2.5), therefore repti nodalplane is e remains in a very small cluster, there was one event on 24 (Figure 6) and about 1 km outside the immediate source region of the main shock. right slip on a Its location If 80. then mot to the south southwest, coupled with a poorly resolvedaetrend in the earthqu k The second e
.l
- ';.;n.
- .; : ,. c*+
~.; a9 ,c.5 ; v ,~ :=r_ v .x.1-
_' q.
. g g; m . 2
,. ~ ~ -
y q~~; ;._ .
, , ~- - _
THE NORTHEASTERN OHIO EARTHQUAKE OF 1/31/86 199 s ,.
OHlo ERz Anmuth skm) Gap . , , LAKE COUNTY l
i 16.8 206 8
.n g l 5.80 92
~ '
,ylB ...
f 1.66 100 0 72 e== a 0.57 73 0 0.68
- O PA 49
-"a,,,,
[
\
48 42 0 EU oE AUGA CcQNTY eau .
o.
lab 69
.s .
1.17 73 g.
0 DestANcs (Ku) 3.32 126 1.66 92 Fic. 6. (Ie/t) Map of aftershock locations within the immediate epicentral region of the main shock.
1.37 103 (R@t) Vertical cross.section perpendicular to N30*E.
0.78 216 1.56 88 epicenters, suggests a short fault segment oriented 25' to 35' east of north, or close 0 86 65 to one of the nodal planes observed in the focal mechanism solution of the main shock. A vertical cross section taken perpendicular to a strike of N30'E (right, 1.24 74 Figure 6) is consistent with rupture initiating at depth on a nearly vertical fault 0.74 75 with a north northeast orientation.
In addition to the tight cluster of aftershocks, one small earthquake (M, = 0.7) was detected near station GS02 (Figure 5) on 12 March.The seismogram as recorded
- on the GEOS instrument (GS02) is shown in Figure 7. Its location based on arrivals l I
Arunuth Plung, at GS02 and the Woodward Clyde stations is 41.727'N,81.170*W, with a focal t 280.4 s.8 depth of 2.0 km. The correspondence of this hypocenter with the base of the
} 69.6 7.3 paleozoic section and its relative proximity to two of the deep injection wellt in
> 72 sj Lake County (less than 3 km) suggest that this single event may have been triggered by adjacent injection activities. It should be noted that similar earthquakes of this 3 21.1 size would not have been detected prior to the occurrence of the 31 January 5 256.0
/0 10.5 10.5 earthquake and the subsequent deployment of sensitive seistnographic eqGpment 4
6j in the epicentral region.
0 b0.9 0.1 Focal mechanism solutions Single-event focal mechanism solutions (lower hemisphere, equal area projec- l 0 43 4 0.8 tions) were constructed using polarity data from nearly all of the ternporary stations l deployed (Figure 8). Nodal planes were determined using the grid searching program l FPFIT (Reasenberg and Oppenheimer,1985). Two general classes of focal mecha- {
0 2 0~9 14j nism solutions were observed. The first type of solution (Figure 8A) exhibited nodal planes oriented northeast and northwest. The P axis is nearly horizontal and varies gg ,'
from east northeast to east. This class includes the two largest aftershocks and
. g i,, p + 0.2, o <
therefore represents some of the best-constrained results. The northeast striking
- f. t 2.5). nodal plane is consistent with the general distribution of the aftershock hypocenters (Figure 6) and would typically be assumed to represent the actual plane of faulting.
.rshock activity f If s , then motion during the earthquakes would have been predorninantly oblique that is located right slip on a nearly vertical fault. l its location to The second class of focal mechanism solutions (Figure 8B), a"lthough somewhat the earthquake i
F
(
l
' l 200 C. NICHOLSON, E. ROELOFFS, AND R. L. WESSON statten.co2 less well resolt oriented more M. L ..
- found within a F "W than one favo-stress associat permit alternaq 0.020 80). These at:
7 A predominance h l."M, . af/pVNM*4,W'. n- _- _-- -
M.0 ggygg 0.015 Relation to reg 3c.oi:-
In general, maximum Cort o.cos -
axis determint stress orientat Furthermore, pq
~
a rKhm- _- - _ _ - - - - - n.so All of the es f interface of t}
common with <
tation of the e; the western fld 26 28 30 .
time (utistsesiotticaiss + secemos d2P monotom. ej I Arch in westem Fic. 7. Seismo am of small event on 12 March near station GS02 within 3 km of the Calhio injection wells (courtesy of Borcherdt, C. Mueller, and G. Glassmoyer). See Figure 5 for location. Province that I of Ohio has be(
explorations, {
A !!V!: 2?t. !?fil JYa !?Y?! L'Y- !!Y!? J7. !!y;; 21. f*yc La/. !?v;f Jy;, gravity and mt A .
basement strual
)
. g . .
$, '
- 4* *
! a * .. .
consists of a sf
,.,* \
- / .'
, the southeast ;
N and Kucks,1%
anomaly map
< (Hildenbrand B !?Vil 'JY. !??J 'J"- !!Vf! J'.'- basement matz 7,7
- same strike asi with the locati a' ', . *
- ., e .t regionalstrikel structures, hoi MacWilliams,l C !?v;, ev. ;?vr! !a,*, ;fy;; ta. !, r?ve Laf;, ;?vt y. observed in thi
- j. y
. . .\,', .t. .'- <
Previous exanj
}(. s . . . ,.
- s .' It has been a Fio. 8. bwer hemisphere, equal. area. single event focal mechanisms determined using a grid search in fluid pressuf for nodal planes. (A) Oblique sli earthquakes (c northeast. (B) Nearly pure stnke:p focal mechanisms slip toechanisms withoriented with a nodal plane a nodal plane north. south.oriented (C) Alter-north northeast-native solutions for poorly constrained focal mechanisms. Solid circles are compressions. and open circles convincing ar@
are dilatations. Legend indicates ongin time, focal depth, and magnitude. character 1sticQ
- Et .: Q -- s meg -
~
. w . . .e . . .e w. % 1r n.' .Raa. '4 r2'A,'E M_' - .%..~. . . . . - -.
1
- - - - - - _ _ _ _ . _ l
l kON THE NORTHEASTERN OHIO EARTHQUAKE OF 1/31/86 201 l l
s m ie .cor less well resolved, exhibits a nodal plane oriented nearly north south and a P axis '
oriented more toward the northeast. This observation of different focal mechanisms found within a single aftershock sequence is not unusual and suggests that more than one favorably oriented, weak fracture is being reactivated by the change in stress associated with the main shock. Several of the smaller earthquakes also permit alternative focal mechanism solutions with large dip slip components (Figuie 8C). These alternate solutions are considered unlikely, primarily because of the predominance of the nearly strike-slip mechanisms exhibited by the better-con-M*0 strained results.
Relation to regional tectonics in general, most of the focal mechanisms are consistent with a horizontal maximum compressive stress field striking northeast to east, consistent with the P axis determined from the main shock (Figure 2) and regional determinations of stress orientation for the Central United States (e.g., Zoback and Zoback,1980).
9 MeSQ Furthermore, because the aftershock slip orientations vary from predominantly strike 8 lip to oblique slip with typically 10* of rake or less, this implies that the vertical stress is probably the intermediate principal stress.
All of the earthquakes occurred at or below a depth of 2 km, corresponding to the interface of the Paleozoic section with the Precambrian granitic basement. In common with other earthquakes in the Eastern United States, no surface manifes-tation of the active fault plane has yet been found. The earthquakes occurred on the western flank of the Appalachian Basin, within which rocks of Paleozoic age so dip monotonically to the east and southeast away from the axis of the Cincinnati Arch in western Ohio. Basement rocks, however, are an extension of the Grenville Oodof Province that forms the eastern section of the Canadian shield. Although this area of Ohio has been extensively surveyed and drilled, primarily for natural gas and salt explorations, few investigations penetrated to the' crystalline basement. Regional
- S L"! !".T .".*; gravity and magnetic studies, however, suggest the presence of several large scale
, . basement structures, one of which defines a northeast-striking lineation. This line k consists of a series of magnetic highs separating a region of low megnetic relief to
., 't * #] the southeast from an area of high magnetic relief to the northwest (Hildenbrand and Kucks,1984a; Weston Geophysical,1986). Moreover, the Bouguer gravity anomaly map shows an odd shaped high centered over much of Lake County (Hildenbrand and Kucks,1984b) that could reflect an area of higher density j k basement material offset in a right lateral sense across a zone that has nearly the '
y' same strike as the ridge defined by the line of magnetic highs. This line is coincident with the location of the main shock and the distribution of aftershocks, and has a
- ~
regional strike of about N40*E. Most recognized subsurface faults, or other basernent structures, however, typically trend northwest or west northwest (e.g., Root and MacWilliams,1986), similar in orientation to many of the auxiliary nodal planes observed in the earthquake focal mechanisms (Figure 8A).
FLUID INJECTION IN NORTHEASTERN OHIO Previous examples of earthquakes and injection well activities
' It has been conclusively demonstrated that under some conditions, the increase in fluid pressure in the earth's crust as the result of the injection of fluid can tigger
Yndd o N.NrtUeEt- earthquakes (c.f. Raleigh et al.,1976). In each of the well documented examples, j ted north south. (C) Alter- convincing argurnents that the earthquakes were induced relied upon three principal
- apressions. and open circles characteristics of the subsequent earthquake activity. First, there is a very close e -
l A
202 C. NICHOLSON, E. ROELOFFS, AND R. L WESSON geographic association between the zone ofincreased fluid pressure and the locations area to become of the earthquakes in the resulting sequence. Second, calculations based on the concentrated ne measured or inferred state of stress in the earth's crust, and the measured injection as far away as e pressure, indicate that the theoretical threshold for frictional sliding along favorably the Ohio eartht l oriented, preexisting fractures was likely exceeded. Third, a clear disparity between mining and his i
' the previous seismicity and the subsequent earthquake activity could be established, operation withi-with the induced seismicity often characterized by large numbers of small earth-quakes that persisted for as long as elevated pore pressures in the hypocentral Solution salt mi region continued to exist. Two of the best documented cases for induced seismicity The associati are summarized below, because they bear directly on whether the' Ohio earthquake i western New Yt may be considered induced, and to illustrate the quality of the evidence for previous operations in nc examples of seismicity related to fluid injection.
of the recent se In 1962, at the Rocky Mountain Arsenal near Denver, Colorado, the injection of mining for salt 17 to 21 million liters / month of waste in to a 3671 m deep disposal well was quickly Nevins,1981) e followed by many felt earthquakes in a region where the last felt earthquake had operations have occurred in 1882 (Healy et al.,1968). A study of event locations showed that the the Silurian Sal earthquakes were occurring in an elongate region about 10 km long and 3 km wide, from Lake Erie.
centered on the well and at depths of 4 to 7 km. A comparison of earthquake be argued that i frequency and average injection rate showed a convincing correlation. Although associated with :
injection was stopped in February 1966, earthquakes continued to occur, not near occurred. In par the base of the well, but primarily within the previously defined linear zone and at located within ti a distance of 4 to 6 km. The largest earthquakes in the sequence (between 5 and 1940, about 50 k 5.5) occurred in April, August, and November 1967, after which activity began to in view of the !
decline. Hsieh and Bredehoeft ,(1981) demonstrated that the records of pressure solution mining.1 falloff at the disposal well were consistent with injection into a long, narrow reservoir; a conclusion supported by the elongate shape of the seismogenic zone. northeastern Ohl it seems reasonal Based on their model, a fluid pressure increase of 32 bars was apparently sufficient solution mining {
to trigger seistnic activity along favorably oriented, preexisting fractures. No hy- activity. Moreov, draulic stress measurements were ever made near the Rocky Mountain Arsenal.
than 50 bars), the Healy et al. (1968) inferred a least cornpressive stress of 362 bars at the bottom of several imperme the disposal well from the pressure at which the volume rate of injection increased rapidly, and estimated a rnaximum compressive stress to be at least the overburden operations mining having within}
si pressure of 830 bars. The estimated formation pressure prior to injection was 269 or having triggert bars. With injection pressures at the Rocky Mountain Arsenal having apparently reached a maximum of 72 bars above ambient, fluid pressures within the reservoir High pressure we were inferred to be capable of initiating failure along favorably oriented fractures with cohesive strengths of as much as 82 to 100 bars. Three high pre Solution mining for salt near Dale, New York, triggered a marked increase in rently operating s region of the ma:
microcarthquake activity in 1971 (Fletcher and Sykes,1977). As many as 80 earthquakes / day were concentrated within 1 km of a 426 m deep injection well in within the Paines salt mining had p an area where the previous record of activity was less than one event / month. Top.
well prior to 31, hole pressure at the injection well typically operated between 52 to 55 bars, or only presure of 55 bars a few bars less than that calculated to induce sliding on preexisting fractures with no cohesion, based on the analysis of hydrofracture stress measurements conducted 1986). The other about 100 km from the activity. The low level of background seismicity prior to poss ble earthqua operation. These t high pressure injection, the dramatic increase in activity following injection, and the rapid cessation of activity following a decrease in injection pressure below about CHf 2, Figure 5) 50 bars strongly suggested that this seismicity was induced. Chemical Compa:
agricultural fungic These two cases demonstrate that, in sufficiently prestressed regions, elevating formation pore pressure by several tens of bars can cause a previously quiescent Natural Resources well began in 1975
_=_ m n..n
_ _ -__ (-. MM y
Q[ MO*
-., ~
m .
-f. y . --
^ ^'
- THE NORTHEASTERN OHIO EARTHQUAKE OF 1/31/86 203 and the locations ns based on the area to become seismically active. Furthermore, although initial seismicity is usually concentrated near the base of the active wells, later seismicity can occur at distances e:sured injection g e.long favorably as far away as several kilometers and long after injection has ceased. In the case of Lisparity between the Ohio earthquake of 31 January, both types ofinjection activities (solution salt
'd be established, mining and high pressure waste disposal) have occurred or are now in current t operation within the northeast region of the. state.
4 of small earth-the hypocentral Solution salt mining
\
duced seismicity Ohio ezrthquake The association of solution mining with the occurrence of small earthquakes in
! ence for previous western New York State (Fletcher and Sykes,1977), and the extensive salt mining operations in northeastern Ohio (Clifford,1973), suggested the possibility that some l
l , the injection of of the recent seismicity in Ohio may be related to solution salt mining. Solution tydl wts quickly rnining for salt began in northeastern Ohio in 1889 (Clifford,1973; Dunrud and
\ quake had Nevins,1981) and continues to the present, although several previously active Id that the operations have been closed down. The target horizon for the mining operations is e.nd 3 km wide, the Silurian Salina formation at a depth of 600 to 900 m, depending on distance n of earthquake from Lake Erie. Based on their spatial proximity and temporal association, it could be argued that several earthquakes in the northeast region of the state could be ation. Although occur, not near associated with solution salt mining operations active at the time the earthquakes iear zone and at occurred. In particular, earthquakes in 1898,1906, and 1907 (Stover et al.,1979) located within the Cleveland metropolitan area, as well as earthquakes in 1932, and (between 5 and etivity began to 1940, about 50 km south of Cleveland (Figure 1), are possible examples. However,
)rds of pressure in view of the large number of earthquakes reported prior to the initiation of a long, narrow solution mining, and the apparent occurrence of at least some earthquakes in ismogenic zone. northeastern Ohio beyond the range of expected influence from mining operations, it seems reasonably clear that at least some of the earthquakes are natural and that rently sufficient actures. No hy- solution mining is not a necessary condition for the occurrence of earthquake
-untain Arsenal.
activity. Moreover, the relatively low injection pressures involved (typically less than 50 bars), the shallow depth of the mining operations (less than 1 km and above it the bottom of ection increased several impermeable shale layers), and the fact that all solution salt mining operations within Lake County have long since ceased, argue against solution the overburden mining having significantly elevated pore pressure in the Precambrian basement, M%n was 269 or having triggered the 31 Janauary earthquake. {
I j,pparently regeryotr High pressure waste disposal operations iented fractures {
Three high pressure fluid injection wells that penetrate into basement are cur-rently operating within Lake County, Ohio and are within 15 km of the epicentral
{
ked increase in
\s many as 80 region of the main shock. One is an oil field brine disposal well (SALT) located njection well in within the Painesville township (Figure 5)in an area where considerable solution at/ month. Top- salt mining had previously occurred. Total volume of fluid injected into this brine 55 bars, or only well prior to 31 January 1986 amounted to about 5 million liters at a top hole g fractures with presure of 55 bars (800 psi) (Ohio Division of Oil and Gas, written communication, tents conducted 1986). The other two deep injection wells are much more likely candidates for
,micity prior to possible earthquake triggering in view of their injection history and length of
- injection, and operation. These two 1800 m deep wells are located near Perry, Ohio (CHf1 and CHp2, Figure 5) and are operated by Calhio Chemicals Division of the Stauffer are below about Chemical Company to dispose of waste products from the manufacture of an tions, elevating agricultural fungicide. The first of these wells, Calhio #1, was completed in 1971 Natural Resources Management Corp.,1971). Full scale injection of waste into the ausly quiescent well began in 1975. A second well, Calhio #2, was completed in 1981 and has been
l l
204 C, NICHOLSON, E. ROELOFFS, AND R. L. WESSON More than 1.19 billion liters (315 the two wells, principally into Calhio #1 (Figure 9) bet ected into million gallons) o
~
ween March 1975 and typical injection rate of 320 liters essures 112 bars top. hole pressure.
/ min at a (85 gal / min) hav reached a maximum of gg, wourn i Although the distance from the Calhio wells to the 31 January Calhio#1
! km, Figure 5) is greater than the corresponding earthquake (12 distances j in eithe th D Dale cases, the total volume of fluid injected into the Calhio! well r e enver or involved are significantly greater. Thus, insorder to assess the and the pressures E d Qig injection activities could have influenced earthquake egree to which fluid activity in h
it is necessary to assess the current statenortheastern of stressOhio, at the botto F calhio #2 <
ann. wou .
the hydrologic properties of m of the wells and to thece, by reservoir examining l into which flu s e ng injected. i ann u Estimation of the state of stress ^d i
i
-- 5
'"d g!
The principle sources ofinformation about existing crustal stre %me i the calhio 92 q of the wells, or in adajcent regions, are: measurements of the instap t sses at the bottom h pressure (ISIP) made during commercial hydrofracture n aneous shut inoperati: "'Q"d"*"#
ons; breakdown
- D.nwa tM pressures sneasured during well stimulation; fracture reopening prcombination.
essures; and focal
! or roechanism orientations of nearby earthquakes,in the case of Lak L = un8ile C sti data from all three injection wells, as well eas the ounty, c y, can be used to recent Ohio, seismi it W*d '
set bounds on each of the three principal stresses Estimates character of the state of stress are available on the regional from hydrofracture stress m several cases, made in Michigan and in western New York (Halmson,1978; measurements Hickman hydrofracture tl State ofstress at bottom ofinjection wells.s eorprincipal Table a .,1985)-4 lists relevant value f stresses available from both existing well data and regional compilati made after ex -
yieldan overe:
uncertainties exist for many of these values particularly the maximu ons. Large compressive stress), Inainlyi because commerc(al m horizontal measurernents to beare more ill ap;'
analysis and because, in nearly all cases, some -suited assumptions for this : int Initial value extrapolations of the existing data had to beerpretations, ,
made toordetermine the v lExtrapo!
bars.
(Michigan ano The preferred values listed at the bottom a uesofcalculated.
the table are not simpl available opinion calculations as to the for that particular parameter, most likely estimate.
y averages of all but representHickman et at our considered Formation p.
apart at the tw known. Density logs taken in theoverburden 8 Calhio is wells indicate an a pump test (init gm/cm throughout the Paleozoic section (Natural Resources Mverage densityand Maynardviof 2.6 This implies a gradient of 0.255 bar/management,1971). or 460 bars at the bottom pressure of th as met identical values of overburden stress were through similar materials (Haimson,1978)n a deep Michigan hole drilled e well measured Nearly i Calhio #1 well 4 (found predomi:
fluid injection ir (bottom of the wells) can be estimated from ISIP 4
a eozoic section hydrostatic if th.
was hydrofractured. This measurernent is valid if the fractuorded while 5). each of the well res produced wells were vertical and propagated parallel to the maximum hu i From the focal in the r zontal compressive compressive stres bottom of the well (BHP) because, g s value although to the most of the the previously de. we fresh water, other material in the injected fluid (acid with re stimulated sand formation breakd density by an unspecified amount. To, salts, etc.) raises simplify its matters a sta dvalues d that range n ar value of 180 bars is assumed information was available fortothe correction indicate a s differ to the bottom 0m unless to ofbanthe corrected 100 the tens Estin well ent value was more approp)r,iate. In attempts to inter'
~
n
SON THE NORTHEASTERN OHIO EARTHQUAKE OF 1/31/86 205 services, Inc.,1980). TABLE 4 e been injected into STRESS AND Pnzasuns ESTIMATES (IN BARS) AT A DErrH OF 1.8 Kw n March 1975 and ,
, n - ,:su=== ramm Peneman,t
~
etion pressures at a s, s., s,: r-u.u%
- hed a maumum of Mich (Haimson) 464 344 603 Wdro New York (Hickman et al) 441 370 570 I
Calhio il (initial) 461 305 tary earthquake (12 488-514 170 187 l
ither the Denver or Calhio i1 (fmal) 340 291 529-655 194 213 18 and the pressures Calhio #2 (initial) 462 320 463-527 195 199 egree to which fluid northeastern Ohio, Calhio f2 (fmal) 346 291 504-567 202 206 Brine wou (initial) 459 271 i of the wells and to ang h eramining Brine well (final) 295 267 ng injected. Adopted value 460 300-320 460-560 200 290
- Measurement made in the Calhio il wou were conducted on 11 April 1971. Measurements mad Jat the bottorn the Cathio "f 2 wou were conducted on 20 August 1979. Measurements made in the Brine w g[gp 6tantaneous shut in ations; breakdown
- Derived from fr.cture breakdown (initial) and fracture reopening (fmall pressuree usins variou 3ressures; and foca) combinations of pore preneure and ISIP, e.g., Pu = 3 (S.) - Sn - P. + T where P. = pore pressure an
- 7. = tensilutnnc h.
> ake County, Ohio, I Measund after standard slug test (initial) and uPon final equilibrium.
- ity, can be used to
!s on the regional ress measurernents several cases, values for the ISIP are measured both early and late into the
- kman et al.,1985). hydrofracture procedure. Table 4 lists both measurements. Since measurements values forprincipal made after extended pumping of the materials used in commercial operations often
)tapilations. Large yield an overestimate of the least horizontal stress, initial values ofISIP are assumed to be more appropriate.
l 1ximum horizontal re ill suited for this Initial values of ISIP corrected to the bottom of the wells range from 271 to 320 interpretations, or bars. Extrapolations from down hole measurements made at regional distances e values calculated. (Michigan and western New York) range as high as 370 bars (Haimson,1978; iply averages of all Hickman et al.,1985). The preferred value is taken to be 300 to 320 bars.
ant our considered Formation pore pressure was measured directly during drill stem tests about 8 yr apart at the two Calhio wells. Table 4 lists values measured just after the etandard g/erburden is pump test (initial) and upon reaching equilibrium (final) in both the Mt. Simon
/
uensity of 2.6 and Maynardville formations. Both sets indicate a change in the formation pore anagement,1971). pressure as measured in the Calhio #2 well, since extensive pumping began in the of the well. Nearly Calhio #1 well 4 yr earlier (1975). This apparent increase in pore pressure with time
- higan hole drilled (found predominantly in the Maynardville) is consistent with calculated effects of fluid injection in the adjacent well. In any case, the values obtained are all close to Paleozoie section hydrostatic if the density of the connate water is assumed to8 be 1.2 gm/cm (Table 5).
. each of the wells
- produced in the From the focal mechanism solutions of the earthquakes, the maximum horizontal
, ontal compressive compressive stress (Su) is at or above the vertical stress (i.e., i:460 bars). Based on this value to the the previously derived values of Su and pore pressure, estimates of Su derived from e stimulated with formation breakdown pressures during well stimulation in the Calhio wells give ts, etc.) raises its values that range as low as 463 bars (initial values, Table 4), but they need to be lard value of 180 corrected the tensile strength of the rock, revising these estimates upwards from 40 (1810 m), unless to 100 bars. Estimates of the maximum horizontal compresalve stress made from e appropriate. In atteropts to interpret the putnping records for fracture reopening pressures are i
I ~
1 1
l l 206 C. NICHOLSON, E. ROELOFFS, AND R. L. WESSON l,
l TM2 5 I
I PHYS! CAL PRoPr.RT!r.s oF Rt.sEnvoin Rocxs INTo WHICH WASTE ls BEINo INJEcrED May ric ML sunon Depth to top (m) Calhio il 1667 1806 Calhio f 2 1672 1811 Permeabilhv (darcies) Calhio i1 4.2-5.4 x 10-** 2.9-5.5 x 10-** 7 Calhio #2 2.1 x 10-81 5.5 x 10-**
Hydraulic conductivity (m/sec)
Calhio A1 4.2 x 10-* 5.5 x 10-* to-Calhio #2 2.1 x 10"* 5.5 x 10-* {
i Thicknew tm) Calhio il .52.7 a 37.8 -
Calhio #2 52.4 35.7 Specine gravity of connate Calhio #1 'I"'
1.194 1.158 brine Calhio f2 y
1.213 1.143 j Transminivity (m'/see) Calhio i1 Calhio #2 2.2 x 10-*
1.1 x 10-*
2.1 x 10-*
2.1 x 10-*
, y Porosity g Calhio #1 87e
- 8.5% t 2 {
Calhio #2 2-4% t 9%$ ,
Minimum storativity Calhio i1 1.25 x 10-* 9.54 x 10-* l Calhio #2 6.25 x 10-* 1.11 x 10-*
Other values e.saumed are: density ofinjected fluid = 1.05 g/cm*; fluid compressibility = 3.03 x 10-"
cm'/ dyne iNatural Resources Management Corp.,1971; Resources Services, Inc.,1980). I
- Drill stem test. 0 j "5 t Core sample.
1
% ,l
$ Welllog.
listed in Table 4 as final values. Measurement derived from well records made i during the stimulation of the brine well near Painesville are suspect, since the hydrofracture procedure was conducted through perforated casing (Petro Evt.luation Services Inc.,1985). Thus, of all the measurements, the value of the maximum compressive stress is the least well known. For our purposes, estimates of the 2
maximum compressive stress based on a lower bound (i.e., the vertical stress of 460 bars) are useful, as they would represent conservative estimates on how close to $ in .
failure conditions are at the top of the basement.
The Mohr. Coulomb failure criterion. Using the adopted values given in Table 4, 5 it appears that without fluid injection, the conditions are near but do not exceed failure at the bottom of the wells. Figure 10 is a graphical representation of the i state of stress and the Mohr. Coulomb failure criterion (c.f., Jaeger and Cook,1976; Simpson,1986) at a nominal depth of 1.8 km. In the presence of a fluid, the effective s stress levels are reduced by the amount of the formation pore pressure which moves y
j the Mohr etrcle to the left toward the failure envelope (middle circle, Figure 10), a This condition is close to, but does not exceed the failure criterion for a fracture with no cohesion. Injection at a nominal pressure of 110 bars, however, would bring the zone immediately surrounding the well bottom to an effective stress state near critical for favorably oriented, preexisting fractures having cohesive strengths of as 0 much as 40 bars and a friction coefficient near 0.6 (left circle, Figure 10). However, as the overburden pressure is only a lower bound for the estimate of the maximum ("5,, ,/,'l compressive stress, the actual conditions for failure at the bottom of the wells may be more critical than the situation shown. Fic. 9. Volume a tion,1986).
State of stress in the hypocentral region. The 31 January earthquake and nearly allits aftershocks were located about 12 km from the wells, and at depths of frorn
77 . --+*r-,_-- _
t ON THE NORTHEASTERN OH10 EARTHQUAKE OF 1/31/86 207 CALHIO #1
' I l 1 I i 1 l l l l t j het sum 1806 1811 O MONTHLY l g i 5.5 x 10-** f g g
- 1.0
{
5 5.5 x 10-* g 10 -
5.5 x 10-* ,
l
(
37.8 35.7 7 m
/ -
l E / i 1.158 2 g
1.143 y ,
2.1 x 10-* $ 8 2.1 x 104 2
g l
.5% t ,
g 8%% CUMULATIVE 3.
9.54 x 10-*
1.11 x 10-*
.rweibility = 3.03 x 10-" I g e.,1980). 0 o
,i... ,i.,. ,i., j ,, , ,i., ,,,,,
i,...,,,i,,,., ,i . , i i, i.. i,..
TIME CALHlO #2 t well records made l l l l l l l l l ; ;
e suspect, since the 1g(Petro Evaluation ue of the maximum P es, estimates of the a !
vertical stress of 460 i jo , E
-10 E' ttes on how close to g $
ie 'ven in Table 4, E. g
(
i o not exceed 7 tation of the E MONTHLY 'l eger and Cook,1976; so 4
-g
'a fluid, the effective p l ressure which moves E j ( .$
le ' circle, Figure 10). $ [ l terion for a fracture 4- 5 CLMU.a ave towever, would bring
- ive stress state near / Q.N/f 1 ,f .
esive strengths of as o /rnT'09 Y ( lAN Figure 10). However, ',' ' 5,,
l
', i,','s , ' 8 8 8ie's2 l
ists lissa iisi l l..
i i ii.3 li ... l ,,, i 1,..
3 ate of the maximum T:ME
.om of the wells may Fic. 9. Volume of fluid injected into the Calhio wells through time (Ohio EPA. written communica.
tion.1986).
rthquake and nearly ad at depths of frorn
- e
l l
\
y
ll l l
s 208 j C. NICHOLSON, E. ROELOFFS, AND R. L. WESSON t = 40 + 0.6 o n 1 Maynardville t = 0.6 o n Porosities and m- and core samp
'ative transmis storativity, wh Hydrostabc No fluid pressure j 8 unit decline I -
I / / I i
j where p is Duic' :
a of the aquifer, a soo j having the thic c (ears) a porosity of 0
' The highest po B, producing vt injec:,on pressure of 11e ears ,
For purposes average fluid ir i liters (310 mill j (1.8 km 1985 (Figure 9 depth).Fic, to. Mohr circle diagram showing inferred state of stress at bottom of the 1:dection wells month. Becausa the distance frc :
a single point i; 2 to 6 km. Estimation of the existing state of stress at such increased hypocentral adjusted so as depths is difficult because simple extrapolation of principal stress cornponents to depth is not theoretically justified, nor is there any indication that the stress ratio maximum valut determined for the lower Paleozoic sedimentary section is maintained into the for a period of investigate the i Precambrian granitic basement. What few measurement 4 do exist for the basement (Michigan and Northern Illinois) indicate that the stress ratio of the minimum Infinite resert
! t as a result of a compressive stress over the vertical stress is between 0.72 to 0.77 (Haimson,1978; directions is giv.
Halmson and Doe,1983) Such high stress ratios imply small stress differentials and consequently, stress conditions that are not as close to failure.
Fluidpressure changes resulting from injection Estimates of the fluid pressure changes near the earthquake hypocenters are difficult to determine because little is known about the hydrologic properties of the in which u = r2 i basement where the earthquakes are actually occurring. The characteristics of the for a hypothetier
~,) coefficients, req'.
reservoir in the vicinity of the wells, however, can be estimated from measurements made during well completion. Using these characteristics, two types of reservoir tively,in order t-models were evaluated in order to determine what the increase in fluid pressure 12 yr of operatic near the earthquake hypocenters may have been as a result of the high pressure head, are shown fluid injection. The first type of model is an infinite isotropic reservoir; the second Although the pre involves reservoirs of finite width (i.e., rectangular cross.section), but of infinite the pressure at e length, extending in the direction connecting the wells and the hypocenters. These tivity yielda a pri models are for the purposes of studying how fluid pressure may have propagated upper bound for l
! horizontally away from the wells and do not address the question of how pressure distance of 12 kn effects could have migrated downward from the injection horizon to hypocentral- Note that in tt depths (Figure 5, bottorn). from the well (br Reservoir properties. For a given reservoir geometry, the fluid pressure field This result confi generated by injection is governed by the reservoir's transmissivity and storativity, time history of f.
In the case of the Calhio wells, waste fluid is injected into the essentially flat lying being induced by immediately after
\
j
, Ey,' .,f \ ,J h 6 7 '
ll 4
l .
.T q z -,,s a . , : a. . . . ~- # -
" - ~ ~
~
~ = m ~
E : ,,n.,; f,W .,,
209 THE NORTHEASTERN OHIO EARTHQUAKE OF 1/31/86 Maynardville and Mt. Simon formations at the base of the Paleozoic section.
Porosities and permeabilities for the various formations, based on drill stem testa and core samples, are given in Table 5. Combining the two formations, a represent-ative transmissitivity for the entire zone of injection is 4.2 x 10-* m'/sec. The storativity, which gives the amount of fluid released per unit column* of aquifer for a unit decline in head, can be calculated from the expression S = pgh(a + n#) (1) where p is fluid density, h is the aquifer thickness, a is the vertical compressibility of the aquifer, n is the porosity, and # is the fluid cornpressibility. For a formation having the thickness of the Maynardville and Mt. Simon formations combined and a porosity of 0.08, setting a = 0 in (1) yields a minimum storativity of 2.2 x 10-8 The highest possible estimate of storativity corresponds to values of a greater than 1
[.% S, producing values for the storativity as high as 2.0 x 10*. '
For purposes of calculating pressure 12 km from the well, it suffices to use an average fluid injection rate. The total volume injected into both wells is 1.17 billion liters (310 million gallons) during the period from March 1975 through November 1985 (Figure 9), corresponding to an average injection rate of 9.0 million liters /
month. Because the distance between the wells (about 800 m) is small compared to
.aion nu (1.8 km the distance from the wells to the hypocenter, the two wells have been modeled as a single point source of fluid. For a given storativity (S), transmissivity (T) was
( tsed hypocentral adjusted so as to maintain the injection pressure at the well below the known l maxirnum value of about 110 bars. In the models tested, injection was continued l l 4 cornponents to for a period of 12 yr (i.e.,1975 to 1986), after which flow was stopped in order to l l t the stress ratio investigate the effects of ceasing injection.
l atained into the Infinite reservoir model(radial flow). The pressure p(r, t) at distance r, and time for the basement t as a result of a constant flow rate Q into a reservoir that extends uniformly in all l of the minimum (Haimson,1978; directions is given by the equation l
ress differentials (2) p(r, t) = 4x 2 . ( d(
in which u = r S/4Tt 2
(e.g, Freeze and Cherry,1979). Evaluation of equation (2) b enters are for a hypothetical well of radius 12 cm, and upper and lower bounds to the storage l properties of thi coefficients, required transmissivities of 5.2 x 10-4 and 5.8 x 10-8 m'/sec, respec-i racteristics of the am measurements tively,in order to match the constraint of 110 bars injection pressure for the entire types of reservoir 12 yr of operation. The results for these two cases,in terms of pressure at the well in fluid pressure head, are shown in Figure 11A as the two very similar curves labeled " infinite."
the high pressure Although the pressure at the wellis not sensitive to the range of storage coefficient, ervoir; the second the pressure at epicentral distances (Figure 11B) is affected. The minimum stora-l n), but of infinite tivity yields a pressure increase of 3.6 bars 12 yr after infection initiation, but if the ypocenters. These upper bound for the storage coefficient is assumed, the pressure increase at a
/ have propagated distance of 12 km is only a fraction of a bar.
n of how pressure Note that in the hypothetical case that injection is stopped, pore pressure 12 km on to hypocentral from the well (based on the radial flow model) continues to rise for at least 2 yr.
This result confirrns the insensitivity of pressure at this distame to the detailed uid pressure field time history of fluid injection at the well, and implies that if seismic activity is ity and storativity. being induced by elevated pore pressure frorn the wells, it will not be suppressed sentially flat. lying immediately after ceasing injection.
4-THI 210 C. NICHOLSON, E. ROELOFFS, AND R. L. WESSON Infinite strip re trending from the ;
stance from 6 j INJECTION PRESSURE AT WELLHEAD model was used by l
!1!IIIIIIIIIIIIIIIIi11II tion around the i 11o - .i 1 INFINITE (hight) evidence that the 1 ' are atill useful, h-8" - 2 ,
- 2. INFINITE (!ow T) 3 ' ! epicentral distane eo - 3 7.5 km wide STRIP history at the we,
"~ 7 g 4 1.0 km wide STRIP injected. t
[ For injection it l E direction, a const. I tu ',o
- 4 c
D l -
.. m so - 4 M
y 6e ,
l
- c. ,: .
/ where u. = (z' '
3 strip. Figure 11 s i l 2 distance of 12 kr ie ' reservoir models ,
l A equal to 6.0 x 1 l
' I n2n*n- i n6 t.h i oo i nn i nh I nnTIME inh l i eh nI 6 inh l ieh lie h i n., transmissitivity -l pressure at the
radial flow mode to a higher value ;
PORE PRESSURE CHANGE 12 KM FROM WELLBORE 36 bars 12 yr af ;
ceases. Figure 1: ,
,, I I I i i i i I i I i I I i i I I I I I I i 1 I l can be used to .
injected. The ac :
g ,, ; ' 4 Calhio #1 and # '
M 1986), inore clo-6 - continually rapi 88 -
w a:
oM to ;
One of the mc :
. 8 _
so few events. I'
@ 3, ;' 3 reported by 15.
in the Eastern 1I iB _ m p f g g g y ,y ,,,,,,,,,',,,,,,,,3 detected in the tik Inh 'uh 'nh ni s Ina TIME i m vs. oo i nh Ink In6 l ieh hek inh l isse Brown and Eb, .
hand, several et l Fio.11. Pressun versus time at wellhead (A) and 1x km from weu ovie (B) for both radial flow a remarkably 8 (infinite width) and finite width reservoir models. An injection rate of 9 million liters / month is used and the m = 5.5 ea' is assumed to cease after 12 yr. Curves labeled 1 and 2 represent transmissivities of 5.8 x 10-* and 5.2 x LO'*m'/see, respectively. Heavy solid line in ( A) represents actual maximum pressures used at the Calhio 1970), and the wells (Ohio EPA, written communication.1966). (B) The lower, hochured arco is bounded above and 5.0) had only t below by curves obtained using the tsdial flow model, and maximum and minimum allowable storativities.
respectively. The large, shaded area is bounded below by the pressure obtened using an infinite strip aftershocks .is -
model 7.5 km wide (3h and above by that obtained using an infinite strip 1 km wide (4). quakes in the e been triggered The 31 Janu
_ + mm 85 %,
THE NORTHEASTERN OHIO EARTHQUAKE OF 1/31/86 211 In/inite strip reservoir model. If fluid flow is confined to a narrow reservoir l trending from the wells to the hypocentral region, then the pressure at a given distance from the well will be higher than for the radial flow models. This type of f model was used by Hsieh and Bredeboeft (1981) to calculate the pressure distribu. '
.1 ! I tion around the Rocky Mountain Arsenal well implicated in the 1960's Denver earthquake sequence. In the case of the Ohio earthquakes, there is no independent i nigh T) evidence that the injection zone has a long, narrow configuration. The calculations I iow T) are still useful, however, in that they illustrate how large a pressure buildup at epicentral distances is possible and because they show how diagnostic the pressure se smip history at the well bore is of the shape of the reservoir into which fluid is being
- se STRIP injected.
For injection into the center of a strip of width w and infinite extent in the x i direction, a constant injection rate Q produces a pressure given by !
p(x, y, t) = 4 w T ..f., . ,,,.
d( t (3) where u,,, = (x' + (y + mw)')S/4Tt, and y is the distance from the center of the strip. Figure 11 shows pressure versus time at the well bore and at the epicentral distance of 12 km, respectively, for two infinite strip reservoir models. Both strip 1 reservoir models have storativities of 5.4 x 10-8 m'/sec. One has a transmissivity 2
, y q,,, equal to 6.0 x 10-' m /see and a width of 7.5 km; the other model has a higher transmissitivity of 2.5 x 10-8 m'/sec and a width of 1 km. For the wider strip, pressure at the epicentral distance is comparable to that for the low storativity radial flow rnodel, except that when injection stops, the pressure continues to rise
!!LLBORE to a higher value. For the narrow strip, pressure at the epicentral distance is about 36 bars 12 yr after beginning injection, and also continues to rise once injection I 1 1 I I ceases. Figure 11A dernonstrates that analysis of the history of injection pressure can be used to discriminate the shape of the reservoir into which fluid is being injected. The actual increase in injection pressures with time, as measured in the Calhio # 1 and #2 wells (Figure 11A, heavy line, Ohio EPA, written communication, 1986), more closely resernbles the radial flow model and is inconsistent with the continually rapid increase in injection pressure required by the narrow strip case, DISCUSSION One of the more notable features of the aftershock sequence was that it contained so few events. No aftershocks were detected in the first 26 br, and only 13 were
~
. 3 reported by 15 April. This contrasts with other recent, moderate size earthquakes 2 in the Eastern United States in which several tens to hundreds of aftershocks where
'"""7.
' f'".h ni h ins detected in the weeks to months following the main shock (Herrmann et al.,1982; Brown and Ebel,1982; Wetmiller et al.,1984; Seeber et al.,1984). On the other hand, several earthquakes in the Central and Eastern United States have exhibited a remarkably similar lack of aftershock activity. Four aftershocks were found for fnN.*/ min'tNe'h.' Man't the m = 5.5 ea:thquake in southern Illinois on 9 November 1968 (Stauder and Pitt, ties of 5 8 x 10"* and 5.2.x
","j,"d,'j '.N,Yni 1970), and the earthquake near Marked Tree, Arkansas, on 25 March 1976 (m. =
mm anowable stontintia. 5.0) had only two (m. - 4.5 and m u = 1.5) (Stauder et al.,1976). The lack of Nj%'" " " aftershocks is thus neither discriminatory nor characteristic of noninduced earth quakes in the east, but is atypical of other earthquake sequences considered to have .
I been triggered by fluid injection (e.g., Talwani and Acree,1986).
I The 31 January earthquake and all of its immediate aftershocks are rather tightly l 8'*.
y l J
i 212 C. SICHOLSON, E. ROELOFFS, AND R. L. WESSON THE clustered about 12 km south of the deep injection wells near Perry (Figure 5, top); that at the bottom '
howeser, there was one small microcarthquake located close to the wells on 12 by a few tens of be March. Its focal depth of 2 km corresponds to the base of the Paleozoic section and judged critical on ti is the same depth at which fluid is injected from the Calhio wells (Figure 5, bottom). these stress estims Small earthquakes in January and November of 1983 may also have occurred in ments made at the close proximity to the Calhio wells and at shallow depths. Since the detection which seismicity et threshold for earthquakes near the wells prior to the installation of portable necessarily be cons equipment following the 31 January earthquake (relying on the seismograph at The state of stre John Carroll University) is estimated to be somewhat greater than magnitude 2.5, well operations cou it is conceivable that additional small earthquakes could have occurred near the small shallow eartt wells between the initiation of injection and the 31 January earthquake. within the upper 2 The tirne lag between the onset of the 1986 sequence and major injection well pressure increase (
operations in Lake County (1975) is unusually long for typical cases of induced location in the Pree seismicity related to deep fluid injection, but is similar in some respects to several the lack of large a well documented cases of earthquakes triggered by reservoir impoundment, partic- remains elevated, a ularly in those cases where the pressure front migrated a considerable distance prior prior to the initiatic to the onset of seismicity (Simpson,1986). Further delays might be expected for January earthquak2 diffusion of pore pressure effects down to hypocentral depths. Permeabilities, as ability that fluid is measured by pumping tests in the deep Ohio wells, are consistent with the observed earthquake must b time delay. Analysis of available stress measurements seems to indicate that the Regardless of w}
state of stress in northeastern Ohio is close to the theoretical threshold for small occurrence of the'e4 earthquakes as predicted by the Mohr-Coulomb failure criterion. This should not conditions for earth be surprising given the history of small to moderate earthquakes in the region, western New York (
Since high pressure fluid injection could have brought at least the region near the et al,1986), that su<
bottom of the wells into a critical stress state, the absence of large numbers of small region that previou earthquakes in the immediate vicinity of the wells suggests that there are not many activity associated e favorably oriented, weak fractures near the wells. Thus, either existing fractures attention. Plans ar@
have cohesive strengths greater than 40 bars, or if weaker fractures do exist, they land Electric Illumii are not favorably oriented for failure in the existing stress field. The predominant County area. Shouk dip of fractures observed in a core taken from the injection zone in Calhio well #2 and the 31 January is 20* (Resource Services Inc.,1980) Such fractures would not be as favorably ranted.
oriented for failure, as shear stress is maximum only for near vertical faults. Thus, the tirne between the initiation of injection (1975) and the 1986 earthquake could be related to the time required for pore pressure effects to migrate out to an area This study was initid with more favorably oriented fractures. Nuclear bactor bgulati authors (C. N. and E. R.
\ The actual pressure elevation in the hypocentral region as a result of the injection ' hip' Thi' P*Per bene 8 operation is certainly no more than 40 bars and more likely only a few bars.
Although previously well documented cases of seismicity induced by fluid injection proh**$y',*,N,*,'g typically involve pressure increases of several tens of bars, cases of reservoir-induced without whien this paper <
Glassmoyer, Chuck Mua seismicity indicate that changes in water height of only a few meters, corresponding discovering the microeaa to pressure changes of less than a bar, have triggered substantial numbers of small P"'id2"8 *** '* th' 8 earthquakes (Simpson,1986). This result, combined with our lack of knowledge
- regarding the state of stress in the hypocentral region, makes it difficult to assess Ite)of,'Nb r, the minimum size of pore. pressure increase sufficient to haved triggered the 1986 collected by the Lamont <
main shock focal mechana earthquakes.
by TEIC: Charles Langu CONCIESIONS intensity results, locatioa Hansen of the Ohio Geoh With our present information, it is not possible to confirm or reject the hypothesis that injection of waste into the Calhio wells triggered the 31 January earthquake and its aftershocks. If the state of stress in the hypocentral region is comparable to "3*
p tor br N
information en their operc
. g d, ,
C~- , f.,;; . . .W, - W;;giw.# - . nr.,,1rggrqcw g
, ',',;" -Q;@Qc Q d 1 7 } %,, j w e Q{.:
- )N THE NORTHEASTERN OHIO EARTHQlJAKE OF 1/31/86 213 rry (Figure 5, toph that at the bottom of the injection wells, then it appears that elevating the pressure
- o the wells on 12 by a few tens of bars would have resulted in a state of effective stress that would be
,leozoic section and judged critical on the basis of the Mohr Coulomb failure criterion. However, because (Figure 5, bottom). these stress estimetes are uncertain, and because they are not based on measure-o have occurred in ments made at the hypocenter, it is not possible to specify a level of pressure below ince the detection which seismicity could not have been triggered, or above which earthquakes would ilation of portable necessarily be considered induced.
he seismograph at The state of stress in the lower sedimentary section is such that deep injection aan magnitude 2.5, well operations could have elevated formation pore pressures sufficiently to trigger occurred near the small shallow earthquakes, and two or possibly three small earthquakes did occur hquake. within the upper 2 km and in close proximity to the wells. However, the small injor injection well pressure increase calculated for the hypocentral region of the main shock, its il cases of induced location in the Precambrian basement (where the stress regime may be different),
respects to several the lack of large numbers of small earthquakes while formation pom pressure aoundment, partic*
remains elevated, and the history of small to moderate earthquakes in the region ;
rable distance prior prior to the initiation of injection all argue for a " natural" tectonic origin for the 31 ;
- ht be expected for January earthquake. Therefore, although triggering remains a possibility, the prob-Perrneabilities, as ability that fluid injection played a significant role in triggering the 31 January b the observed earthquake must be regarded as low.
hate that the Regardless of whether the 31 January earthquake is conaidered induced, the bnold for small occurrence of the earthquake itselfimplies that the regional state of stress is near
- n. This should not conditions for earthquake generation. In view of this situation, and of results in Ikes in the region. western New York (Fletcher and Sykes,1977) and in southwestern Ontario (Mereu the region near the et al.,1986), that suggest induced seismicity may be more prevalent in the northeast te numbers of small region that previously had been supposed, the possibility for future earthquake there are not many activity associated with injection operations in this region should be given continued
- existing fractures attention. Plans are currently underway by John Carroll University and the Cleve-
.ures do exist, they land Electric Illuminating Company to continue monitoring seismicity in the Lake
. The predominant County area. Should additional activity occur near the wells, or between the wells e in Calhio well #2 and the 31 January earthquake, further examination of this issue would be war-ot be as favorably ranted.
ertical faults. Thus, 6 earthquake could ACKNOWLEDGMENTS rate out to an area This study was initially undertaken on behalf of, and with financial support from, the Office of Nuclear Reactor Regulation, U.S. Nuclear Regulatory Commission. This work was done while two of the sult of the . .mjection authors (C. N. and E. R.) held National Research Council U.S. Geological Survey Research Anociate.
ships. This paper benefited greatly froni discussions with John Bredeboeft, Keith Evana, and Leon V only a few bars. Reiter. The authors acknowledge and appreciate the cooperation, contributions, and exchange of data uid injection provided by a number of other groups involved in studies of the 31 January earthquake and its aftershocks.
(% 'oir induced I.et corresponding without which this paper could not have been prepared. Particularly helpful were: Roger Borcherdt, Gary Glassmoyer, Chuck Mueller, and Carlos Valdes for access to, and help with the GEOS data, and
&scovenng the microcarthquake of 12 March: Richard Holt, Gabriel LeBlanc, and Preston Turner for il numbers of small provi&ng accen to the considerable data set collected by Weston Geophysical; Tom Statton, Richard lack of knowledge Quittmeyer, and Kathy Mroteck for generous access to the data collected by woodward.Clyde Consult.
t difficult to assess ants; John Armbruster, leonardo Seeber, and David Simpson for discussions and arrival time data triggered the 1986 collected by the Lamont group; Robert Herrmann and Steve Neiers for providing their analysis of the main shock focal mechanism and data collected by St. louis University;Jer Ming Chiu for data collected by TEIC; Charles Langer, Margaret Hopper, Jim Dewey, and Russell Needham for arrival time data, intensity results, location, and teleseismic first motions of the main shock: Horace Collins and Mike
. Hansen of the Ohio Geological Survey, and Dennis Crist and John Gray of Ohio DNR for information nect the hypothes.ts on locai and regional subsurface geology, and data on the brine well in Painesville; Gerry Meyers of Obia anuary earthquake EPA for bringing to our attention the waste disposal wells in Lake County and providing considerable an is comparable to information on their operation; Warren Latimer and Jay Henthorn for providing well completion reports; 3
THE C. NICHOLSON, E. ROELOFFS, AND R. L. WESSON 214 Lahr, J. C. (1985). H' William Toth, Jim Rhodes, and Joe Fischer for ad&tionalinformation on the Calbio wells Dale Wedge bypocentral parae l and Pat Linn of the Geauga County DSA for damage reporta, and locations cf the disturbed shallow 1 519,35 pp.
water wells, and Pradeep Talwani for a preprint of this manuscript. We thank Jim Dewey, Dave Mereu R F., J. Brunet. ,
Oppenheimer, Paul Segall, and an anonymous reviewer for their careful and thoughtful reviews of the . the Gobles oil field '
manuscript. ! Musman, S. A. and T.
E strains (abstract), f REFERENCES $ Natural Resources Mar 2
subsurface disposa!
Borcherdt, R D. (E6 tor) (1986). Preliminary report on aftershock sequence for the earthquake of January 31,1986, neat Painesville, Ohio, U.S. Geol Suru., Open file Rept.66-181,109 pp.
I NEIC (1986). Prelimina '
!, Bulletin,24 pp.
Borcherdt, R D. and G. Glassmoyer (1987). On the aftershock sequence for the earthquake of January Nuttli, O. W., W. Stauc 31,1986 in northeastern Ohio: effects of bandwidth and local geology on observed high frequency [*
United States, Ear:
ground motion, Paper Presented at Earthquake Groundmotion Estimation in Eastern North i Petro Evaluation Servi America. EPRI, March 31-April 2 Palo Alto, California,46 pp. Lake County, Ohic Borcherdt, R. D., J. B. Fletcher, E. G. Jensen, G. L Maxwell, J. R VanSchanck, R. E. Warrick, E. Press, F. (1966). Seisrr Cranswick, M. J. S. Johns;on, and R McClearn (1985). A General Earthquake Observation System Am. Mem. 97,195 I
(CEOS), Bull Seism. Soc. Am. 7 5, 1783-1823. Raleigh, C. B., J. H. He.
Brown. E. J. and J. E. Ebel (1982). A study of selected aftershocks of the January,1982 earthquake near {. Colorado, Science ~
Laconia, New Hampshire (abstract), EarthquoAe Notes 53,14. Reasenberg, P. and D. (
\ Chaplin, M. P., S. R. Taylor, and M. N. Toks62 (1980). Coda. length magnitude scale for New England.
I for calculating anc Earthquake Notes 51,15-22. l 85J39,109 pp.
Cleveland Electric Illuminating Company (1982). The Perry Nuclear Power Plant Unita I and II: Final
- Resources Services, In<
Safety Analysis Report, Cleveland, Ohio. I well, injection We Clifford. M. J. (1973). Silurian rock salt of Ohio, Ohio Geol. Suru. Rept. o/ Investigations, vol. 9, Columbus, Rice, J. R and M. P.
Ohio,42 pp. porous media with Dewey, J. W. and D. W. Gordon (1984). Map showing recomputed hypocenters of earthquakes in the Roeloffs, E. (1987). H entern and central United States and adjacent Canada,1925-1980, U.S. Geol. Surv. Map MF.1699.
Basis for Intermes Dunrud, C. R. and B. B. Nevins (1981). Solution mining and subsidence in evaporite rocks in the United Monterey, Califor.
States, U.S. Geol. Surv. Map I 1298. Roeloffs, E. and J. D.
Dziewonski, A. M. (1986). Global seismicity of 1985 (abstract), EOS, Trans. Am. Geophys. Union 67 to the August 4 309. Union 66,986.
Evans, K. (1987). In. situ stress discontinuity across the Appalachian Plateau decollement (abstract), Root, S. L and R H. B Paper Presented at AGU Spring Meeting. Baltimore, Maryland, p. 66.
163.
Fletcher, J. B. and L R Sykes (1977). Earthquakes related to hydraulic mining and natural seismic Sbar, M. L and L R activity in western New York State, J. Geophys. Res. 8 2, 3767-3780.
America: an exan:
Freeze, R A. and J. A. Cherry (1979). Ground Water, Prentice HallInc., Englewood Cliffs, New Jersey, Seeber, L, E. Crans, 604 pp. aftershock sequer Glassmoyer, G., R Borcherdt, J. King C. Dietal E. Sembera E. Roeloffs, C. Valdes, and C. Nicholson Trans. Arrt Geop*
(1986). Source and propagation characteristics of aftershock sequence near Painesville, Ohio (ab-Simpson, D. W. (1986 stract), EOS, Traru. Aa Geophys. Union 67,314. Stauder, W. and A. B Haimson, B. C. (1978). Crustal stress in the Michigan basin, J. Geophys. Res. 83, 5857-5863. November 9,196r Haimson, B. C. and T. W. Doe (1983). State of stress, permeability, and fractures in the Precambrian Stauder, W., S. Schs granite of northern Illinois, J. Geophys Res. 8 8, 7355-7372. bulletin, vot 7, S Healy, J. H., W. W. Rubey, D. T. Griggs, and C. B. Raleigh (1968). The Denver earthquakes, Science Stauder, W., R Herrt 16 1,1301-1309.
Central Mississir Herrmann, R. B. and B. V. Nguyen (1986). Focal mechanism studies of the January 31,1986 Perry Ohio Stover, C. W., B. G.
earthquake (abstract), Earthquake Notes 57,107. Geol. Surv. Map Herrmann, R. B., C. A. Langston, and J. E. Zollweg (1982). The Sharpsburg, Kentucky, earthquake of Talwani, P. and S. At 27 July,1980, BufL Seism. Soc. Am. 7 2, 1219-1239. of January 31,15 Hickman, S. H., J. H. Healy, and M. D. Zoback (1985). In situ stress, natural fracture distribution, and Vorhis, R C. (1968).
borehole elongation in the Auburn geothermal wall, Auburn, New York, J. Geophys. Res. 90,5497-National Acaden 5512. Wakita, H. (1975). %
Hildenbrand. T. G. and R P. Kucks (1984a). Residual total intensity magnetic map of Ohio, U.S. Geol.
Weuon, R L and C.
Surv. Map G P.961.
a report to the U Hildenbrand. T. G. and R P. Kucks (1984b). Complete Bouguer gravity anomaly map of Ohio, U.S.
pp.
Geol. Surv. Map GP.962. Weston Geophysical Hsieh, P. A. and J. D. Bredehoeft (1981). A reservoir analysis of the Denver earthquakes: a case of issues: northeas-induced seismicity, J. Geophys Res 86,903-920.
Illuminating Co.
Jaeger, J. C. and N. C. W. Cook (1976). Fundamentals of Rock Mechanics, John Wiley and Sons, Inc.,
Weston Geophysical New York,585 pp.
- 2.M
~
.. --, . .. s - p ' 7, .- y . 4 . _, . . L _d '* . . ,
..-s ISSON e Gmlhio wells; Dale Wedge THE NORTHEASTERN OHIO EARTHQUAKE OF 1/31/86 215 ns of the esturbed shallow 14hr, J. C. (1985). HYPOELLIPSE/VAX: a computer program for determi e thank Jim Dewey, Dave bypocentral 519,35 pp. parameters, magnitude and fint motion pattern, U.S. Geol Suru d thoughtful reviews of the Menu, R. F., J. Brunet, K. Morrissey, B. Price, and A. Yapp (1986), A study o the Gobles oil field area of southwestern Ontario, Bull Seisa Soc. An 7 6, 1215-1223.
Musman, S. A. and T. Schmidt (1986). The relationship of intraplate mismicity to otrains (abstract), EOS, Trans. An Geophys. Union 67,307. .
nee for the earthquake of Natural Resources Management Corp. (1971). Report on the dnlims, testing and c st.86-181,109 pp. subsurface esposal well #1, Calhio Chemicals, Inc., Perry, Ohio,81 pp.
the earthquake of January NEIC (1986).
Bulletin,24 pp. Preliminary determination of epicenters-Monthly listing January,198 in observed hich frequency mstion in Eastern Nonb Nuttli, O. W., W. Stauder, and C. Kisslinger (1969). Travel time tables for earthquak United States, EarthquaAe Notes 40,19-28.
3 I
i
- chaeck, R. E. Warrick, E. PetroLakeEvaluation County, Ohio,5 Services, pp. Inc. (1985). Well completion record, saltwalter issection w ,
iquake Observation System Preu,Am.F,Mem. (1966). Seismic velocities, in Handbook of Physical Constants, S. D. Clark, E 97,195-218.
.-~M2 eanhquake near Raleigh. C.Science B., J.191,1230-1237.
H. Healy, and J. D. Bredeboeft (1976). An experiment in earthqu
)
.e for New England, Colorado, Raasenberg, P. and D. Oppenheimer (1985). FPFIT, FPPLOT and FPPAGE: Fortran Plant Units I and II: Final for calculating 85J39,109 pp. and displaying earthquake fault. plane solutions, U.S. Geol Suru., Opeq
.tgarions, vol. 9, Columbus, Resources Services, Inc. (1980). Report on the drilling, testing and completion of the su well, injection Well #2, Calhio Chemicals, Inc., Perry, Ohio,84 pp.
.ters of earthquakes in the Rice, J. R. and M. P. Cleary (1976). Some basic stress diffusion solutions for fluid. saturat Geol. Sury. Map MF 1699. porous mess with compressible constituents, Rev. Geophys. Space Phys. 14, 227-241.
aporite rocks in the United Roeloffs, E. (1987). Hydrologic precursors: a review, Paper Presented at Physical and Obs!
Basis for Intermediate Term Earthquake Prediction, U.S. Geological Survey Redbookl Monterey, California,34 pp. 1
. An Geophys. Union 67 l Roeloffs, E. and J. D. Bredeboeft (1985). Coseismic response of water wella near Parkfiel au decollement (abstreet), to the66,986.
Union August 4,1985 North Kettleman Hills eanhquake (abstreet), EOS, Trans. Am. Geo iining and natural seismic Root,163.
S. I. and R II. MacWilliams (1986). The Suffield fault, Stark County, Ohio, Ohio .l. Sci l
(
lewood Cliffs, New Jersey, Sbar, M. L. and L. R. Sykes (1973h Contemporary compressive stress and meismicity in eas America; an eaample ofintra plate tectonics, Geol Soc. Am. Bull 84,1861-1882.
Valdes, and C. Nicholson Seeber, L., E. Cranswick, J. Armbruster, and N. Brastow (1984). The October 1983 Goodn '
i near Painesville, Ohio (ab. aftershock sequence: regional seismicity and structural features in the Adirondacks (ashtra Trans. Am. Geophys. Union 65,239.
e7-5863. Simpson, D. W. (1986). Triggered eanhquakes, Ann. Rev. Earth Planet Sci. 1 4, 21-42.
the Precambrian Stauder, W. and A. M. Pitt (1970). Note on an aftershock study, south cenml Illinois earthquakl November 9,1968, Eull Seism. Soc. Am. 6 0, 983-986. (
over earthquakes, Science Stauder, W., S. Schaefer, J. Best, and S. T. Morrissey (1976). Southesst Missouri Regional Seis Bulletin, vol. 7, St, Louis University, St. Louis, Missouri,25 pp.
nuary 31,1986 Perry Ohio Stauder, W., R. Herrmann, S. Singh, C. Nicholson, D. Reidy, R. Perry, S. Morrissey, and E.
Central Mississippi Valley earthquakes-1979, Earthquake Notes 52,26-31.
. Kentucky, earthquake of Stover, C. W., B. G. Reagor, and S. T. Algerminen (1979). Seismicity map of the state of Oh Geol. Surv. Map MF 1142.
fracture estribution, and Talwani, P. and S. Aeree (1986). Deep wellinjection at the Calhio wells and the 14roy, Ohi I, Geophys. Res. 90,5497- of January 31,1966, A report to the Cleveland Electric Illuminating Co., Cleveland, Ohio,92 pp Vothis, R C. (1968). Eifects outside Alaska,in The Great Alaska Earthquake of 1964:
Hydrology, Part A, National Academy of Sciences Publication 1603,140-189.
c map of Ohio, U.S. Geol.
Wakita, H. (1975). Water wells as possible in6cators of tectonic strain, Science 189, 553-555.
omaly map of Ohio, U.S. Wesson, R. L. and C. Nicholson (1986). Stu&es of the January 31,1986 nonheastern Ohio ear a report to the U.S. N.iclear Regulatory Comminien, U.S. Geol Sure., Open. File Rept. 86 331,131 pp.
er earthquakes: a case of Weston Geophysical Corporation (1986). Investigations of confirmatory seismological and geo
,hn Wiley and Sons, Inc., issues. northeastern Ohio earthquake of January 31, 1986, A report to the Cleveland Electrie Illuminating Co., Cleveland, Ohio,233 pp.
Weston Geophysical Corporation (1987a). Quarterly progress report, CEI seismic monitoring pr
_6.h. -
e - - - ~ . ,
I 216 C. NICHOi. SON, E. ROELOFFS, AND R, L. WESSON > '
for northeastern Ohio, Octoger 15,1986-January 15,1987, A report prepared for Cleveland Electric Illuminating Company, Cleveland. Ohio,12 pp. !
Weston Geophysical Corporation (1987b). Second quarterly report, CEI seismic monitoring network , LAKEER, '
January land, Ohio,11 15.-April pp. 15,1987, A report prepared for Cleveland Electric illuminating Company, Cleve.
~
Wetmiller, R J., J. Adams. F. M. Anglin, H. S. Hasegawa, and A. E. Stevens (1984). Aftersbock sequenew of the 1982 Miramichi. New Brunswick earthquakes, Bull Scien Soc. An 74, 621-653.
t
)
Zoback, Geophys. M.ResL 8and M. Zoback (1980). Interpretative stress map of the conterminous United States, .r.
5, 6113-6156.
U.S Grotoc CAL SURvry RESToN, Y!RolNIA (C.N., RLWJ U.S. GeoLoclCAL SURVEY MENLo PARK, CALIFORNIA (E.R) !
Manuscript received 2 February 1987 ADDENDUM ~
Subsequent monitoring in the Lake County region by Weston Geophysical on behalf of Cleveland Electric illuminating Company has detected 13 more mi- ;
croearthquakes during the period 15 April 1986 to 15 April 1987 (Weston Geophys- }
' ical,1987a, b). Three of these events are located within the previously def'med "*~ t aftershock cluster of the 31 January 1986 earthquake (Figure A1, top). One earth- '
quake occurred less than 3 km east of the brine injection well (SALT, Figure A1) north of Painesville. The other nine microearthquakes occurred within 5 km of the Pom deep injection wells operated by Calhio and all at depths of 2 km or less (Figure"A1, i I'
top and bottom). Magnitudes of these small events are all about M, = 1.0 or less. "'
The distance of these later earthquakes from the Calhio wells (<5 km) corresponds 2 80 -
to an inferred pore. pressure increase of at least 15 bars (Figure A1, middle). An increase of 15 bars Guid pressure corresponds with the pore-pressure increase E so -
- go -
calculated to trigger slip on favorably oriented, weak fractures, based on ont inferred values of the principal stress components (see Figure 10) These results suggest that , , _ i the state of stress in northeastern Ohio is suf6ciently close to failure that elevating w E ,, y formation pore pressure by a few tens of bars can trigger small shallow earthquakes.
A recent compilation of relative stress values (Evans,1987) indicates that similar 88}
stress conditions near the base of the Paleozoic secticn prevail over much of the c i*
Appalachian plateau.
a satr e= - - - -
.as - - - -
I 1
o F10. A1. (Top) Map e fluid injection wells (sol (Midd'e) Mazimum iner hypotheticalprofile A-A 2 and 9 million liters /mo:
12 yr, respectively. (Bot!
located within the rectan,
( .. . ..
^
h-
, _ n i l M '
^-
- ^
..y.,
.. . . , , e.o . .+. _- n
.. m <,r a
- . zs %. . y..
amm... l* hk.?.$$.. "f(.,,. &QK(
Vo a_.t,. gvW - ~' r*W
. .r m_ _ .~
,. . ~. , a em*-w . . mm . .. .
o de6-