Earthquakes and hydrocarbon production in the Fort St. John area of northeastern British Columbia

Full text

C.wADIAN JOURNAL OF EXPLOAzaTION GEOPHYSlCS
“ix 30, NO, 1 ,J”NE 19941, P 38~50
EARTHQUAKES AND HYDROCARBON PRODUCTION IN THE FORT ST. JOHN AREA OF NORTHEASTERN
BRITISH COLUMBIA
R.B. HORNER’, J.E. BARCLAYS AND J.M. MACRA$
M
ABSTRACT
Eanhquakes a6 hrfe as magnitude 4.3 have occumd neiv Pan St.
John. B.C., snce 1984 an* show both spatial and tcmpord comls,io”s
with oil cxtxmiun and asvriatcd high-pressure wlcr injection hn
he Brlhy Fmnarim ,Ptmnia”~ iI IhC FJl;iFle we>, ;mJ Elglc fields.
The majority of the larger ranhquakrs cm he gruupcd into three dis-
ttncl cIu~tcw in time and space. Each hes a duradon of about one
month. The first in No”cmbcr~“ecemtxr 14x4 appears LO have been
cenlrcd “SX Ihe western b”““dilry of Ihe FA& WCS, field. me iecond
in January-February 19% and the third in Drcemhc~ 1992~January
1993 werr huLh Ikratrd ahnut IO-15 kilometres (km) funhrr cast. pre-
sumahly nYC, the Eagle field. Earthquake* rronm, the Eagle field may
have begun IuI early as 1986. Fluid injection 10 humme recovery wib
initiiltrd at Eagle West in 1980 and at Eagle in 19X5, in hnth cases prior
to the msct of rrismicity. As production declined in the Eagle West
field in Ihe late 198Ua. so too did the sriamicity. Mast of the larger
emhqwkcs have nccumd during the winter months hut this is perhaps
not significant il rhey we related to only a few episodes of strain
~elciw This scason~l variation ii not rdlccted hy the pnxluction or
mjrction cata No sarthquakes were Iocated in the Fan St. John wea
betim 19X4. Them are no rcpms ofrarlicr kh ~~rnls and the scismo-
graph network in Western Canxh waukJ have permitted lamion of
earthquakes with mapniluder eqrlal 10 thorr in 1984 since he Inid-
I9h”S.
A field survey in January-March 1993 found epicentres of Lowe
magnitude ea*hqwkcr cxclusivrly in the Eigle field within a few kilns
metccs 01 both production and in,icctim wells. Focal depths could only
he determind with an accuracy of 3-4 km hut were consistent with
injectim depths ~fabout 2 km. The high rullacs injection pressurrh of
up 10 ahout 25 MPa, the fiulted nature of this region ol the Peace Riw
Arch. and the spetiul and temporal co~mli~lions indicate that ftuid injec-
tim should be considered 8s a possible cause of this reismicily. The
lnlervd of up lo 46 years heWren the unset of in.iection and reirmiciry
might retlea Lhc tire it ldikcr fur pressures away from the in,iection
wellh lo increilsr lo levels that could ildiille movement on prrrrirhg
fal,tr. Ahhwgh ““id pressure reduction does “0, appear t,, he signi,i-
cant. it is not pxsiblr lo fully r~illuilte the inlluencr 01 oh3 pamid
Ilitors with availahlr *ala. Hydmulic fracturing indicate5 a compres-
siw stress regime is, dqxhs of I 11, ? km where S,,,,w,r > S,,,>;” > s, an*
where Ihmst andlnr strike-slip tiluking would hr explcd.
The largest carthqu&kes to date have occurred in 1992, 1993 and
1Y94 and appal 10 retlect an increase in the fitqurncy and magnilude
of rvmts over the Eagle ficld aomparcd to the Eagle Wrrt field.
Although there has hem no significant danqr, many of the canh-
quakes were felt whh Mdifird Mexalli intensities as high as V. Felt
areas were a6 Ikrgc as 2ooO km”. The ha,ard pcesentcd hy thk seismic-
ity dill rneeds 10 he adequately assessed.
muscript received by the Editor April X, 1994: revised ~nenusfript rccci~ red,
INTHOI)IICTION
Beginning in 1984, a number of felt earthquakes us Iurge
as magnitude (M) 4, have occurred nettr Fort St. John in the
Peace River ttrea of northeastern British Columbiu. These we
unusual events in il region of typically very low-level seis-
micity on the Interior Platform eust of the Rocky Mountain
Foreland Fold and Thrust Belt (Figure I). This is ills” u
major oil and gas producing urea in British Columbia and the
earthquakes appear to cluster nex some of the fields just
north of Fort St. John (Figure 2). The first gas discoveries
were made in the early 1950s. Oil wells were first drilled in
these fields in the mid- 1970s; however, significant oil pro-
duction und enhanced recovery by water injection did not
occur until the early 1980s. just u few years beftxe the earth-
quakes were first observed.
There are numerous instances of earthquakes induced by
oil snd gas production in other basins elsewhere in the world.
They are usually attributed to fluid injection (Healy et ill..
1968; Raleigh et ul., 1972, 1976) or fluid pressure reduction
(Yerkes and Castle, 1976; Pennington et ill.. 1986; Segull,
1989; Grasso and Wittlinger, 1990). but there mey br il vttri-
ety of factors (Doser et al., 1991, 1993). In Canada. the only
documented case of induced seismicity by oil or g;ts entrac-
tion was ttt the Strachan gas field nrx Rocky Mountain
House, Alberta (Rebollar et al., IYXI; Wetmiller. 1986).
Milne (1970) implied thet a 1970 M 4.6 earthquake nex
Snipe LBkr in northwestcm Alberta ww related to oil pn,-
duction but there were insufficient data to tmake a definite
c”lKl”SiOn.
The purpose of this paper is to document the observed
record of eerthquakes near Fort St. John to September lY93
and to present results of un initial seismicity survey that wets
carried out from January to March, 1993. We make an initial
examination of the possible relationship between the seistnic-
ity and hydrocarbon production and use hydraulic fracturing
““C 14. 1994: manuscript accqml Aq!“Sl I I, 1w4.
‘Geolocical Survey 0fCsnada. Pecific Geoscience Centre. Bon hMX1. Sidnrv. Rritish Columbia V81. &I2
‘Fmne~ly,
Ge~~lo&al Survey of Canada. Institute of Sediicntary and Pet&m Geology. 1203 33rd Slrcct N.W.. Calgary. Alhrna T21. X7: presently. tt;ar*y Oil
and (ias Canada Limited. 1tKlO 530 8th Avrnuc S.W.. Calgary. Alhcna TZP XX
‘B.C. Ministry of Energy. Miner and Pemlcum Resources, Pcmleum Grulogy Hcanch. IX10 Blanchad Slrcet. 6th floor. Victoria, British Columhi;~ WV IX4
We gratefully xknowlrdgc the cmrihution of Greg Kuran 01 Home Oil Company Limited, Ceignry. who provided pmduclion. iqirclion and hydraulic tractwing
dilla di WPll iis much wl”ah,e disc”,,\i<,n and c~~,“me”,. We dlill ,ha,,h M. Rest. J. Casiidy. E. Davis. G. Rogers. D. IlFfetm >>“‘I Il. Wrichert. ilS KCll ilS three am,ny-
Incus rcvicwm lor their comments. Wc ue very grateful tier Ihc logistic support provided by Davr Johnson and his staff al the B.C. Ministry of Energy. Mints and
Pstmleum Resources office in For! St. John during khc 1W1 field survey. The field work could not have hem accan~pliihrd whhout the cnthusiaaic assi5limcc of
Jnhn Carter from the Pacific Geoscience Centre. We also thank Mary Wade, Ncii Thompson and Terry and Carolyn Wood and lhcir hnilics hr operating the heirs
mograph at Chartie Lake, Cecil Lake and North Pine. respectively, and for their cncauragemm~ and ruppon. This is Geoltrgicd Survey of Canaih Cwwihu~ion
NO 10394.

Fig. 1. Seismicity in western Canada cuer a IO-year period from 1982-1991, superimposed on the morphogeological belts of the Canadian Cordillera.
data to infer the existing stress regime. This is extremely
important in order to assess the potential hazard such seis-
micity might present. not only to the industry but to the pub-
lic as well. Further analysis will await the results of a
detailed seismicity survey conducted over the Eagle Field
from November 1993 to March 1994.
EARTHQUAKE HISTORY
All of the earthquakes that have been located in the Fort St.
John region to September 1993 are listed in Table I. Most of
the events over M 2.5 were felt locally. The largest earthquakes
were generally felt to distances of about 25 km with
Modified Mercalli intensities as high as V. There were
reports of cracked plaster, fallen pictures and small objects
knocked off shelves. In many cases, one strong thud was
reported as though there had been an explosion or a vehicle
had run into the building. People sleeping were awakened.
An isoseismal map shown in Figure 3 for one of the large
earthquakes on January 9th, 1993, is typical of the intensity
distributions produced by the larger events. The felt area is
about 2000 km2.
The first earthquakes were not observed until 1984, even
though the existing seismograph network would have permitted
CJEC 40 ,unr Iv*

t:/\I~I’HQIIAKIS AN0 IIYI~KO~‘AKM,N ,‘K00,:(‘710N IN ‘I’HI: I’OKI 5,‘. 10,1K ,w,:n
121’Oo’W 1:
Fig. 2. Epicentres of all earthquakes located in the Fort St. John area
in relation to the oil and gas fields. Event numbers refer to Table 1, -F’
indicates an epicentre determined during the January-March 1993
field survey (see also Figures 6 and 11). -s” are locations of the tem-
porary seismographs (Table 2). ‘“84’ marks the location where the first
tremors in 1984 were mainly felt and represents the most accurate
epicentre for these events.
i PINE . II-111 \
/ T - _z
(-,: ~ , “i’
k;-”
IV-V \
,~s
:T &;;;L )
0 20krn
I I I I I I y
I I I
12100’ 12(1x
Fig. 3. Modified Mercalli intensities produced by the M 3.9 earthquake
on January 9th. 1993, at I&05 UT (event 19 in Figure 2 and Table 1).
the location of events like these. i.e., about M 2.5 and
greater, since the mid-1960s (Figure 4). Stations at Fort St.
James, Mica Creek, Edmonton and Yellowknife, installed
within a few years of each other, dramatically lowered the
location threshold from about M 5 that existed in the 1950s.
There is no record of moderate or larger events before the
1960s. If they had occurred they would have been felt and
reported.
The accuracy of the epicentres shown in Figure 2 (exclud-
ing those from the field survey) cannot be considered better
than about IO km because of the seismograph distribution
(Figure 4) and the relatively low magnitudes. The nearest
station to Fort St. John is Bennett Dam, about 100 km to the
west; the others are all over 300 km away. Accurate focal
depth determinations are also precluded, although the rela-
tively high intensities do suggest a shallow, upper crustal
source (e.g., Gendzwill et al., 1982).
Despite the uncertainty associated with the computed epi-
centres the macroseismic data confirm the general pattern
shown in Figure 2 and suggest spetial correlations specifi-
cally with the Eagle West and Eagle oil fields. Although the
epicentres of the initial sequence in 1984 (events l-5) are
scattered outside the Eagle West field in the Stoddart and
Stoddert South fields, they were primarily felt on the east
side of Charlie Lake near the west boundary of the Eagle
West field and probably occurred within a few km of that
location. At a residence in IO-85.l9Wh (‘84’ in Figure 2),
event 4, only M 2.8, was strong enough to knock ornaments
from a shelf and felt like an explosion. Another small unlo-
cated earthquake felt there the next day was not felt by
neighbours about 2 km away (M. Smith, pen. comm., 1984).
In contrast, the earthquakes in 1992.1993, and perhaps as
early as 1986 (event 7). appear to be further east over the
Eagle field. This is evident by both the overall IO-IS km
eastward shift observed in the computed epicentres (Figure
2) and the intensity distributions determined from interviews
conducted in January and February 1993. The isoseismal
map shown for event I9 in Figure 3 is typical. In addition.
residents on the east side of the Beatton River over the Eagle
field reported many smaller felt events not detected by the
regional seismograph network. Also, none of the larger earth-
quakes were felt strongly at North Pine, as would be expected
if events IS, 16, II and 22 had occurred where they are plot-
ted in Figure 2. In fact, all were felt with higher intensity at
Fort St. John (Figure 3) suggesting true epicentres further
south, over the Eagle field. As a result the apparent north-
south epicentral trend over the Eagle field and other apparent
lineations in Figure 2 cannot be considered real.
An important feature of the earthquake distribution is the
clustering in both time and space. Nineteen or 20 of the 24
events in Table I can be grouped into 3 distinct clusters,
each having a duration of about I month and apparently con-
strained to a much smaller area than indicated by the com-
puted epicentres on the basis of macroseismic data discussed
above. The first in November-December 1984 (events l-5)
was centred near the western boundary of the Eagle West
field. Event 6 in March 198.5 could very likely have occurred
c,ea 41 ,““C I!?24
I

R.“. HOKNER. J.E. RAKCLAY md J.M. MACKAE
Table 1. Earthquakes located in the Fort St. John atea to September 1993. Also included are preliminary details of the May ,994
ewnt.
EW”l Date Time (UT)
NO. yy-mm-dd hh:mm:ss
::
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
84-1 l-09
05:30:18
84-11-10 ,5:34:34
84-1126 02:35:52
84-12-09 04:,7:32
64-12-12 12:07:40
85-03-13 20:19:36
66-02-10 12:09:,6
67-02-12 ,2:45:,1
87.11-12 05:30:19
89-01-23 07:35:53
92-01-23 04:44:22
92-01-23 05:52:20
92-0*-o* 21:5,:,2
92.o*-09 02:27:,*
92-02-I 1 ,0:25:09
92-02-15
92-12-26
93-01-08
93-01-09
93-01-09
93-01-09
93-01-09
93-01-09
93-01-30
94-05-22
07:42:56
20:52:40
oK32:50
l&05:47
18:12:13
16:4,:39
19:31:02
2,:1604
,6:03:45
,5:06:46
Lat. Long. Depth
Deg N Deg w km
56.398
56.274
56.391
56.367
56.405
56.394
56.407
56.450
56.354
56.397
56.3,1
56.337
56.375
56.343
56.378
56.355
56.366
56.326
56.348
56.310
56.321
56.365
56.344
56.351
56.30
120.929 5.0G
2.6
121.050
5.0G
2.2
120.676 5.0G 2.6
121.041 5.OG 2.6
120.967 5.OG 2.6
121.017 5.OG 2.4
120.779 5.OG 3.1
120.927 5.OG 2.5
120.685 5.OG 3.0
120.878 5.0G 3.1
,*o.i%o 5.0G 3.2
120.654 5.0G 2.6
120.877 5.0G 2.5
120.637 5.OG 2.6
120.723 5.0G 3.5
120.741 5.OG 3.2
120.762 5.OG 4.1
120.741 5.OG 2.6
120.725 5.OG 3.9
120.730 5.0G
120.877 5.0G
120.722 5.0G
120.694 5.0G
120.665
120.75 5.OG
2.5
2.4
3.7
4.0
3.5
4.3
Ja”“ary-March 1993 Field survey
25. 93-02-I 1 09:54:39 56.303 120.716
26. 93-02-12 ,3:46:25 56.303 120.737
27. 93-02-13 09:43:12 56.312 120.699
26~ 93-03-16 03:“3:26 56.292 120.736
29. 93-03-29 063531 56.303 120.753
2.OG
0.5
4.0
4.0
0.0
blag. comments
<I
Cl
1.6
1.8
1~3
Felt at FOrl St. John
Felt on east side of Charlie Lake.
Felt (IV-V) on ea5.t side of Charlie Lake.
Felt on east side of Charlie Lake.
Fe,, at Fort St. John.
Felt in the Forl St. John area.
Felt in the Forl St. John area.
Felt in the Fort St. John area.
Felt in the For, St. John area.
Largest earthquake to date. Felt strongly (IV) in
Fort St. John. Many People awakened. No
reports 01 damage.
Felt (IV) in Fort St. John.
Largest earthquake to date. Felt strongly (IV-V) in
For? St. John and area. similar to event 2 weeks
later on Jan. 9. ,993. Several small aftershocks
in next few days.
Not reported felt.
Felt strongly (IV-V) just no”h 01 Fort St. John; (I”)
at Fort St. John: (Ill-l”) at Cecil Lake and Charlie
Lake; (Ill) at Montney; and (II) at North Pine and
Rose Prairie. Also felt at Taylor. Perceptible to
distances of about 50 km. About 20 aftershocks
observed in the next 5 hours including felt events
at ,6:41, 1931 & 2,:16. See Figure 3.
NO, reported felt.
Felt mildly in the Fort St. John area.
Felt in the Fort St. John area. Less severe than
event 19.
Felt strongly in the Fort St. John area. Similar to
event 19.
Felt in the Fort St. John area.
Felt (IV-V) in the area bounded by For? St. John.
Charlie Lake, North Pine and Cecil Lake. Similar to
event 19. Foreshock a, ,5:02, M 2.6 was also felt.
Regional solutions (I-24) are “of accurate to much better than 10 km and there is no cOntr0 on focal depth,
Field survey epicentres are accurate to better than 2 km, focal depths to 3-4 km.
at this Iwzation as well. The second and third clusters in
January-February 1992 and December 1992.January lY93
(events I I- I6 and 17.24, respectively) are centred over or
very near the Eagle field.
This distribution suggests a cusal
relationship, i.e., the earthquakes within each cluster are not
independent events but are
related to a common strain release
episode, much like a main shock-aftershock sequence.
The seismicity over or “ear the Eagle field has also been
the most intense and includes all of the ler~er earthquakes. M
3.5 and greater. This activity has continued with probably the
largest earthquake to date occurring on May 22nd. 1994 at
15~06 UT. The preliminary magnitude is 4.3 and the epicen-
tre appears to be in the same area as those in January 1993.
The felt area is at least as large as that shown in Figure 3
with similar intensities.
Until the May I994 earthquake occurred, an intriguing fea-
ture of the temporal distribution was the apperant confinement
to the winter months. All of the larger earthquakes had
occurred between November and March (Figure 5). This is
now probably not significant, particularly if these events are
related to only a few episodes of strain release.
JANUARY-MARCH 1993 SEISMICWY S~IHVEY
Following the January Yth, 1993 earthquakes, two seismo-
graphs were deployed in the epicentral region to improve
monitoring of any low-level seismicity. In February a third
seismograph was installed along with two portable instru-
ments. Sites are shown in Figure 2: coordinates and operating
times are listed in Table 2. About a dozen small earthquakes,
42 ,W’C i’i’li

,::\,~,,,~~~,~,\l\,:‘; ;\N,) ,,YI~i<Ol’.\I<IK~I l’i<ollr”lIoN IN I’llI: I,OI( I 5 I, 1011Y ?Aiil ,!
Fig. 4. The current distribution 01 seismograph stations in western
Canada, showing the operating periods of those closest to ForI St. John.
Fig. 6. The monthly distribution of earthquakes, M 2 and greater
(Table I), located near Forl St. John. Included is the M 4.3 event on
May 22”d. ,994.
as well as the M 3.5 cvcnt on January 31lth. wcrc dctcctcd up
to the end of March IYY?. Five of these events were recwdcd
on at least three of the seismographs and could he located.
The hypocrntres (Table I) are considered accurntr to helter
than 2 km horizontally and y-4 km vertically. For thih study
we used a velocity model from u regional refraction survey
(Zrlt and Ellis. 1989). The seismograph distribution and anile
logue data precluded a more refined model.
The epicentres all lie in the Eagle oil field (Fi~urcs 2 and
6) at depths of less then about 5 km in what appews to he a
somewhat ENE-WSW elongated zone, about 5 km in length.
This distribution is also supported by the variation in
P
arrival time differences observed at CHBC and CEBC for a
number of common events (Figure 7). The I. I second varia~
tion would translate to a horizontal distance of ;Ihwt 5 km
assuming a
P
velocity of 5 km/xc. This zone is also very
likely where the larger events in January occurwd and sup-
ports the eastward shift discussed earlier. It is also important
to note that there were no wcnts located outside of the Eagle
field. This seismograph network would have permitted location
\
-k I \I \ 1
\ I
-‘ho,”
,! *- j
, 7.85
1 EAGLL
WEST
,’ c-7 5
“‘20 &,.c
<n .,iJ:
Fig. 6. Epicentres determined during the January-March 1993 sur-
vey (Table 1) superimposed on a stwcf~re map on the top 01 the
Selloy Formation, also showing some of the inferred faults and the
outline of the Eagle West and Eagle oil units (note that the unit
boundaries are different from the field boundaries shown in Figure
2). Contours are metres below sea level. These epicentres are con-
sidered accurate to better than 2 km. Seismograph locations are
shown in Figure 2.
Table 2. Operating parameters for the Fort St. John seismographs during the January-March
1993 seismicity survey. See Figure 2 for station lx&ions.
Station Name Code La, Long Elev Opened/Closed Paper Speed
N w m 1993 mm/min
Charlie Lake CHSC 56.3146 121.0076 ml Jan 14 Apr 1 60
Cecil Lake CEBC 56.3006 1*0.5501 720 Jan 15 Apr 1 60
North Pine NPBC 56.4071 120.6473 750 Feb9-Aprl 60
Airpolt FSBI 56.2391 120.7074 695 FebS-Febl2 120
Pineview FSS2 56.3412 120.7623 670 FebS-Febl2 120
<‘JMi 43 lll”Y I’,‘,,
I

I” $1 FVFNT
r
$4
mm3
322
=lL
0’
‘0.5 ‘1 .o’ ‘1.5’
‘2.6
SECONDS
CHBC-CEBC
Fig. 7. Differences in P arrival times at CHBC and CEBC for 11
earthquakes recorded by both seismographs. See Figure 2 for station
locations and Table 1 and Figure 6 for event numbers.
of similar magnitude events in the Eagle West, Stoddart, Fort
St. John, Cecil Lake or North Pine fields-had they occurred
during this period.
GEOt.OGICAt. SF.TTING
This seismicity occurs in an area of the Western Canada
Sedimentary Basin known as the Peace River Arch. The
“Arch” is B Precambrian granitic basement-cored regional
crustal structure covered with Proterozoic and Phanerozoic
strata that display thickness and facies variations as a result
of its tectonic activity. This cratonic arch is located near the
western margin of the Basin and is oriented NE-SW, perpen-
dicular to the depositional strike of the Basin (Figure I). It
has been tectonically active, although relatively stationary
geographically, since the Proterozoic and perhaps earlier
with a Proterozoic to Devonian uplift phase (Peace River
Arch), a Carhoniferous to Triassic downwarp phase (Peace
River Emhayment) and a Jurassic to Recent regional suhsi-
dence phase with some subtle uplift (Table 7). Numerous
smaller episodic movements have also occurred during these
three tectonic phases along numerous high-angle hasement-
rooted faults within two NW- and SE-trending orthogonal
sets. These faults were continually active and appear to have
absorbed the large-scale movements of the structure
(Sikahonyi, 1957; Sikahonyi and Rodgers, 1959).
Since the first well was drilled in the Arch area in 1949
(deMille, 195X), a variety of structural and structura-strati-
graphic traps related to the active tectonic history have been
investigated. Prolific oil and gas fields have been found, such
as Normandville (Devonian), Dunvegan (Carhoniferous),
Eagle (Permian), Boundary Lake (Triassic) and Elmworth
(Cretaceous), yet despite over 40 years of exploration and
research in this area, researchen have not been able to singu-
larly explain the origin(s) of the Arch and its several phases
(e.g., O’Connell et al., 1990).
The Eagle, Eagle West and nearby Stoddart and Cecil
fields (Figure 2) occur in one of the more tectonically active
areas in late Paleozoic time. They are situated near the crest
of the Proterozoic Arch and the north rim of the
Carhoniferous grehen on the eastern flank of the Manias
High, a north- and northwest-trending Cretaceous-Tertiary
anticlinal fold which is an inversion structure imprinted on
the downwarped Carhoniferous-Perminn grahen (David G.
Smith, pus. comm., 1990; Barclay et al., 1990). The main
R.B. H”RNER.J.E. BARCLAY mdJ.M. MACRAF
producing oil and gas reservoirs lie at depths (below surface)
of about 2 km within the NW-trending shoreline-related
deposits of the Imainly Permian Belloy Formation. The
Brlloy consists of interspersed sandstone and carhonute units
that were deposited on an elevated shallow marine shelf on
the northern rim of the grahen during its Peace River
Emhaymrnt downwarp phase (Barclay et al., 1990: Leggett
et al., 1993). During Permian time. while the grahen was in
its decaying phase and suhbidence and block faulting was
reduced compared to Carhoniferous time, the northern rim
appears to have persisted and was affected by significant
continued faulting compared to other areas (e.g., Barclay et
al., 1990, their figures Xh, c). Fault displacements were in the
order of 30 m and may have been up to IS0 m in some
regions.
PKOOIICTION HISTOKY
Petroleum and natural gas exploration and production
have played an important role in the economic activity of
northeastern British Columbia since the early 1950s.
Significant gas discover& in the Belloy Formation were
made near the city of Fort St. John at the Fan St. John field
in 1952 and 1953 (Figure 2). at the Fort St. John Southeast
field in 1952, at the Stoddan field in 1957 and at the Stoddart
West field in 1961. Significant Belloy oil discoveries were
made in 1970 at the Stoddert West field, in 1972 at the Eagle
field and in 1976 at Eagle West field.
The Belloy gas pools have produced in excess of 75% of
their hydrocarbon resource originally in place. No pressure
maintenance projects have been put into place in there pools
and, as a result, average reservoir pool pressures have
declined to 25% or less of their original pressure. The
Stoddart West. Eagle and Eagle West Belloy oil pools were
initially produced by solution gas drive the primary expan-
sion of the crude oil and dissolved solution gas in response to
a controlled pressure release at the surface. This type of
depletion mechanism is inefficient, so pressure maintenance
by water injection was introduced into Eagle West in 1980,
into Eagle in IYXS and into Stoddart West in 1991 to
enhance the producing rate and oil recovery. In contrast
to
the gas pools, original reservoir pressures were not allowed
to drop more than 50% before waterflooding was introduced.
Since then pressures have been maintained or increased
slightly from those that existed prior to the start of the flood.
Cumulative voidage (the total volume of fluid extracted
minus that injected) taken from the two Eagle pools where the
earthquakes appear to have occurred is shown in Figure 8 along
with the earthquake history from Table I. No direct cowelation
is noticeable, other than that the earthquakes start when
approximately I .7 x IO’ m3 of fluids or 4.4%, of the hydrwar-
bon pore volume had been removed from these reservoirs.
Waterflood stun-up times are also indicated. At Eagle West
there wils a 4.year interval before the first earthquakes were
observed. At Eagle the interval was 6 years if, in f&t, the I992
events were the first at that field. The 19X6-1989 earthquakes
(events 7. IO) are not well Iouted and the availehle mwxoseis-
mic data are not sufficient to resolve which group they might
belong to. It is also evident that earthquake frequency and
ClEG 44 lY”C IW1

,:;\,~‘,‘,,,~,~:\h,:S ANIl iiYIn<ot’\l<IIoI\ ,‘1<01l,~r’1’10\ IK’IIII~ IO1~‘1’51’ IOlIN ;\,<I \
Table 3. Peace River ArchiEmbayment history.
Time Geometry
ARCHEAN ?
PROTEROZOIC ARCH
CAMBRlAN ARCH
ORDO”,ClAN ARCH?
SlL”R,AN ARCH?
DEVONIAN ARCH
Description
Basement terranes 8 CWS, & mantle stw~b,res crosscu1 late, “Arch’ trends.
Truncated Upper Proterozoic strata in outcrop.
Shallow water deposits: Middle, Vpper Cambrian missing.
No sediments in region.
No sediments in region.
Basement emergent B hinged by reefs, sands. Terrestrial arkoses on crest
areas. Progressive transgressive onlap throughout. 600 m total relief on Arch.
Active NW 8 NE basement-rooted high-angle faults.
CARBONIFEROUS EMBAYMENT & GRABEN Old Arch blanketed by carbonates throughout embayment & graben. “Peace
River Embayment’developed as broad NW-SE downwarp. Later. embayment
shrunk lo central. elastic-filled E-W graben. Old NW & NE faults very active in
graben.
PERMIAN EMBAYMENT Sands & dolostones deposited throughout Peace River Embayment downwarp.
Graben filled 8 less active. NW-SE faults locally active. e.g.. in Eagle field
area. Eagle fields in shoreline deposits.
TRlASSlC EMBAYMENT Return lo broad Peace River Embayment downwarp.
Ctastic-carbonate-evaporite deposits. NW-SE faults active with reduced
throws.
JURASSIC REGIONAL SUBSlDENCE
8 SOME UPLIFT Subtle extra subsidence more than rest of Western Canada Basin (Columbian
foredeep). Middle Jurassic absent indicating uplift. Then westerly-derived
mud and sand starting in Upper Jurassic.
CRETACEOUS
CENOZOIC
SUBSIDENCE & DRAPE
STRUCTURES
SUBTLE ARCH
Episodic extra foredeep subsidence. Drape folding & fracturing over old Arch
8 graben. reets and faults.
Laramide deformation inducing subtle uplift, persists into present. Teniary
removed from here 8 most of western Canada. Arch activity thus unclear.
Recent earthquakes. Eagle fieid. production-induced rejuvenation of old
faults?
Compiled from deMi,,e. 1956: Lavoie, 1956; Williams, 1956: Pugh. 1973; Porter et a,., 1962; Cant. 1966; ROSS and Stephenson, 1969: Stephenson et
al.. 1969: Zelt. 1969: Barclav et al.. 1990: Hart and Plint. 1990; Leckie et al., 1990; McMechan. 1990; Norford, 1990; O’Connell and Bell, 1990;
O’Connell et a,., 1990; P&n et al.: 1990; Floss, ,990. O’Connell et a,., 1990; P&n et al.: 1990; Floss, ,990
EAGLE WEST AND EAGLE EAGLE WEST AND EAGLE
0 0
- -
z z -100,000 -100,000 .: .:
:: :: . .* . .* - - 2 2 -800.000 -800.000
> >
Y Y -1200.000 -1200.000
? ?
: :
3 3 -,.500.000 -,.500.000
0 0
L % L %
,j~~* J,’ ,j~~* J,’
,’ ,’ _’ _’
i’ i’ / /
*” *” A, 8, A, 8,
magnitude are higher at the Eagle field. All of the M 3.5 and
larger events have occurred there since 19Y2.
t
Figure Y shows oil, gas and water volumes produced from
the Eagle West and Eagle pools from January 1977 to March
1993. Peak oil production of about 40 000 m3/month.
reached in the early 1980s in Eagle West, started tu decline
about 1987 and by the beginning of 1993 was d own to about
12 000 m?. In Eagle, peak oil production of about I2 000
m3/month was not reached until 19X6. In hoth Eagle West
and Eagle the earthquakes lagged peak production periods.
The seasonal variation exhibited by the earthquakes (Figure
5) is not evident in the production rates.
Figure IO shows average monthly injection rates and aver-
age wellhead injection pressures over the same period as
Figure 9. At Eagle West there was an incrense in average
surface injection pressure from ahout 20 MPa to 27 MPa in
1984, just before the first earthquakes were ohserved.
Injection pressures were reduced in 1987 and appear to he
reflected by a coincident decline in seismicity. At Eagle,
slightly higher injection pressures of ahout 2.5 MPa were
r
-2mo~ooo I ‘,B’ Id ‘82’ ‘84’ ‘B6’ ‘es’ ‘90’ ‘32’
Fig. 6. Combined cumulative reservoir voidage for the Eagle West
and Eaale units from Januarv 1977 to March 1993. in relation to the
observed seismicity. Most &hquakes about M Z.&and greater could
have been located by the existing seismograph network (Figure 4)
since the mid-1960s. A-A’ and B-B indicate the commencement of
enhanced recovery operations by water injection and the earthquakes
possiblv located in the Eaale West and Eaqle fields. respectivelv.
&thq&kes at the Eagle field may have begun as early as i966. attained near the end of 1988 and have hccn maintained

R.R. HOKNER, I.E. BARCLAY and J.M. MACRAE
50000
1 EAGLE WEST
Fig. 9. Monthly oil, gas and water volumes produced frown the Eagle
West and Eagle units from January 1977 to March 1993. Note the
scale differences. The times and number af earthquake wcurrences
are marked.
since. Again, there is no evident seasonal variation in pres-
sure, nor is there the same apparent direct correlation
between seismicity and increased injection pressure observed
in Engle West.
Also included in Figure IO is an approximate estimate of
voidage replacement. Voidage replacement ratio (VRR) is
the ratio of monthly fluid volumes injected at reservoir tem-
perature and pressure divided by the total fluids produced. A
ratio of one indicates that replacement equals withdrawals. In
contrast to Eagle West where VRR was maintained near I,
VRR at Eagle was only about 0.5 except for a Z-year period
from 19X8 to 1989 when it was I or slightly higher. During
this time there was only one earthquake.
DWJ.JSSlON
Although a detailed analysis is precluded by the relatively
poor location accuracy of the larger magnitude events and
the absence of any mechanism solutions, there do appear to
be spatial and temporal correlations between the earthquakes
and oil production in the Eagle West and Eagle fields. Fluid
injection in particular must be considered as a possible
*oooo
EAGLE WEST
!
2
-l
5 (NUMBER) W(l)(l) (l)(l) (1) (6) (8)
__ WATLR VOL. IN,. (m’)
INJ. PRESS. (LPO)
-----~ “RR (x,0) /r-t _’
20000 - _’ :;:
EAGLE
~0000 -
0 ‘88’
‘92’
’
Fig.
10. Monthly water vot~mes injected. average monthly injection
pressures and
voidage replacement ratios in the Eagle West and
Eagle units from January 1977 to March 1993. Note the scale differ-
ences. The times and number of earthquake occurrences are marked.
cause. The surface injection pressures of about 25 MPa are
much higher than injection pressures in other oil fields that
have been demonstrated to induce seismicity (e.g., Davis and
Frohlich, 1993) and the earthquakes located during the
January-March 1993 field survey all lie within about O-4 km
of both injection and production wells (Figure I I). taking
into account epicentral uncertainty of I-2 km. Focal depths
are less accurate but are consistent with injection depths of
about 2 km.
Earthquakes induced by fluid injection are usually
explained by the theory of effective stress (Hubbert and
Ruby. 1959). Increasing fluid (pore) pressure will reduce the
frictional resistance to fracture by decreasing the effective
normal stress xross the fault plane according to the equation
(e.g., Davis and Pennington. 1989):
L, = ?I + Ff(% - P),
where zCrit is the critical shear
stress
needed
to
cause slip on a
fault, ~~~ is the inherent shear strength of the rock. 0, is the
normal stress across the fault, ,I is the pore pressure and Pr is
45 ,,rnc 19%

R18W6M R17
0 Battery
A
Satellite
0 Field Header
0 Oil Well
F Shut In Oil
0 Gas Well
,?X Shut In Gas
4 D&A
fi Injection
@ Epicentre
0 1 2km
I I /
Fig. 11. Distribution of well-determined field epicentres (Table 1) in relation to production and injection wells in the
Eagle field
the coefficient of friction. In Figure 12, Mohr circles are used
to illustrate the inferred state of stress at the bottom of injec-
tion wells in the Eagle field and for comparison at Rangely,
Colorado, where seismicity was clearly related to fluid
injection (Raleigh et al., 1972). At Rangely, surface injection
pressures above about 6.5 MPa combined with hydrostatic
pressures of about 19 MPa were sufficient to exceed the
Mohr-Coulomb failure criterion and induce earthquakes.
These ranged in magnitude up to M 3.5 and occurred at
depths of about 1.X-3.7 km.
In situ stress magnitudes in the Fort St. John area can be
inferred from hydraulic fracturing data in the Eagle and
Stoddart fields (Table 4). Instantaneous shut-in pressure
(ISIP) in three different formations at depths of l-2 km
yield estimates for the minimum compressive horizontal
stress. s,,,i”, (Haimson and Feirhurst, 1970) that are an
average of about 3 MPa higher than corresponding vertical
stress values, S,, calculated assuming B lithostatic gradient
of 25 kPe/m (Figure 13). No direct measurements of the
larger horizontal compressive stress (.Q,,,,) are available
but Kry and Gronseth (1983) calculate S,,,,,:S,,,, ratios
of I .3-l .6: I .O in the Peace River Arch area (see also Bell
and McCallum, 1990). At injection depths of about 1900
tn. assuming a lower SHmnx:SHmin ratio of I .3: I .O, we
obtain
o, = SHma = 65.0 MPa
> 09 = .&, = 50.0 MPa
> q = s, =41.5 MPa ,
IOT
RANGELY, COLORADO
Fig. 12. Mohr circle diagrams showing the inferred stress conditions
at the bottom of injection wells at Rangely, Colorado (Raleigh et al.,
1972) and at the Eagle field near Fort St. John. B.C. (Table 4 and
Figure 13). n, and o3 are maximum and minimum principal stresses.
respectively. 0,’ and 03’ are effective stresses indicating the com-
bined influence of hydrostatic and surface injection pressure.
Portions of the circle to the left of the Mohr-Coulomb failure line indi-
cate pressures are more than sufficient to induce movement on
favourably oriented preexisting faults with zero strength. The failure
criterion for Fort St. John is not known so two lines are plotted to
indicate a possible range.
C,Ki
47 ,unr 14%

where 0,. IS? and o3 are the maximum, intermediate nnd
tninimum principal stresses, respectively. These values indi-
catc u compressive stress regime where thrust and/or strike-
slip faulting would be expected (since 0, and G, have similur
magnitudes).
The Eagle surface injection pressure of 25 MP” ond a
hydrostatic pressure of about 18.6 MPu are both subtracted
from the maximum and tminimum principal stresses in Figure
12. Since the f;tilure criterion is not known for this case. two
tailurr lines are plotted to indicate a possible range and illus-
trate the effect of different coefficients ot twtion. The result-
ing Mohr circle lies to the left of both lines. If a larger
S .S ratio had been used both failure criteria wuuld
,h.Lx ,,mn
have been exceeded by tm even larger degree. On the other
hand, friction losses that would lower bottomhole pressures
by ahout 0.5 MPn have been ignored. Nonetheless. it does
appear that injection pressures at the Eagle field could be
sufficient to promotr failure on favuurahly oriented preexist-
ing t:wlts.
Waterflooding was initiated on the west side of the
Beatton River in IYXS and extended eust of the Beatton
River in 1986 (Figure I I ). The delay of up to 4-6 years after
injection was initiated and before earthquakes were observed
might be related to the time it takes for pressures away from
the in,jection wells to incrrnse to levels that could initiate
~novrment on preexisting faults. There is abundant evidence
for faults in this region of the Peace River Arch (such as dis-
placements seen on seismic sections and displucements seen
on thickness, structure and facies maps of stratigraphic units:
see Cant, 1988: Barclay et al., IYYO: O’Connell et a., IYYO:
~ , 1; 2; , :” L “p L 5; 1 6!
\’
\‘\ EAGLE AND
STODDART FIELDS
K.“. HOKNFR. J.R. BARCLAY ;,nll J.M. MACKAE
Table 4. Minimum compressive
horizontal stress (S,,,,) from
hydraulic fracturing data in the Eagle-Stoddarl area of Felt St. John.
ISIP is instantaneous shut-in pressure at the surface. Pressure gradi-
ents of the fluids used for hydraulic fracturing range from about 7.8.
10.0 kPalm. Ignored are friction losses that would lower S,,,, values
by about 0.5 MPa.
Well Location
05.26.84.18WW
IO-*2-84.18W6M
1*-04-86-20W6M
06-04-620W6M
02-36-84~l@mM
14.19-8519W6M
15-35.85.zoww
FO,tTlC+fiO” Depth UP Static pressure Gm,,,
m MPa MPa MPa
&thing -946 21.2 7.4 28.6
Gething -973 19.5 7.6 27.1
ooig -1601 27.2 16.0 43.2
ooig -1608 27.9 12.5 ‘lo.4
Belloy -1873 34.0 14.7 48.7
Belloy -1940 34.0 19.0 53.0
Belloy -1940 36.0 15.2 51.2
and others listed in Table 3) and the Eagle field itself is dis-
sected by small NE- and NW-trending normal faults (Leggett
et al., 1993).
Davis and Frohlich (1993) developed tt criterion to assess
whether injection is likely to have induced the ohserved
seismicity. They pose seven questions based on the historical
earthquake record, temporal and spatial correlations and
injection practices and state “that in every case we studied
where five or more of the questions had “yes” answers. most
professional seismologists would conclude that injection
induced the earthquake sequence”. Their questions are:
I. Are these events the first known earthquakes of this char-
acter in the region’!
2. Is there a clear temporal correlation between injection and
seismicity’?
3. Are epicentres within 5 km of injection wells?
4. Do some earthquakes occur at or near injection depths?
5. If not, are there known geologic struct”res that may chan-
nel flow to the sites of eanhquakes?
6. Are changes in tluid pressure at well bottoms sufficient to
encourage seismicity’!
7. Are changes in tluid pressure at hypocentral locations suf-
ficient to encourage seismicity?
In this study we can address the first six questions and would
answer “yes” or “apparently yes” to all of them.
It is difficult to evaluttte the possible intluence of other
factors with available data. Pool depletion does not appear to
be as significant as high-pressure tluid injection. If the large
volumes of gas that have heen produced from nearby gas
pools have had no precipitating effects for earthquakes it is
unlikely that the relatively small ttmount of unreplaced
voidnge in the Eagle pools would have an effect either.
BELLOY
\ \
C,li<i 48 ,u,/r IUUI
Fig. 13. Minimum horizontal compressive stress values. S,,,,, from
hydraulic fracturing data in the Eagle-Stoddart area of Fort St. John
(Table 4). S,,,,, values are an average of about 3 MPa higher than
comparable lithostatic pressures. Also indicated is Eagle injection
pressure of 25 MPa at a de,,th of about 1900 m.
There are severnl observations that point to a correlation
between earthquakes and oil production in the Fort St. John
area.
I. These earthquttkes occurred in a region of typically very
low-level seismicity.

The majority of the larger earthquakes are grouped into
three distinct clusters, each with a duration of about one
month and located over or very near to the Eagle West or
Eagle oil fields.
The first cluster in November 19X4 was apparently centred
neu the west side of the Eagle West field. This field was
discovered in 1976; however. peak production and associ-
ated high-pressure water injection to enhance recovery did
not occur until 1980. Average surface injection pressures
were increased from about 20 MPa to 23 MPa in 19X4,
just before the first earthquilkes were observed. As injec-
tion pressures and production declined in the late 1980s.
so too did the seismicity.
In January-February I992 and December 1992.January
1993 most of the earthquakes and all M 3.5 and greater
occurred in two clusters over or very neu the Eagle field.
Although this field was discovered in 1972, significant
production was delayed until Eagle West production
started to decline. Water injection was initiated here in
I985 and 1986 with pressures reaching about 25 MPa,
about 2 MPa higher than at Eagle West. Earthquakes pos-
sibly occurred at the Eagle field as early as 1986.
A field survey in January-March 1993 found low-magni-
tude events exclusively in the Eagle field. Epicentres were
within a few km of both production and injection wells
and although focal depths were not accurate t” better than
3-4 km they were consistent with reservoir depths of about
2 km.
The injection pressures of about 25 MPa are perhaps high
enough to induce failure on favourably oriented preexist-
ing faults. The delay of up to 4-6 years after injection has
been initiated and before earthquakes are observed is per-
haps attributable to the time it takes for fluid pressure
away from injection wells to rise to the level necessary to
initiate failure. Hydraulic fracturing indicates a compres-
sive stress regime at depths of I to 2 km where S,,,,,,
’ hli” > S, and where thrust and/or strike-slip faulting
would be expected.
There is abundant evidence for tiults in this region of the
Peace River Arch that has been tectonically active since at
least the Proterozoic.
Although high-pressure fluid injection must be considered
as a possible cause of this seismicity, we cannot rule out the
possible contribution of other factors until more accurate
hypocentre and mechanism solutions, in particular, are avail-
able.
The earthquakes in December 19Y2, January 1993 and,
most recently, in May IY94. with magnitudes up to M 4.3.
would he among the largest associated with any secondary
recovery or pressure maintenance program. All were strongly
felt with Modified Me&Ii intensities as high as V and felt
areas up to about 2000 km2. Davis and Frohlich (1993)
report magnitudes as high as M 4.6 in the Cogdell oil field in
Texas. How large the earthquakes in the Fort St. John area
could become is not known. Also, the hazard presented by
this seismicity still needs to be adequately assessed and war-
rants further investigation.
(‘JlKi 49
KIWERliN(.l<S
Barclay. J.E., Krausr. F.K. camphell. K.,. iilld uting. J.. 1wn. I~ynamic
carting ;md gnw,h raulting ,*I the huwn (‘leek Gr;lhen (‘olnpler:
C;irh,,nilemui-Prr,,,ian Peace KiWl Emhayllwll. westurn (‘anad;,. in
O’Cooncli, s.c and Rdl. .I.% Edr.. Gcolo~y ,,I lhe Peace KiVU Alrll
Bull. can. l’elr. (ied 38.4 I 15~145.
“dl. I.5 illld Mc(‘dhm. R.li.. 1’)w. 11, ii,” ilW\S in ,,,c PPBm RiVCl Arc,,
ilrca. waswrn Canad;,. i,/ O‘(‘<,mell. S.C. iwd Hell. J.S.. EdS.. Ce,,hpy of
the Pcacc KiW Arch: Hull. can. Pm Garl. 3XA. 270.ZXI.
c;in,. D.,.. 19x*. Reg’mal tlmclure and ‘levckqwlwn~ or the Pexe River
AVCh. A,hcrta: il Palcomic f.drd-rin \ywn’!: Rul,. ~‘S,~. Pew GCd 36.
2x4.2%
Davis. S.D. il”d Fmhlich. c.. 141)X
Did lor will) thirl injeclion cmx wrtlb
quakes’! crilcci~ fb ic mimal ;issrsrmen~: Seis. Ke*. L.ac. 64. 207-221.
ixntl Pennington. W.IL IWU, Induced seismic d~+~xm;iti~m in 0~
Cog&d oil lield of west Tcwr: Hull. Seir. Sm. Am. 79. 1477-1&U
delrlillr. C.. 195X. Fle~Mirrissippi;w hiitwy of lhc Pcacc River Arch J.
Alla. sot. Pru. Cd cihl-hX.
Doser. L.1.. Raker. M,K., Luo. iv.. M;wrwjoin. P.. H;~lltrkxx L..
Kingwell.
1.. Diw. H.L.. and Kaip. G.. IWZ. ‘The mu so simpls r&uionship hrlween
seirmicily and oil pmduclion: Pure Appl. Cieophyr. 139. 4X I.
_._~~, ~~~ and Mason. D.B.. 1991. Seismicity in the W;cWink gas
field, Dclawarr Basin, west ‘I-was, ;and its I-ehtkmrhip to jpetrdew~, pw
ductinn: Bull. Seis. SW Am XI. W I-W>.
Gendrwi,,. ” .,.. timer, R.H. md tl;i\ugiiun, H.S.. IW?. ,nducrd Lw”b
quakes il, a polarh Illi” “CX Sikml<m. Canada C;m I. ,i:lC,h sci. 14.
466~475.
Grario. J.R. and Wittlinger. C.. 1990. ‘Ten yean of seiblic ~mcmiroring 0ve1
3 gas field: Bull. Scis. Sot. Am. IUJ, 4X-472,
Haimum. B.C. md Fairhunt, C., 1970, IrhCnt stress dctsmina~im~ at great
dqm hy llleil”I or hydraulic kxt”ring. i,, S<m,Lwm. \I’,. Ed.. Rock
mechanics: theory and pl-aclice: Pwcrrdinpr of the I Ilh Sympmiom cm
Rock Mrchi~nics: Am. Inst. Min. Eq.. New Ycrrk. W-5X4.
Han. B.S. and Plinl. A.G.. 1990. Upper Cwlaceous wwping imil fiwll ~nxwe~
,nent 011 Ihe southel-1, llank <If ,be Peace River Arch. Alhcvta ir,
~YConncll. S.C. Wld Hell. J.S., Edi.. Geok~py d ,,w Psxe RiVW Arch:
Buli. csn. Petr. Gd ,X.4 IW I’>S.
Healy. J.H., Ruhey. W.W.. Griggs. D.‘I’. and Raleigh. C.B.. IW+. The
“en”erranhq”nkcs: Science 161. 1301-131”
Hubhrrt. M.K. and Ruhey, W.W.. IWJ, Role of hid ~ILI~SLICC in wcrthrnsl
faul,ing: Hull. Crd. SW A,,,. 70. I ,S~2,K,.
Kry. P.R. and Cil-onwh. J.M.. lW3, 10 iim 51r~sses and hyfkidic tlscwcing
in Ihc deep hasin: J. Cm Pew Tech. 32. no IO. il-25.
Lawie. D.K 19% .I’he Pwcc Rives Arch during Missirsippi:m iand Psrmw
Pe”nsylvnnian time: J. Ah sot. l’C,l. GC<,,. 6. 6’1.74.
Leek ,>.A.. Stnnil;,“d. MR. and HilyCI. B .,.. 14w. Repional ,,,a,” or ,hC
Alhian Pracc Rim and Ikwcr Shaftrshwy I)~matim~s irn the Peace River
Arch. owthwcrtecn Alherl;l ;md nwthe;mlern British Columbia. in
O‘Connell. S.C. and Rell. J.S.. Eds.. C~olcrgy of the Pc;m River Arch:
Bull. can. IJetr. Geol. .%A. 176-,X9.
Legfen. S.. Young. K. and F.ihcnrclwt/. D.. 1993, Facieh or 165 Hrlloy
F’~rnwlioo (Loarr Fcrmi;~nj. nonhr;~~~err~ Hcitish Cdumhia. in Cow wrrk~
shop guidehook. Carhonifemur 10 Jurarsic. Pangu. August 15~ 19. IU91.
Calgary. Alk”ii. cmadn: can. sot. Fwl-, ckxrl.. 88~ I,,X.
McMechan. M.E.. IWO. Upper Prwxwoic to Middle Can~hri;w hixory 01
Ihc Peace River Arch: rvidencu lnm lhe Kwky Mwnmin*. iri O‘Cunncll.
s.(‘. and Bell. J.S.. Eds.. Geology of “x PWCC Rives Arch: Hllll. c’nn.
Pm Gcol. 38A. 36-44.
Milne. W.C.. 1970. The Snipe Lake. Alhe~ti~. ~wlhqui&x of March X. 1970:
can. 1. F.anil sci. 7, ,x34-I%,.
Nwfimi. B.S.. 1990. Chdovicinn and Silurian rtratigraphy. pak~peography
and dspositiond hismry in the Pciioe Rivrv Ad :are;t. Alhem md Rrilish
Columbia. in O‘Connell. S.C. and Hell. J.S.. FA.. Gedopy 0I lhc Pcacc
River Arch: Rull. Can. Petv. Ged %A. 45.54.
O’Connell. S.C. snd tkll. J.S.. El.. IUW. Cedogy 01 the Pcacc River
Arch: Ball. can WC. Grol. MA. 2x I p.
_, Ilix. G.R. ;md Ha~lay. J.E.. 1940, ‘I-he wigin. histwy and qimal
blrllC,“lai dr”el,,pmcnt of the Peace River Arch. western C;mKla. i,,
O‘Connell. S.C. and Bell. J.S.. F.dh., Gx~logy d the Pcxc River Axh:
Bull. cm. Par. Gee,. 38A. 4~24.
,ll”C ,4Y,

R.R. HORNEK. J.E. “ARCLAY ;hld J.M. MACKAF
Pwar. J.W.. PliCC. K.A. and Mc~r,l\s;ln. R.G.. IW2. ThC WCSlCrn cm;h
Scdimenury B;isitl. in The wolu~ion (I/ rrdimrnrary hasins: Roy. SW
Lond<m. Phil. Tr;UlS. A305 lh%, X?.
Pwlton. T.P.. Ti,lcIIIo~c. J. uld Ddhy. G.. 1440. lur;iisic s,riltil 01 nmh-
weitem (and wcrl&cenkal~ Alhcm and tndxartcm Rrilirh Culumhia. in
O‘Connell. S.C. and Rdl. J.S.. Fads.. Geology of rhr Prace River Arch:
Bull. Can. PLY,. c;COl. 38A, 150-175.
Pugh. D.C.. 1073. Suhsurfxce Lower Paleozoic rmligraphy in nw~hsm and
cen,m, Alhal-,n: Cirol. sot. runada, f?,prr 72.12. 54,‘.
Raleigh. C.R.. tkdy. I.H. and Brrdrhocf,. ,.I)., 1972. Faulling and Crustill
s,vxs 8, Rangely. C,~li>r;do. i,, Heard. KC.. Fhrg. I.Y., C;,,,rr. N.L. illld
Kayleigh. C.R., Edr.. Fkw and fiacturr of TOCLI ,mgg\ “<d”IIIc,: A,,,.
Gcl,phys. union. Geophys. Mwx,. 16. 275-2.84.
..~. ~ nnd ~, 1976, An cnpcrimrnt in earthquake cmtm a!
Rangely. Colorado: Science 191. 1230~1217.
Kshoihr. c .J., Kanaswich. C.R. and Nyland. E.. 14XI. source parmcwn
from shallow cwnls in Ihe Rocky Mountain House earlhquakc swiml:
cm J. Ear,,, sci. 19, ‘XW’~1X.
RW\. C.M.. ,w,, Drsp cr11st and hasenwn, l,,“Cl”rr (II ,hr Peace River
Arch region: consIr:3int* on mechanisms of lonna~im. io O‘Connell. S.C.
and Hell. J.S.. Edh.. Cmhgy <II ttr Prxs River Arch: Bull. Can. Par.
Crd 3XA. 25.35.
~ snd Svqhrmon. K.A.. I L)W, Cryswllinr hasemsnt: Ihc I)rundatinnr
uf Wsrtcrn Canada Scdimrtmry Rabin. in Rickem. H.Il., Ed.. Wehicrn
Cawda Scdilncnury Basin -- a case hiwry: Cm SW Prtr. Geol.. Spec.
Pub. NC,. 30. Calgary. Alhrrw. 33-45.
Segal. P.. IWI. Ewthqoaker Iriggcred hy lluid rxtraclim Gcdqy 17.
‘)4?~‘)46.
Sikahonyi. L.A.. 1%7. Majw Icclonic trends in the prairie region cfCanad;c:
J. Al,B. sot. Par. Gro,. 5. 23~2X.
_ and Rodgrrr. W.J.. !1)59. P*lroa~ic trctmicr and scdimenlaiion ill the
nwlhcm hallollhe West Canak,” “mi”: J. All& S,r. Petr. Gmi. 7, 191-216.
Slcphenrun. R.A.. Zrlt. C.A.. Ellk, R.M.. tk,jnal. %.. Mnrclk-I‘Huissicr. P..
Mm”. R.F.. Nonhey. I).,., WCS,, G.F and Kmarcwich. E.R., Iwl. crust
and upper ~niilntlr wucturc and Ihe origin of the Peace River Arch: Bull.
(‘an. Pal-. Gel,,. 37.224-m
Wrtmikr. I<.J.. IWh. Emlhquakrs ncilr Kc&y Mowmin Hou*r. Alkna. and
their rclalionship to gas productiw rslcililies: can. ,. tarill sci. 23. 172-,x,.
Willima, G.K.. 1%X. Inllwnw d the kace Kimr Arch m Mcsomic
IInl~l: J. AI,;,. SW P‘s mwl. 6. 71.8,.
Ycrker, K.F. and Cartlc. R.O., 11)X Srismiciry and faulting attrihuwhlr to
fluid extraction Eng. Cieol. 10, IS I-167.
Zelt. C.A.. IW9. SCi\,lliC ~Ir”EI”IC of the U”\I and upper lllantle in Ihl‘
Peace River Arch region Ph.D. thr*is, Univ. of Hrilirh Columbia. 223p.
~ and Ellir. R.M.. IWd. Seismic rlrucure of the cwst and upper ~mil~l-
ile iI3 the Pracr River Arch repicm. Canxla: J. Ceophyr. Kes. 94. RS.
ST?‘)-5744.
c,,i<i 50 J”“C 14w