-
ABSTRACT
The Beekmantown Group in the Quebec Low-lands was deposited as
part of an extensive EarlyOrdovician coastal and shallow marine
complex onthe eastern margin of the North American craton.The
Beekmantown is stratigraphically equivalent tothe Beekmantown,
Knox, Arbuckle, and Ellen-burger rocks of the United States, and is
subdividedinto two formations: the sandstone-rich TheresaFormation
and the overlying dolomite-r ichBeauharnois. Dolomites of the
Beekmantown pro-vide an important exploration target in both
theautochthon and the overlying thrust sheets of theCanadian and
U.S. Appalachians.
The reservoir potential of the autochthonousBeekmantown Group in
the Quebec Lowlands canbe determined from seismic data, well logs,
cut-tings, and petrographic analyses of depositional anddiagenetic
textures. Deposition of the Beek-mantown occurred along the western
passive mar-gin of the Iapetus Ocean. By the Late Ordovician,the
passive margin had been transformed into aforeland basin. Faulting
locally positioned UpperOrdovician Utica source rocks against the
Beek-mantown and contributed to forming hydrocarbonreservoirs. The
largest Beekmantown reservoirfound to date is the St. Flavien
field, with 7.75 bcf oforiginal gas (methane) in place in fractured
and pos-sibly karst-inf luenced allochthonous dolomiteswithin a
thrust-fault anticline.
The Beekmantown below the thrust sheetsforms a
northward-thinning wedge of peritidal and
subtidal deposits. Seven major depositional unitscan be
distinguished in cuttings and correlatedwith wireline logs. Most of
these units form north-ward-thinning sediment wedges and were
deposit-ed on a gently dipping ramp. Quartz sandstonesdominate
updip, whereas shallow, subtidal, pelletalto skeletal limestones
dominate downdip. Awidespread blanket of shaly dolomite is the
upper-most unit of the Beekmantown, but is of poorreservoir
quality.
Dolomites in the Beekmantown contain vuggy,moldic,
intercrystalline, and fracture porosity. Earlyporosity formed at
the top of the major deposition-al units in peritidal dolomites;
however, much ofthis porosity was later filled by late-stage
calcitecement after hydrocarbon migration. Thus, a key tofinding
gas reservoirs in the autochthonousBeekmantown is to define
Ordovician paleostruc-tures in which early and continuous
entrapment ofhydrocarbons prevented later cementation.
INTRODUCTION
Quebec is commonly considered to be relativelybarren of
hydrocarbons, but significant oil seepswere noted in the province
at least as early as themid-1800s, mainly at the end of the Gaspé
Pen-insula as reported by Logan (1846). A particularlyintriguing
aspect of Logan’s report is that manyyears before the importance of
anticlines in trap-ping hydrocarbons became widely recognized,
henoted that the seeps were most common along thecrest of an
anticlinal fold in rocks now known to beSilurian in age. Following
the discovery of oil bydrilling in Pennsylvania in 1859, the seeps
nearGaspé became a focus for exploration, and twoshallow wells were
drilled in 1860 (Hume, 1932).Unfortunately, neither yielded more
than a trace ofoil, and the drilling of dozens of additional wells
onthe peninsula in the years since has met with simi-larly
disappointing results.
During the late 1800s and early 1900s, explo-ration efforts in
Quebec moved westward along theSt. Lawrence River past Quebec City.
There, farmersencountered significant amounts of gas in shallow
513AAPG Bulletin, V. 79, No. 4 (April 1995), P. 513–530.
©Copyright 1995. The American Association of Petroleum
Geologists. Allrights reserved.
1Manuscript received June 1, 1994; revised manuscript
receivedNovember 15, 1994; final acceptance December 6, 1994.
2Talisman Energy Inc., 6611 Longmoor Way S.W., Calgary, Alberta
T2P3R2, Canada.
3Consulting Geologist, 701 Harlan, #E-69, Lakewood, Colorado
80214.We thank Les Beddoes, Jeff Chisholm, and Henri Lizotte for
useful
discussions and suggestions on early drafts of this manuscript.
AAPGreviewers Rick Major, John Hobson, and Kenneth Stanley also
providedmany useful comments. Thanks to Bow Valley Energy, Inc.,
and TalismanEnergy, Inc., and their partners for permission to
publish this paper. BruceBailey prepared lithology logs for all
study wells. The authors gratefullyacknowledge the financial and
technical cooperation of the SociétéQuébécoise d’Initiatives
Pétrolières (SOQUIP).
Gas Reservoir Potential of the Lower OrdovicianBeekmantown
Group, Quebec Lowlands, Canada1
John C. F. Dykstra2 and Mark W. Longman3
-
wells drilled for water. Several enterprisinglandowners in the
area were able to recover enoughgas for home heating and domestic
use. The mostsignificant shallow gas discovery was made in 1955at
Pointe du Lac field, located near Trois Rivières.Initially, this
was another example of a local farmerseeking gas to heat his home,
but his test wellencountered methane-filled unconsolidated
glacialsands encased in clay at a depth of about 100 m(320 ft) and
blew out. After six months, the wellwas finally controlled, and for
the next 35 yr thefarmer used the gas to heat his home and those
oftwo of his children.
The blow-out stimulated additional interest inthe shallow gas
field, and Pointe du Lac field wasextensively developed between
1962 and 1972,producing about 2.5 bcf before being converted toa
gas storage reservoir. Based on thermal matura-tion indices,
Bertrand and Dykstra (1993) suggest-ed that the shallow gas was
generated in theOrdovician Lorraine and Utica formations andseeped
upward into the glacial sands within thepast 80 k.y.
The discovery of gas fields in the Alberta over-thrust belt, in
combination with the gas found atPointe du Lac, spurred exploration
of the Ordovi-cian section in the Appalachian overthrust belt ofthe
Quebec Lowlands. That search resulted in thedrilling of about 40
deep (>1500 m or >5000 ft)wells, and the discovery of St.
Flavien field in 1972.This field, which consists of two producing
wellsand six dry holes, has now produced about 5.7 bcfof gas out of
7.75 bcf in place in an allochthonous(thrust-faulted) section of
Beekmantown in a hang-ing-wall anticline. St. Flavien is currently
in the pro-cess of being converted to a gas storage
reservoir.Reservoir rocks are dolomites with secondaryporosity in
which possible karst-related porosityhas been enhanced by faulting
and fracturing (B.Bailey, 1992, personal communication). A numberof
exploration companies joined the search for sim-ilar allochthonous
Beekmantown reservoirs duringthe 1970s, but the dozens of wells
drilled to testthe thrust sheets resulted in no other
commercialdiscoveries.
A new approach to exploration for Beekman-town reservoirs in the
Quebec Lowlands began in1989, following the discovery of
OrdovicianArbuckle dolomite pay in Wilburton field inOklahoma in
1987 (Petzet, 1992). The Arbucklereservoir rocks at Wilburton field
are stratigraph-ically equivalent to the Beekmantown Group
inQuebec, and were originally estimated to contain asmuch as 600
bcf of gas (Petzet, 1990). The gas istrapped in a structurally high
fault block beneathseveral thousand meters of thrust sheets.
UsingWilburton field as an analog, it appeared that
theautochthonous Beekmantown section in the
Quebec Lowlands, which has been similarly faulted,should also
offer promising exploration targets.
This study describes the reservoir potential of theautochthonous
Beekmantown Group in the QuebecLowlands. We emphasize understanding
the deposi-tional units and controls on porosity that
wouldinfluence reservoir development. The gas produc-tion from St.
Flavien field, the presence good poros-ity locally within the
autochthonous Beekmantownsection, and the common occurrence of
bitumenand gas shows all suggest that the Quebec Lowlandscontains
deeply buried gas reservoirs.
Specific objectives of the study were to (1) definethe major
depositional facies in the Beekmantown,including grain-rich versus
mud-rich facies, relativeshaliness, and the degree of early
dolomitization; (2)interpret the diagenetic history of the
Beekmantownwith particular emphasis on the origin and texturesof
the dolomites; and (3) describe the origin and dis-tribution of
porosity and pore types.
STUDY METHODS
The wells penetrating the Beekmantown section,and particularly
the autochthonous Beekmantownbelow the thrust sheets, in the Quebec
Lowlands(Figure 1) provided the basic data for this study.Wireline
logs proved to be of limited use in differen-tiating lithologies,
so detailed lithology logs wereprepared from the cuttings for each
well. A fewcores from the Beekmantown were also availablefor study.
Selected intervals were sampled and 125standard petrographic thin
sections were prepared.These thin sections were used to identify
the litho-logic variations in the Beekmantown, as well as todefine
depositional and diagenetic fabrics. Specialemphasis was placed on
sampling those intervalsreported to have yielded hydrocarbon shows
andthose with visible bitumen stain. The samplesselected for study
range in depth from 1356 to 4120m (4449 to 13,517 ft). The
petrographic data werethen integrated with the wireline and
lithology logsto subdivide the Beekmantown Group into
sevendepositional units. Interpreting these units and
theirassociated porosity forms the basis of this study.
The depositional and structural interpretationspresented in this
paper are supported by the in-house evaluation of over 2000 km
(1200 mi) ofrecently reprocessed seismic data and 917 km (570mi) of
recently acquired seismic lines.
TECTONIC EVOLUTION OF THE QUEBECLOWLANDS
Understanding the depositional and tectonic his-tory of the
Quebec Lowlands area is important to
514 Beekmantown Group Gas Potential
-
understanding the nature of the Beekmantown andits potential for
reservoir development. St. Julienand Hubert (1975), Tremblay
(1992), Dykstra(1993), and Shaw (1993) all provided
regionalinformation on the rocks and structural history ofthe
Quebec Lowlands area. Their work is com-bined with our own to
provide the following sum-mary of the tectonic evolution of this
interestingand complex region.
Most of what is preserved in the sedimentaryrock record was
deposited during the Cambrianand Ordovician (Figure 2).
Tectonically, the evolu-tion of the basin can be divided into six
majorepisodes: (1) Middle to Late Cambrian, when nor-mal faulting
of the passive margin occurred andsandstones of the Potsdam Group
were beingdeposited; (2) Early to Middle Ordovician (Ibexianto
Mohawkian), when gentle subsidence allowedwidespread deposition of
the carbonate-r ich
Beekmantown, Chazy, and Black River rocks; (3)Middle to early
Late Ordovician (late Mohawkianto early Cincinnatian), when initial
continentalconvergence resulted in the onset of subductionfar to
the east as the limestones of the Trentonwere being deposited in
the Quebec Lowlands; (4)early Late Ordovician (early Cincinnatian),
whencollision of continental plates and the onset of dis-tal thrust
faulting provided a f lood of shale (theUtica Shale) across the
Quebec Lowlands area; (5)Late Ordovician (late Cincinnatian), when
proxi-mal thrust faulting of the lower Paleozoic sectionoccurred as
the Lorraine f lysch and Citadelwildf lysch were being deposited;
and (6) post-Ordovician, which has been characterized mainlyby
relative tectonic stability in the QuebecLowlands region. Each of
these events is illustratedin Figures 3 and 4, and described
briefly in the fol-lowing paragraphs.
Dykstra and Longman 515
����QUEBEC
QUEBEC CITY
175A'
167
St. Flavien Gas Field / Storage
Pointe du LacGas Field / Storage
����
CAN
ADA
U.S.A.
QUEB
EC
QUEBEC
VERMONT
MAINE
MONTREAL
NEW YORK
��
��
A
161
187156
222158
185
163
165A
214A
162
157
166
DRUMMONDVILLE
NEW
HAM
PSHIRE
EasternCanada
Quebec City
Ottawa
Toronto
Ontario
Quebec
A-A'
Wells penetrating autochthonous Beekmantown or older section
(but no samples studied)Wells chosen for this study (with samples
studied)Other study wells not penetrating autochthonous
BeekmantownLocation of cross section
N 44˚ 30'
W 74˚ 00'
W 74˚ 00'
N 47˚ 30' W 70˚ 00'
N 47˚ 30'
W 70˚ 00'
N 44˚ 30'
26
Montreal
TROISRIVIÈRES
St. LawrenceRiver
�
Figure 1—Distribution of wells used in this study of the
autochthonous Beekmantown section in the Quebec Low-lands. Other
wells that penetrated only allochthonous Beekmantown in thrust
sheets are indicated with a squaresymbol. Line of Figure 6 cross
section is shown.
-
Normal Faulting of the Passive Margin
During Potsdam deposition in the Middle to LateCambrian, the
Iapetus Ocean was widening asLaurentia (paleo-North America) and
Baltoscandiawere moving apart (Scotese and McKerrow, 1991;Huff et
al., 1992). The area which later became theQuebec Lowlands was then
along the westerncoastline of the Iapetus Ocean. New oceanic
crustwas forming at a mid-ocean ridge far to the east(present-day
coordinates), much as it is forming inthe Atlantic Ocean today
(Figure 3A). The conti-nental margin was under tensional stress due
to theopening of Iapetus and formed a passive marginwith a series
of tilted fault blocks bounded bydown-to-the-basin and antithetic
faults. Significantthickness variations in the Potsdam are seen
onseismic lines, indicating that at least some faultswere active
during Potsdam deposition.
Sandstones of the Potsdam Group were derivedmainly from the
Laurentian highlands to the pres-ent-day northwest. Ephemeral sand
dunes probablylined the shore, but most of the preserved
Potsdamrocks were deposited in a shallow subtidal settingwhere
waves and currents reworked the sedi-ments. The net result was
deposition of a fairlymature, clean, quartz-rich to arkosic sand.
By theend of Potsdam deposition, sand covered virtuallythe entire
coastal margin complex, forming a nearlyflat plain that extended
many kilometers basinwardfrom the shoreline.
Subsidence of the Ramp
Deposition of the carbonate-rich Beekmantown,Chazy, and Black
River groups occurred during atime of relative tectonic stability.
These sequencesincrease in thickness very gradually into the
IapetusOcean, indicating that deposition occurred on anextremely
broad, very gently dipping ramp. Minorfluctuations in thickness are
attributed to subtletopographic highs and local sags. Carbonate
depo-sition occurred immediately following majormarine
transgressions, but siliciclastic sedimentswere transported onto
the ramp during times of rel-ative sea level lowstand. The nature
of this gentlydipping passive margin is shown in Figure 3B.
During the Early Ordovician when the Beek-mantown was deposited,
the North American cra-ton was located along the paleoequator with
theQuebec Lowlands area lying between 15 and 20°Slatitude (cf.
Lindsay and Koskelin, 1993). This
warm, tropical setting favored the deposition ofcarbonates, many
of which accumulated in veryshallow subtidal to peritidal settings.
Aspects ofLower Ordovician deposition on what has come tobe known
as the “Great American Bank” have beensummarized by Wilson (1993).
Distinctive featuresinclude the common occurrence of
upward-shoal-ing carbonate units ranging in thickness from 1–2m to
tens of meters, extensive dolomitization of theupdip parts of these
cyclic deposits, presence ofsandstone marker beds reflecting times
of relativelylow sea level, and a major unconformity (the
Saukunconformity) that terminated deposition at theend of the Early
Ordovician.
Onset of Crustal Convergence
After the deposition of the Black River Group,the ancient
continents of Laurentia and Balto-scandia started converging. This
movement isrevealed by faulting visible on regional seismic
sec-tions that show that the Trenton Limestone overly-ing the Black
River varies significantly in thicknessacross some faults. Many of
the passive margin nor-mal faults were reactivated during this
time.Houseknecht (1986) reported similar patterns offaulting on the
southern margin of the NorthAmerican craton at the onset of thrust
faulting inthe Ouachita thrust belt during the Pennsylvanian.Figure
3C, based in part on the work of Jacobi(1981) and Stockmal et al.
(1987), illustrates ocean-ic crust being subducted beneath
Baltoscandia. Amajor marine transgression occurred during thistime,
drowning the continental margin as far inlandas Ontario and burying
the shallow-water sedi-ments of the Black River beneath deeper
waterlimestones of the Trenton.
Crustal Collision and Distal Thrusting
During upper Trenton to lower Utica deposition,we interpret from
the presence of a restrictedfacies that Laurentia and Baltoscandia
began collid-ing. This collision marked the first phase of
theTaconic orogeny, which mainly involved thrustfaulting of
deep-water rocks along the edge of whatwas once the continental
margin of Laurentia(Figure 4A). Distal flysch sediments from the
thrustsheets were deposited, causing the TrentonLimestone to grade
eastward and upward into theUtica Shale. Once the basinal thrust
sheets had
516 Beekmantown Group Gas Potential
Figure 2—Stratigraphic column for the Quebec Lowlands. The study
area does not include the shaly Levis and Que-bec City formations,
but they are shown here for the sake of completeness. Adapted from
Dykstra (1993).
-
Dykstra and Longman 517
������������������������������������������������������
������������������������������������������������������
��������������������Unit 6
Organic-Rich Shales Potential Organic- Rich Shales
���������������
���������������
��������������������
�������������������
��������������
���������������������
TectonicEvents
QUEBEC LOWLANDS STRATIGRAPHYN. W. S. E.
Coastal Onlap Curve
Landward Oceanward
Hydrocarbon Occurrences
Pointe du LacTertiary
Cretaceous
Jurassic
Holocene
Cen
ozoi
cM
esoz
oic
Triassic
Permian
PennsylvanianMississippian
Devonian
Silurian
Scale Changex 25
Subsidence of Ramp due toLoading ofThrust Sheets
OR
DO
VIC
IAN
CIN
CIN
NA
TIA
NM
OH
AW
KIA
NW
HIT
ER
OC
KIA
NIB
EX
IAN
LAT
EM
IDD
LEE
AR
LY
Precambrian
Pro
tero
zoic
AppalachianAllochthon
Devonian limestone deposited but completely eroded. Onlyremnants
in breccia remain
������
������
����������������
����������������
��������
������������������������
����������������
Devonian Limestone
Utica Grp
Lorraine / Queenston
Sauk Unc.
Chazy Grp
Black River GrpTrenton Grp
Unit 6
Unit 5 Unit 4
Unit 2
Unit 3
Potsdam GrpCairnside Fm
Potsdam GrpCovey Hill Fm
Pot
sdam
Gro
upB
eekm
anto
wn
Gro
up
Villeroy(FracturedShale)
St. Flavien
St. Simon(CO2)
*Absolute Ages From AAPG Cosuna Charts
570
540542
550
560
520
530
500
510
515
480
485
490
460
470
440
450
455
200
355
405
425
250
290
67
140
0
Atlantic PassiveMargins
Initiation of
Atlantic Rifting
Allegheny Orogeny
(No effecton Quebec)
Acadian Orogeny
Cretaceous Intrusives Devonian Carbonates PresentWhen
Monteregian Intrusives Formed
St. PierreSand Glacial Till
Champlain Clay
TA
CO
NIC
OR
OG
EN
YP
AS
SIV
E W
ES
TE
RN
MA
RG
IN O
F IA
PE
TU
S O
CE
AN
PA
LEO
ZO
IC
Trenton
Blk River
��������������������
Chazy
Sauk Unconformity
Beauharnois
Unit 5
Unit 4P
hili
psb
urg
Sta
nb
rid
ge
Sill
ery
Gro
up
Le
vis
[ ]
St. Rosalie Grp
Utica
MMyr BP*
CA
MB
RIA
N+ ++
++
+ +++
++
+ + +++
++
+ +++
++
++
Unit 2
Unit 1
Unit 7
Qu
eb
ec
Cit
yO
rle
an
s G
rou
p
Lorraine
Unit 3Theresa
Citadel
Unit 1
Unit 7
Potsdam Grp
Potsdam Grp
Queenston
Cairnside Fm
Covey Hill Fm
[ ]
[ ]
[ ]
[ ]
[ ]
-
518 Beekmantown Group Gas Potential
(A)
NO
RM
AL
FA
UL
TIN
G O
F P
AS
SIV
E M
AR
GIN
Dep
ositi
on o
f: P
otsd
am S
ands
Ero
sion
of H
ighl
ands
(La
uren
tian
Mou
ntai
ns)
Gre
nvill
e
Yam
aska
Fau
lt
Sea
Lev
el
Laur
entia
Iape
tus
Oce
an
Con
tinen
tal S
helf
Edg
e M
id-I
apet
us S
prea
ding
Rid
ge
Acc
retio
n
(B)
SU
BS
IDE
NC
E O
F T
HE
RA
MP
Dep
ositi
on o
f: B
lack
Riv
er
Cha
zy
Sea
Lev
el
Cla
stic
Inpu
t Red
uced
S
ea L
evel
Iape
tus
Oce
an
Sag
ging
Dor
man
t Mid
-Oce
anic
Rid
ge
Laur
entia
(C)
ON
SE
T O
F C
RU
ST
AL
CO
NV
ER
GE
NC
E
Tre
nton
Me
lan
ge
We
dg
e
with
Op
hio
lite
s
For
earc
Bas
inIa
petu
s O
cean
Subduct
ion
Laur
entia
Ba
ltosc
an
dia
Bac
karc
Bas
in
Act
ive
Vol
cani
c A
rc
Pot
sdam
Bla
ckR
iver
/Cha
zyB
eekm
anto
wn
Bal
tosc
andi
a
Bal
tosc
andi
a
Bas
inal
Sha
les
SA
UK
UN
CO
NF
OR
MIT
Y
Bee
kman
tow
nB
asin
al S
hale
s
Yam
aska
Fau
lt
* T
ecto
nica
lly q
uies
cent
*
Gro
wth
sta
rtin
g al
ong
the
Yam
aska
Fau
lt af
ter
deve
lopm
ent o
f the
Sau
k U
ncon
form
ity
* S
aggi
ng o
f mid
-Iap
etus
rid
ge c
ausi
ng s
ome
exte
nsio
n on
the
ram
p,
re
sulti
ng in
sub
tle s
aggi
ng in
loca
l are
as a
long
the
ram
p.
* G
row
th a
cros
s fa
ults
due
to e
xten
sion
and
con
tem
pora
neou
s
depo
sitio
n*
Ver
y lit
tle r
emai
ning
topo
grap
hy o
n th
e ra
mp
by th
e en
d of
Pot
sdam
dep
ositi
on
* R
eact
ivat
ion
of m
any
of th
e or
igin
al p
assi
ve m
argi
n fa
ults
due
to
cr
usta
l ben
ding
rel
ated
to th
e su
bduc
tion
of th
e oc
eani
c cr
ust
* C
ontin
ued
grow
th a
long
the
Yam
aska
Fau
lt
Oce
anic
Cru
st
Oce
anic
Cru
st
Oce
anic
Cru
st
Dep
ositi
on o
f: T
rent
on L
imes
tone
B
asin
al S
hale
s
Fig
ure
3—
Tec
ton
ic e
volu
tio
n o
f th
e Q
ueb
ec L
ow
lan
ds
area
fro
m t
he
Cam
bri
an t
o t
he
Mid
dle
Ord
ovi
cian
. (A
) N
orm
al f
ault
ing
of
pas
sive
mar
gin
,(B
) su
bsi
den
ce o
f th
e ra
mp
, (C
) o
nse
t o
f cr
ust
al c
on
verg
ence
.
-
moved far enough, a restricted basin formedbetween the thrust
front and the shoreline. Therestriction of the basin, combined with
the pres-ence of abundant organic material, contributed tothe Utica
Shale becoming a rich petroleum sourcerock locally. Gradual loading
of thrust sheets uponto the ramp reactivated many of the
previouslyformed normal faults. In some places, this
faultreactivation juxtaposed Utica source rocks againstBeekmantown
reservoir rocks.
Thrusting of Carbonates
As the collision of the crustal plates continued,carbonate rocks
of the shallow-water ramp wereincorporated into the thrust sheets
(Figure 4B).The Utica Shale was conformably covered by thesilts and
fine-grained sands of the Lorraine flysch,which prograded through
time onto the craton.The Queenston Group is a coarser equivalent
ofthe Lorraine, which was deposited closer to thethrust front.
Shales of the Lorraine are not as rich inorganic matter as those of
the Utica because rapiddeposition diluted the available organic
matter. Asthe thrust front advanced, the thrust faults
becameshallower and involved younger rocks. Eventuallythe Lorraine
Group itself became imbricated alongthe foreland thrust belt.
Silurian to Present
Virtually no post-Ordovician rock record hasbeen preserved in
the Quebec Lowlands, but thearea was once covered by younger
strata. Silurianrocks are completely unknown, but Devonianrocks are
known from inclusions in the CretaceousMonteregian intrusives,
which occur on the islandof Sainte-Hélène in Montreal (Globensky,
1987).These intrusions produced contact metamorphismwhen they
passed through the Ordovician carbon-ates that resulted in the
release of carbon dioxideinto porous intervals in the Beekmantown.
As indi-cated by carbon isotopes analysis (Bertrand andSavard,
1992), it was this type of carbon dioxidethat was tested in one of
the deepest and southern-most wells in the Quebec Lowlands (St.
Simon 1Awell) at a depth of 4110 m (13,500 ft). Similar con-tact
metamorphism of carbonate strata reportedlyaccounts for the
majority of the world’s naturallyoccurring carbon dioxide trapped
in subsurfacereservoirs (Farmer, 1965).
Figure 4C is a present-day interpretation of theQuebec Lowlands
area based on regional seismiclines and work by St. Julien et al.
(1983). TheOrdovician rocks of the Quebec Lowlands areunconformably
overlain by glacial sediments that
were deposited from approximately 80,000 to lessthan 9500 yr ago
(Lamothe, 1989). Fine-grained flu-vio-deltaic sands were deposited
in glacial LakeChamplain as the ice sheets retreated to the
north-west. Several accumulations of methane, includingthe Pointe
du Lac field, have been discovered insand lenses encased in the
Champlain Clay.
STRATIGRAPHY
Beekmantown rocks were first described inQuebec and Ontario as a
calciliferous sandstone bySir William E. Logan in 1864. Clarke and
Schuchert(1899) were the first to assign the name“Beekmantown
Group” to the type section in NewYork state. Based on extensive
field work by Ells(1896) and Ami (1900), Raymond was the first
touse the term “Beekmantown” in Canada in 1913.
The stratigraphic position of the LowerOrdovician
(Ibexian–Whiterockian) BeekmantownGroup is shown in Figure 2. The
Beekmantown istraditionally subdivided into two formations: alower
interval rich in quartz sandstones, called theTheresa Formation,
and an upper interval consist-ing of relatively clean dolomites,
named theBeauharnois Formation. The Theresa has beeninterpreted as
resting conformably on the UpperCambrian Potsdam, but the presence
of a peritidaldolomite interval at the base of the Theresa in
thedowndip St. Simon 1A well suggests that this con-tact is
unconformable in updip areas to the north.The contact between the
Theresa and Beauharnoisappears to be conformable. The contact of
theBeauharnois with the overlying Chazy Group is awidespread
unconformity commonly referred to asthe Sauk unconformity (also
known as the St.George unconformity in Newfoundland; see Knightet
al., 1991).
Dolomites are the primary exploration target inthe Beekmantown
for two reasons. First, they con-tain good porosity locally.
Second, they are strati-graphically equivalent to similar dolomites
in theArbuckle and Ellenburger groups in the southernUnited States;
these groups have yielded largeamounts of hydrocarbons (e.g., see
Holtz andKerans, 1992; Bebout et al., 1993). Furthermore,dolomites
of the Beekmantown form the reservoirfor the St. Flavien gas
field.
The Lower Ordovician section in the QuebecLowlands represents
cratonic deposition in periti-dal environments ranging from exposed
tidal flatsto shallow-marine environments. Deeper waterwas to the
south and east where the ramp carbon-ates grade into dominantly
shaly facies (this facieschange occurs outside the immediate study
area).Well-rounded and windblown quartz sand grainsderived from the
Laurentian highlands, part of the
Dykstra and Longman 519
-
520 Beekmantown Group Gas Potential
(A)
CR
US
TA
L C
OL
LIS
ION
AN
D D
IST
AL
TH
RU
ST
ING
Dep
ositi
on o
f: U
tica
Sou
rce
Roc
k Utic
aS
ea L
evel
Dor
man
t Vol
cani
c A
rc
Iape
tus
Oce
an
Col
lisio
n of
Con
tinen
ts
(B)
TH
RU
ST
ING
OF
CA
RB
ON
AT
ES
Lorr
aine
Sea
Lev
elS
ea L
evel
1
2
(C)
PR
ES
EN
T-D
AY
SC
HE
MA
TIC
INT
ER
PR
ET
AT
ION
Sch
emat
ic In
terp
reta
tion
alon
g S
eism
ic L
ine
200
(Ada
pted
from
St.
Julie
n et
al.,
198
3)A
fter:
* I
mbr
icat
ion
of L
orra
ine
Fly
sch
*
Dev
onia
n de
posi
tion
and
eros
ion
*
Aca
dian
Oro
geny
(D
evon
ian)
* M
onte
regi
an In
trus
ives
(C
reta
ceou
s)
*
Ter
tiary
ero
sion
Pro
to L
aure
ntia
Pro
to B
alto
scan
dia
Aca
dian
Def
orm
atio
n (D
evon
ian)
Atla
ntic
Pas
sive
Mar
gin
Late
Tria
ssic
Ope
ning
St.
Dan
iel O
listo
stro
me
(with
Oph
iolit
es)
Gua
dalu
pe F
ault
Asc
ot-W
eedo
nV
olca
nic
Arc
Ass
embl
age
St.
Wen
cesl
as 1
(p
roje
cted
)S
t. F
lavi
en 3
Cha
udie
re K
lippe
Lorr
aine
N
appe
s
St.
Cro
ix 1
Can
adia
n S
hiel
d(G
renv
ille)
2 1
* T
hrus
ting
of d
eep-
wat
er s
hale
s*
Con
tinue
d re
activ
atio
n of
man
y of
the
orig
inal
pas
sive
mar
gin
fa
ults
and
form
atio
n of
new
nor
mal
faul
ts d
ue to
dis
tal l
oadi
ng
of
thru
st s
heet
s*
Res
tric
tion
of th
e ba
sin
* Im
bric
atio
n of
dee
per
wat
er fa
cies
1 a
nd la
ter
th
rust
ing
of c
arbo
nate
faci
es 2
ont
o th
e ra
mp
* F
orm
atio
n of
new
nor
mal
faul
ts d
ue to
the
load
ing
of th
rust
shee
ts a
long
with
con
tinue
d re
activ
atio
n of
orig
inal
pas
sive
mar
gin
faul
ts
Dep
ositi
on o
f: L
orra
ine
Fly
sch
Cita
del W
ildfly
sch
Imbr
icat
ion
of
Car
bona
te F
acie
s
Imbr
icat
ion
of
Bas
inal
Fac
ies
Fig
ure
4—
Tec
ton
ic e
volu
tio
n o
f th
e Q
ueb
ec L
ow
lan
ds
area
fro
m t
he
Mid
dle
Ord
ovi
cian
to
th
e p
rese
nt.
(A
) C
rust
al c
oll
isio
n a
nd
dis
tal
thru
st-
ing,
(B
) th
rust
ing
of
carb
on
ates
, (C
) p
rese
nt-
day
sch
emat
ic i
nte
rpre
tati
on
.
-
Canadian shield, are abundant in the TheresaFormation and occur
locally in the Beauharnois andChazy intervals.
Lower Ordovician dolomites in the United Stateshave been studied
much more extensively thanthose of the Quebec Lowlands Beekmantown,
inpart because excellent outcrops and large hydro-carbon reservoirs
are present there. Thus, theseequivalent rocks provide a wealth of
informationwith which to interpret the Beekmantown. Amongthe
equivalent rocks are the Beekmantown typesection of New York state
(Fisher, 1968), theArbuckle Group of Oklahoma, the EllenburgerGroup
of Texas, the Knox Group of the Illinoisbasin, and the Prairie du
Chien Group of theMichigan basin. Features shared by all of these
rockunits include relatively great thickness (up to a fewthousand
meters); locally interbedded sandstones,particularly in the lower
parts of the section; depo-sition mainly in peritidal environments,
whichcommonly contain distinct shallowing-upwarddepositional units
up to a few meters thick; exten-sive dolomitization, much of which
was penecon-temporaneous with deposition; and porosity that
isdominantly secondary related to karst processesand/or
fracturing.
Porosity development in these Lower Ordoviciandolomites has been
the focus of numerous studies.Amthor and Friedman (1991) described
dolomitetextures and porosity development in theEllenburger Group
of west Texas, noting that up to12% porosity was present at a depth
of 6477 m(21,250 ft) in the Delaware basin. They also dis-cussed
the relative importance of karst processesand tectonically induced
fracturing in these deeplyburied carbonate reservoirs. The
importance ofkarst porosity in Lower Ordovician dolomites has
been emphasized by Loucks and Anderson (1985)and Kerans (1988),
among others. Interestingly, nodistinct evidence of karst processes
was observedin the Beekmantown samples examined for thisstudy.
A standard tool for defining depositional trendsis the isopach
map. Unfortunately, the alloch-thonous Beekmantown sections in the
Appalachianthrust sheets are incomplete and have been trans-ported
too far to offer much insight into the origi-nal thickness of the
Beekmantown, but theautochthonous Beekmantown is another story.
Bydefinition, the autochthonous Beekmantown isessentially in situ,
and seismic lines show that it isdipping very gently to the east
and southeast. Wellspenetrating the autochthonous Beekmantown
indi-cate that the section (and each of the two forma-tions
comprising the Beekmantown) thickens veryregularly from north to
south (Figure 5). This trendsuggests that the autochthonous
Beekmantownwas deposited on a very gently southward-dippingramp.
Minor irregularities in thickness probablyindicate subtle
paleohighs and paleolows on theramp. The fact that the Theresa and
Beauharnoisformations vary directly in thickness in the studyarea
indicates that these paleodepositional featurespersisted throughout
Beekmantown deposition.
From this data on thickness and assuming littleor no
post-Beekmantown erosion (an assumptionsupported by the blanketlike
nature of the shalydolomite capping the Beauharnois Formation),
onecan calculate the angle of dip of the “ramp” onwhich the
Beekmantown was deposited. The for-mation thickens consistently in
a southward direc-tion (Figure 6) by 200 m (656 ft) over a lateral
dis-tance of 140 km (86 mi). Converted to degrees, thismeans that
the dip on the Beekmantown “ramp”
Dykstra and Longman 521
Figure 5—Cross-plot of the thick-ness of Theresa and
Beauharnoisformations in the BeekmantownGroup relative to distance
from aninferred paleoshoreline. Both for-mations steadily increase
in thick-ness from north to south acrossthe study area.
-
was about 0.083°. Although the dip at any giventime during
deposition may have been somewhatgreater than this extremely gentle
slope, it is clearthat the Beekmantown units were deposited on
anearly flat surface.
DEPOSITIONAL UNITS IN THEBEEKMANTOWN
Based on sparse outcrop data in the QuebecLowlands, the
Beekmantown Group has traditional-ly been subdivided into the
sandstone-rich TheresaFormation and the overlying, dominantly
dolomiticBeauharnois Formation (as described by Globensky,1987).
Based on subsurface information, includinglithology logs and
cuttings, the Beekmantownsection can be subdivided into seven
deposition-al units, with three in the Theresa Formation andfour in
the Beauharnois Formation. These unitsare herein numbered in
ascending stratigraphicsequence. The distribution of these units is
sum-marized on the cross section in Figure 6, andeach unit is
described brief ly in the followingparagraphs.
Theresa Formation (Unit 1)
Above the sandstones of the Potsdam Group, thefirst significant
occurrence of dolomite marks thebase of the Theresa Formation. The
lowest unit inthe Beekmantown is dominated by dolomite-
andquartz-cemented quartz sandstones across most of
the study area. In general, these sandstones are tootight to
offer any reservoir potential. In the south-ern part of the area,
however, between the St.Simon 1A and St. Armand 1 wells, unit 1
thickensabruptly and the amount of dolomite present grad-ually
increases upward over a few tens of meters toa point where it
becomes the dominant lithology.Thick oolitic and peloidal dolomites
in this intervalare nonporous in the St. Armand well, but
couldoffer some reservoir potential if located withinstructural
closure.
Theresa Formation (Unit 2)
An interval of relatively pure dolomite called unit2 occurs in
the southern part of the study area.This unit consists of very
finely crystalline andunfossiliferous mudstones that were
probablydeposited in peritidal to very shallow subtidal
envi-ronments. More northerly wells contain little ornone of this
dolomite, and sandstone dominatesdue to the more updip (proximal)
position. Thedisappearance of the peritidal dolomites updip
sug-gests that the Theresa/Potsdam contact is uncon-formable
farther updip to the north (Figure 6).
Unit 2 is a very important interval because itforms a promising
reservoir objective. Inter-crystalline and vuggy porosity occur in
cores takenfrom this sequence in both the St. Armand 1 and St.Simon
1A wells. The porous dolomite in St.Armand 1 was not tested because
no hydrocarbonshows were observed, but in the St. Simon 1A wellthe
interval was heavily stained with bitumen and
522 Beekmantown Group Gas Potential
��������������������������������������
������ �
SOUTHA #166
ST. ARMAND-OUEST1
#157BROSSARD
1
#214AST. SIMON
1A
#162ST. OURS
1
#156GENTILLY
1
#187DU CHENE
1
#161STE. FRANCOISE
1
#167STE. CROIX
1
#175LES SAULES
1
NORTHA'
Bea
uhar
nois
Fm
.T
here
sa F
m.
ST PT
ST PT
ST PT
TD = 11,335 ft.
TD = 8570 ft.TD = 6156 ft.TD = 13,920 ft.
TD = 4754 ft.
ST PT
ST PT
ST PT
TD = 12,483 ft.
Potsdam
Unit 1
Unit 2
Unit 3
Unit 4
Unit 5
Unit 7
Unit 6
BE
EK
MA
NT
OW
N G
RO
UP
7
2
3
4
6
Black RiverChazy
48 km 58 km 30 km 22 km66 km 8 km35 km 36 km23 km
ST PT
TD = 10,413 ft.
ST PTTD = 9385 ft.
Canadian Shield(Grenville)
ST PT
TD = 6137 ft.
ST PT
TD = 3186 ft.
7
3
4
6
7
3
4
6 3
SAUK UNCONFORMITY
#222ST. WENCESLAS
1
Potsdam
Potsdam
Beekmantown Facies Vertical Scale
500 m
ST PT
Subtidal Peritidal
Figure 6—Regional cross section across the Quebec Lowlands area
showing the geometry of the various units in theBeekmantown Group.
Location of this section is shown in Figure 1.
-
did yield gas (mainly carbon dioxide, as describedin detail by
Dykstra, 1993). This gas-rich interval,from 4107 to 4112 m (13,474
to 13,491 ft), con-tains up to 17% porosity.
Theresa Formation (Unit 3)
The upper part of the Theresa Formation, herecalled unit 3, is
composed primarily of sandstoneon the updip part of the ramp. This
unit thickensgradually downdip and changes into
interbeddedsiltstones and limestones. Cuttings samples suggesta
fair degree of homogeneity in the lithologies in agiven well,
indicating that conditions on the rampremained fairly constant
during deposition. On theupdip part of the ramp where units 1 and 2
areabsent, unit 3 rests unconformably on the Potsdam.Updip erosion
of the Potsdam probably providedsome of the sand in this interval.
Almost every-where the rocks of unit 3 are quite tight, and eventhe
cleanest sandstones are so extensively cement-ed with dolomite and
quartz that they offer essen-tially no reservoir potential.
Lower Beauharnois (Unit 4)
A gradual decrease in the rate of clastic depositionat the end
of unit 3 deposition, which was probablycoincident with a continued
marine transgression,
resulted in the onset of carbonate deposition. Theserelatively
clean carbonates mark the base of theBeauharnois Formation and the
deposition of unit 4conformably over the Theresa. A regional
limestonewith scattered skeletal fragments marks the base ofthe
unit and is absent only in wells that were drilledon paleohighs
(e.g., Ste. Francoise Romaine 1 well).
Unit 4 consists of subtidal limestones that gradeup into
peritidal dolomites (Figure 7). In wells thatwere drilled into the
more distal portion of theramp (e.g., St. Simon 1A and St. Armand
1), thisunit consists mainly of limestone and shaly lime-stone with
little or no reservoir potential. In updipwells, however, unit 4 is
a promising reservoirobjective. Not only does it contain the
thickestdolomite interval in the Beauharnois, but it also haslocal
porosity, and has yielded gas shows and bitu-men stain in the
autochthon. This unit may alsoform the reservoir in the
allochthonous thrustsheet that produces at St. Flavien field.
Unit 5
Unit 5 is a relatively thin wedge of rock that ispresent only in
the deeper autochthonous wells.Downdip wells, such as St. Armand 1
and St. Simon1A, exhibit a dominant limestone and shaly lime-stone
lithology, and the unit appears to lie con-formably on unit 4.
Farther updip, the unit is domi-nated by finely crystalline
(peritidal?) dolomite.
Dykstra and Longman 523
����������������
South
Below Fair Weather Wave Base
Deep RampShallow Basin
Shallow Subtidal toSupratidal Flats
Shallow Ramp
Increasing Faunal Content & Decreasing Clastics
Passive Margin-Basin Sag
St. Armand 1 St. Simon 1A
St. Ours 1St. Wenceslas 1
Gentilly 1
Du Chene 1Ste. Francoise Romaine 1
Ste. Croix 1Les Saules 1
Precambrian Grenville Basement
Wave AgitatedClean Dolomites Dolomites & Shaly Dolomites
Dolomites & Sands
Low Energy Peritidal
North
+ ++ +
+
+
++
+ + +
+
++
++
+ +
+
+
+ + +
+
++
+ + +
+
+
+
+ + +
+
++
+
+
+
+
+
+
+
+ + +
+
++
+ ++
+
++
+
+
+
+ ++
+
+
+
+
++
+
+
+
+ ++
+
+
+
+
+ +++ ++
++
++
+
+
+ + +
+
+
+
+
+ ++
+
+
+
+
++
++ +
+
+ +
+
+
+
+
+ ++
+ + ++ + +
++
+ + ++ + +
++
+
++
+
+ ++ +++ + +
++ + ++ ++
+ ++ + + ++ + ++ + +
++ + ++ ++
+
++
+
+ +
+
++
+
+ ++ +
+
+
+
+ + +
+
+
+
++
+
+
+
+ + +
+ + ++
+
+
+
+
+ +
+ +
+
+
+ + +
+
+
+
+
++
++
+ +
+
+ + +
+ + ++
+ + ++ + +
++
+
+ + ++ + +
+
+
+ ++ +++ + +
++ + ++ ++
+
+ ++ +++ + +
++++
+
+
+
+ +
UNIT 3
UNIT 4
UNITS 1 & 2
POTSDAM
Figure 7—Schematic depositional profile of the carbonate ramp
during unit 4 deposition.
-
The interval grades upward from limestone todolomite and
represents a shallowing-upwarddepositional unit. More open-marine
areasdowndip were subject to less dolomitization andless erosion.
The reservoir potential of unit 5 is lim-ited by the fact that most
of the updip shallow-water dolomites have been eroded. There may
beareas where porous peritidal dolomites in this unitwere deposited
and escaped erosion, but they havenot yet been found.
Unit 6
Unit 6 is very similar in lithology to the middlepart of the
Beauharnois Formation. This unit repre-sents another
shallowing-upward depositionalsequence and unconformably overlies
unit 5 (far-ther updip where unit 5 is absent it overlies unit
4).Like the other units, the thickness of unit 6 increas-es in a
basinward (southeasterly) direction.Lithologically, unit 6 is very
similar to unit 4 becauseit was subjected to the same diagenetic
processes.Limestones are more common in the lower part ofthe unit,
particularly in the downdip wells, andgrade updip into dolomites.
Even midway up theramp in the St. Ours 1 well, samples reveal that
thedominant lithology is peritidal dolomite that oncecontained
well-developed intercrystalline porosity.In that well, most of this
porosity is now filled withlate-stage poikilotopic calcite, but
these peritidaldolomites could form a good reservoir facies
whereearly entrapment of hydrocarbons prevented thelate-stage
cementation by calcite.
Unit 7
The uppermost part of the Beauharnois Forma-tion consists of a
shaly dolomite mudstone interval
that is remarkably uniform in thickness andlithology across the
study area (Figure 6). Thisinterval is absent only in the
updip-most wellswhere erosion has removed the top of theBeekmantown
section. This unit was depositedwhen a major marine transgression f
looded theramp and just before the major regression thatcoincided
with formation of the widespreadSauk unconformity. The blanketlike
nature ofunit 7 is evidence that there was not a largeamount of
differential erosion on the ramp fol-lowing Beekmantown deposition.
We postulatethat unit 7 was relatively impermeable
duringpost-Beekmantown exposure. This, in combina-tion with a
moderately arid climate, limited largeamounts of meteoric water
from penetrating theunderlying dolomitic units. Thus, the
presenceof this shaly dolomite partly explains why evi-dence of
karst processes is so rare in the underly-ing Beekmantown
section.
PORE TYPES IN THE BEEKMANTOWN
Five types of porosity occur in the Beekman-town Group in the
Quebec Lowlands. Listedapproximately in order of decreasing
abundance,these are (1) vuggy (and minor moldic); (2)
inter-crystalline; (3) fracture; (4) interparticle in carbon-ate
grainstones; and (5) interparticle betweenquartz grains in
sandstone. Except for the interpar-ticle pores in sandstones,
examples of these poretypes are shown in Figure 8. Similar former
poreswere commonly observed in other Beekmantownsamples to be
filled with late cements, such as cal-cite, quartz, and
dolomite.
Vuggy pores in the Beekmantown are irregular inshape and
distribution. They formed by dissolution
524 Beekmantown Group Gas Potential
Figure 8—Examples of pore types in the Beekmantown Group. (A)
This thin-section photomicrograph from unit 2in the St. Simon 1A
well shows a finely crystalline dolomite stained with pyrobitumen.
Scattered vugs are impreg-nated with blue epoxy. Some vugs probably
represent former fossil fragments and are now enlarged moldic
pores.Other vugs are very irregular in shape and probably formed
during dolomitization and dissolution of the matrix.Total porosity
measured from core at this depth is about 17%. (B) The small vug
visible just above the photo labelprobably formed by dissolution of
a fossil fragment, such as an echinoderm columnal. Open
intercrystalline porescontain a residue of black pyrobitumen.
Sample from unit 2 in the lower part of the Theresa Formation. (C)
Thiscutting from unit 6 shows a small vug (at upper right) that has
been partly filled with quartz cement (which appearslighter beige
than surrounding ferroan dolomite crystals). The crystals lining
the vug are clearly zoned with iron-rich rims (stained blue). The
fact that bitumen lines the vug beneath the quartz crystal
indicates that the quartzpostdated hydrocarbon migration. (D) This
cutting from unit 6 contains mainly intercrystalline porosity
betweenthe dolomite rhombs. Some quartz sand grains (light beige to
the far right and far left) with well-developed quartzovergrowths
are present in the dolomite. The residue of black bitumen on the
dolomite rhombs indicates that thispore network once contained
hydrocarbons. (E) This photomicrograph shows partly cemented
fractures. The hostrock is a sandy dolomite, and the fractures
contain both quartz and calcite cements. (F) This dolomitized ooid
grain-stone is from the Beekmantown in one of the thrust sheets
associated with St. Flavien field. Interparticle pores arepresent
locally, but are partly filled with dolomite cement. Such
interparticle porosity was not observed in theautochthonous
Beekmantown samples examined for this study, but its presence in
the thrust sheets indicates itcould be important locally in forming
reservoirs.
-
and some are simply enlarged moldic pores.Examples are shown in
Figure 8A–C. Total vuggyporosity in the samples studied nowhere
exceeded5%. Montañez and Stefani (1993) reported that simi-lar
vuggy pores are locally present in the LowerOrdovician Knox Group
of the U.S. Appalachians in
peritidal (cyclic) deposits, but that total vuggy poros-ity
there typically averages only 3%.
Intercrystalline porosity occurs between crystalsof relatively
similar size (e.g., Figure 8D) and iscommon in some peritidal
dolomites. Some inter-crystalline porosity can also be seen in
Figure 8A
Dykstra and Longman 525
-
and B near the vuggy pores. Much of the intercrys-talline
porosity observed in the Beekmantown sam-ples is lined with a black
hydrocarbon residue(bitumen or pyrobitumen). Other former
intercrys-talline pores are filled with calcite cement. Thistype of
porosity was originally quite common inthe peritidal dolomites at
the top of the deposition-al units, but is now present only
locally, particularlywhere bitumen staining is common (e.g., in
unit 2of the St. Simon 1A well).
Fracture porosity is difficult to recognize in cut-tings samples
because the cuttings break along thesurfaces of open fractures.
However, as shown inFigure 8E, examples of fractures were observed
insome core samples. Open fractures do not appearto be particularly
important in the autochthon, butare probably important in the
Beekmantown inthrust sheets. Bertrand and Savard (1992,
personalcommunication) suggested that fracture porosityplays an
important role in the reservoir at St.Flavien field. Most of the
fractures observed in thisstudy were cemented with some combination
ofcalcite, quartz, and dolomite cements.
Interparticle porosity in carbonate grainstoneswas not observed
in samples from the autochthon,but does occur in dolomitized ooid
grainstones inthrust sheets near St. Flavien field (Figure
8F).Considerable dolomite cementation between theooids filled much
of the former interparticle porosi-ty, but as much as 12%
interparticle porosityremains in a few beds. Such a network of
interparti-cle pores can provide excellent reservoir quality, soits
presence in the Beekmantown could be impor-tant. However, no porous
grainstone samples wereobserved in the autochthon.
The least common type of porosity observed inthe Beekmantown is
that occurring between quartzgrains in the sandstones. This was
observed in onlya few samples of the Theresa Formation. In the
fewsandstones with interparticle pores, the porositywas preserved
by the presence of authigenic clays(mainly illite) which inhibited
cementation byquartz and dolomite. These same clays serve
toseverely limit the reservoir potential of this type ofporosity.
The highest interparticle porosityobserved in Beekmantown
sandstones is only about3%, and the pores are too isolated to be
productive.Several factors contribute to the paucity of
interpar-ticle porosity in the Beekmantown sandstones.
Mostimportant among these are the early cementation ofthe sands
with dolomite, and the extensive forma-tion of quartz overgrowths
on sand grains in thosesandstones with little or no dolomite.
In summary, the best potential reservoir rockslikely to occur in
the Beekmantown are dolomiteswith a combination of vuggy,
intercrystalline, andfracture porosity. These porous dolomites are
mostlikely to occur near the tops of the shallowing-upward
depositional units (units 2, 4, and 6) where peritidaldolomites
accumulated. A factor that decreases thereservoir potential of the
peritidal dolomites is thepresence of siliciclastic sediments
(clays and sand)which seem to have aided porosity destruction. It
isalso worth noting that early entrapment of hydrocar-bons probably
helped preserve porosity locally inthese peritidal dolomites. This
is indicated by thecommon occurrence of hydrocarbon residues
inintercrystalline pores, and by the fact that off-struc-ture,
formerly porous dolomites now tend to be tight-ly cemented, mainly
with late-stage calcite cement.
DIAGENETIC SEQUENCE
Because of their age (more than 450 m.y.), previ-ous deep
burial, and complex tectonic setting, therocks of the Beekmantown
Group have undergoneextensive diagenesis. Important diagenetic
eventsinclude dolomitization, neomorphism (of bothdolomite and
limestone), precipitation of quartzovergrowths and ferroan
dolomite, and precipita-tion of late-stage fracture-filling quartz
and calcite.The relative timing of these events can be deter-mined
from crosscutting relationships and othercharacteristics. This
allows development of a para-genetic sequence (Figure 9) that
provides insightinto the evolution of the pore system in the
reser-voir rocks. The major diagenetic events affectingthe
reservoir rocks are discussed briefly.
The earliest diagenesis in the Beekmantownoccurred
penecontemporaneously with deposi-tion. Carbonate muds (micrites)
deposited in periti-dal settings were probably dolomitized shortly
afterdeposition. This early dolomitization produced
theaphanocrystalline dolomites seen in the upperparts of the
Beekmantown depositional units. Atthe same time, other processes,
including down-ward seepage of hypersaline brines and/or mixingof
fresh and marine waters, contributed to dolomi-tization of
underlying calcareous sediments. In theupdip parts of the
Beekmantown units, almost allof the carbonate sediments were
dolomitized.Farther downdip, however, the peritidal environ-ment
was more ephemeral (if it formed at all), anddolomitization was
much less extensive. Thus, thedowndip parts of the Beekmantown
units are com-monly less than 50% dolomite. A useful summarypaper
on the nature and origin of dolomite inLower Ordovician peritidal
cyclic deposits (in theKnox Group of the U.S. Appalachians) was
pub-lished by Montañez and Read (1992). Many of theirfindings can
probably be applied directly to dolomi-tization in the Beekmantown
units.
Much of the dolomitization of the peritidal sedi-ments produced
relatively low-porosity dolomitemudstones, but formation of
euhedral crystals in
526 Beekmantown Group Gas Potential
-
certain beds produced highly porous dolomites inother intervals.
Thus, some of the porosity thatformed in the Beekmantown during
early peritidaldolomitization has survived more or less intact
duringthe hundreds of millions of years since the Early
Ordo-vician. Similarly, most of those dolomites that weretightly
cemented early remain tight dolomites today.Only fracturing has
significantly improved reservoirquality in the low-porosity
dolomites (and significantoccurrences of fractures are very
localized).
During and after this early penecontemporane-ous dolomitization
event, some skeletal fragmentsremained calcareous, in part because
their largergrain size inhibited dolomitization. Some of
theseskeletal fragments were subsequently dissolved toform moldic
pores. This dissolution most likelyhappened while the sediments
were still at fairlyshallow burial depths.
An interesting anomaly in the Beekmantown ofthe Quebec Lowlands
is that there is almost noevidence of subaerial exposure and karst
processes.
This absence is surprising considering that mostof the
depositional units are capped by peritidaldeposits, and the fact
that the Beauharnois is sepa-rated from the overlying Chazy and
Black Rivergroups by a major subaerial unconformity. De-spite what
must have been extensive subaerialexposure, karst features, such as
solution cavernsor cave calcite, were not observed in the
Beek-mantown [although B. Bailey (1992, personalcommunication)
reported that some of the porosi-ty in the St. Flavien gas field
may be karst related].Soil zones and pedogenic calcretes are
alsoabsent, although this is partly due to the absenceof land
plants during the Early Ordovician. Thepresence of regional
truncation of some units(and the formation as a whole) updip to the
north(Figure 7) proves that there was such subaerialexposure, but
perhaps exposure occurred underrelatively arid conditions where
limited amountsof fresh water limited the development of
karstfeatures.
Dykstra and Longman 527
Figure 9—Paragenetic sequence for the autochthonous Beekmantown
Group in the Quebec Lowlands. Fracturing,which was also an
important porosity-enhancing process in the Beekmantown, is not
shown, but would haveoccurred at a number of times across the time
spectrum indicated.
-
After the Beekmantown was buried beneathyounger Ordovician
sediments, a variety of neomor-phic events occurred in the
limestones anddolomites. Small micron-size crystals
commonlyrecrystallized to somewhat larger crystals, some
pre-cipitation of calcite and dolomite occurred, and thesediments
became completely lithified. Much of theporosity associated with
deposition was probablydestroyed through compaction during this
burial.
Still later, possibly after several thousand feet ofburial,
quartz overgrowths began forming in thesandstones. The source of
the silica for the forma-tion of these ubiquitous silica cements is
unknown,but may be related to compaction in the
sandstonesthemselves or to silica carried upward from moredeeply
buried sandstones in the underlyingPotsdam Group. In any case, the
net result is thatalmost all sandstones in the Beekmantown
becametightly cemented with the quartz overgrowths,destroying
reservoir quality prior to the generationof hydrocarbons.
Another event of some significance during thistime of burial was
the formation of stylolites in thelimestones (and, less commonly,
in somedolomites). High-amplitude stylolites indicate thatlarge
amounts of limestone were dissolved in thelower parts of the major
Beekmantown units. Someof the vast quantities of calcium carbonate
releasedduring this dissolution was transported in the for-mation
waters up and out of the formation, butsome of the remaining
calcium carbonate probablyprecipitated as calcite or dolomite to
plug much ofthe porosity remaining in the formation. The threemost
common cements formed during this burialdiagenesis (but prior to
hydrocarbon generationand entrapment) were equant calcite,
ferroandolomite, and saddle dolomite.
Toward the end of the Ordovician, the Taconicorogeny resulted in
burial of the Beekmantown andassociated formations to depths
sufficient for theonset of hydrocarbon generation. Some of
thehydrocarbons probably migrated along the sty-lolitic pathways
and fractures in the Beekmantownto fill whatever pores remained
open in the paleo-structures existing at the time. Small amounts of
secondary porosity may have been created in CO2-enriched waters
migrating just ahead of (orwith) the hydrocarbons, but this
late-stage porosityis relatively unimportant. Once trapped,
furthercooking of the hydrocarbons produced bitumenresidues that
coated the walls of many of the poreand fracture surfaces in the
rocks where the oil wastrapped. Gas was also generated during this
ther-mal breakdown of the hydrocarbons. The bestpotential for
Beekmantown reservoirs occurs inporous peritidal dolomites capping
the deposition-al units on paleostructures where the gas wastrapped
as soon as it formed.
Much Beekmantown porosity in the form of frac-tures, vugs, and
intercrystalline pores containing aresidue of bitumen somehow
survived deep burial,but was later cemented. Such late-stage
cementscan be clearly identified because they postdate thebitumen
residue (e.g., Figure 8C). In such cases,three stages of late
cementation can be identified.The first stage consisted of
precipitation of a fewdolomite crystals, the second was
precipitation ofquartz in the form of euhedral crystals, and
thethird was the precipitation of coarse equant calcitecement. It
is these late-stage (post-hydrocarbon)cements that have now so
completely destroyedthe reservoir-quality porosity of most of the
former-ly porous dolomites in the Beekmantown. Onlywhere
“permanent” gas entrapment predates theselate-stage cements is
reservoir quality preserved inthe Beekmantown.
This paragenetic sequence makes it clear why itis so important
that exploration for Beekmantownreservoirs focuses on finding
Ordovician paleo-structures capable of trapping hydrocarbons.
Onlyin these settings was the extensive late-stagecementation
inhibited sufficiently for commercialamounts of hydrocarbons to be
produced. Thehighly porous (and bitumen-r ich) peritidaldolomites
at a depth of more than 4000 m (13,000ft) in unit 2 of the lower
Beekmantown in the St.Simon 1A well clearly show that burial depth
alonewas not a cause of porosity destruction. Instead, upto 17%
porosity was preserved at depths that wereonce as great as perhaps
9000 m (30,000 ft) (basedon burial history reconstructions by
Bertrand andDykstra, 1993), because the trapped hydrocarbonsand
associated CO2 gas prevented precipitation oflate-stage
cements.
CONCLUSIONS
(1) Based on sparse outcrop data, the Beek-mantown Group has
traditionally been subdividedinto the sandstone-rich Theresa
Formation and theoverlying dolomite-rich Beauharnois
Formation.Based on subsurface information, including theanalysis of
cuttings and wireline logs, theBeekmantown in the Quebec Lowlands
can be sub-divided into seven depositional units, hereinlabeled 1
to 7 in ascending stratigraphic order.
(2) Units 2 and 4 (and to a lesser extent unit 6)contain the
best potential reservoir rocks in theBeekmantown Group. In the
autochthonousBeekmantown section, the most favorable rocksare found
in areas along the depositional rampwhere peritidal dolomites with
early porosityescaped later porosity-destructive processes suchas
late-stage calcite cementation due to the earlyentrapment of
hydrocarbons.
528 Beekmantown Group Gas Potential
-
(3) Also offering some reservoir potential in theallochthonous
Beekmantown section are fracturedand brecciated dolomites, such as
those seen in thethrust-faulted anticlinal trap at the St. Flavien
field.
(4) During development of the Sauk unconformi-ty, the slightly
shaly unit at the top of theBeauharnois Formation (unit 7), in
combinationwith a moderate to arid climate, limited largeamounts of
meteoric water from penetrating intothe underlying dolomitic units.
This partly explainswhy karst processes did not play a major role
inporosity development in the Beekmantown.
(5) Although limestones within the Beek-mantown are present in
almost all study wells, par-ticularly downdip on the depositional
ramp, nonewere observed to have significant porosity.Stylolites are
common in the limestones and appar-ently released large volumes of
calcium carbonate,some of which reprecipitated as cements in
porousdolomites after significant burial.
(6) As a general rule, most of the dolomites inthe Beekmantown
are nonporous, but much of thelack of porosity is due to late-stage
cementation bypoikilotopic calcite. This calcite clearly
postdateshydrocarbon generation and migration because itcommonly
fills vugs and intercrystalline pores linedwith bitumen.
(7) The common occurrence of bitumen andpyrobitumen in the study
wells suggests that oilwas once widespread in the
Beekmantowndolomites. Ordovician paleostructures, where
earlyhydrocarbon entrapment limited later burialcementation, offer
the best reservoir potential.These prospects can be defined on time
structuremaps generated from seismic data (e.g., on thehorizon at
the base Theresa).
(8) The findings from this study should be appli-cable beyond
the area of the Quebec Lowlands toother deep Beekmantown prospects
in theautochthon elsewhere in the Canadian and
U.S.Appalachians.
REFERENCES CITEDAmi, H. M., 1900, On the geology of the
principal cities of eastern
Canada: Royal Society of Canada, Transaction Series 2, v. 6,sec.
4, p. 125–173.
Amthor, J. E., and G. M. Friedman, 1991, Dolomite-rock
texturesand secondary porosity development in Ellenburger Group
car-bonates (Lower Ordovician), west Texas and southeasternNew
Mexico: Sedimentology, v. 38, no. 2, p. 343–362.
Bebout, D. G., W. A. White, and T. F. Hentz, 1993, Atlas of
majorMidcontinent gas reservoirs: Gas Research Institute and
Bureauof Economic Geology, University of Texas at Austin,
SpecialPublication, 85 p.
Bertrand, R., and J. C. F. Dykstra, 1993, Organic metamorphism
andburial histories in the St. Lawrence Lowlands and in the
externaldomain of the Quebec Appalachians (abs.): Geological
Societyof America, Abstracts with Programs, v. 25, no. 6, p.
A-31.
Bertrand, R., and M. Savard, 1992, Isotopic study of St. Simon
no. 1
core dolostone, Beekmantown Group, St. Lawrence
Lowlands:Proprietary Study prepared for Bow Valley Industries, 8
p.
Clarke, J. M., and C. Schuchert, 1899, Nomenclature of NewYork
series of geological formations: Science, New Series 110,876 p.
Dykstra, J. C. F., 1993, Recent gas exploration in the St.
LawrenceLowlands, Quebec: BVI et al. St. Simon no. 1: Ontario
Pe-troleum Institute, Proceedings of the 31st Annual Conference,15
p.
Ells, R. W., 1896, Paleozoic outliers in the Ottawa basin:
RoyalSociety of Canada, Proceedings and Transactions 12, section
4,p. 21–30.
Farmer, R. E., 1965, Genesis of subsurface carbon dioxide, in
A.Young and J. E. Galley, eds., Fluids in subsurface
environments:AAPG Memoir 4, p. 378–385.
Globensky, Y., 1987, Géologie des Basses-Terres du Saint
Laurent:Gouvernment du Québec Publication MM 85-02, 63 p.
Holtz, M. H., and C. Kerans, 1992, Characterization and
classifica-tion of west Texas Ellenburger reservoirs, in M. P.
Candelariaand C. L. Reed, eds., Paleokarst, karst-related
diagenesis andreservoir development: examples from
Ordovician–Devonianage strata of west Texas and the mid-continent:
1992 AnnualField Trip, Permian Basin Section, SEPM, Publication No.
92-33,p. 45–54.
Houseknecht, D. W., 1986, Evolution from passive margin to
fore-land basin: the Atoka Formation of the Arkoma basin,
south-central U.S.A.: International Association of
Sedimentologists,Special Publication No. 8, p. 327–345.
Huff, W. D., S. M. Bergstrom, and D. R. Kolata, 1992,
GiganticOrdovician volcanic ash fall in North America and Europe:
bio-logical, tectonomagmatic, and event-stratigraphic
significance:Geology, v. 20, no. 10, p. 875–878.
Hume, G. S., 1932, Oil and gas in eastern Canada:
CanadaDepartment of Mines, Economic Geology Series No. 9, 187
p.
Jacobi, R. D., 1981, Peripheral bulge—a causal mechanism for
theLower/Middle Ordovician unconformity along the westernmargin of
the northern Appalachians: Earth and PlanetaryScience Letters, v.
56, p. 245–251.
Kerans, C., 1988, Karst-controlled reservoir heterogeneity
inEllenburger Group carbonates of west Texas: AAPG Bulletin,v. 72,
no. 10, p. 1160–1183.
Knight, I., N. P. James, and T. E. Lane, 1991, The Ordovician
St.George unconformity, northern Appalachians: the relationshipof
plate convergence at the St. Lawrence promontory to
theSauk/Tippecanoe sequence boundary: Geological Society ofAmerica
Bulletin, v. 103, no. 9, p. 1200–1225.
Lamothe, M., 1989, A new framework for the Pleistocene
stratigra-phy of the central St. Lawrence Lowland, Southern
Quebec:Géographie Physique et Quaternaire, v. 43, no. 2, p.
119–129.
Lindsay, R. F., and K. M. Koskelin, 1993, Arbuckle Group
(LateCambrian–Early Ordovician) shallowing-upward parase-quences
and sequences, southern Oklahoma, in D. R. Kellerand C. L. Reed,
eds., Paleokarst, karst-related diagenesis, reser-voir development
and exploration concepts: examples fromthe Paleozoic section of the
southern mid-continent: 1993Annual Fieldtrip Guidebook, Permian
Basin Section, SEPM,Publication No. 93-34, p. 45–65.
Logan, W. E., 1846, Report of progress 1844: Geological Survey
ofCanada, p. 5–66.
Logan, W. E., 1864, Rapport sur la Geologie du Canada:
Com-mission Geologique du Canada, Rapport de Progres Jusqu’a1863,
p. 983.
Loucks, R. G., and J. H. Anderson, 1985, Depositional facies
andporosity development in Lower Ordovician EllenburgerDolomite,
Puckett field, Pecos County, Texas, in P. O. Roehland P. W.
Choquette, eds., Carbonate petroleum reservoirs:New York,
Springer-Verlag, p. 1–19.
Montañez, I. P., and J. F. Read, 1992, Eustatic control on
earlydolomitization of cyclic peritidal carbonates: evidence from
theEarly Ordovician Upper Knox Group, Appalachians:
GeologicalSociety of America Bulletin, v. 104, no. 7, p.
872–886.
Dykstra and Longman 529
-
Montañez, I. P., and A. B. Stefani, 1993, An overview of
factorscontributing to the evolution and spatial distribution of
reser-voir properties in Knox cyclic carbonates, in D. R. Keller
andC. L. Reed, eds., Paleokarst, karst-related diagenesis,
reservoirdevelopment and exploration concepts: examples from
thePaleozoic section of the southern mid-continent: 1993
AnnualFieldtrip Guidebook, Permian Basin Section, SEPM,
PublicationNo. 93-34, p. 67–83.
Petzet, G. A., 1990, E & D drilling hits fast clip in Arkoma
basin: Oil& Gas Journal, October 1, v. 88, no. 40, p.
34–36.
Petzet, G. A., 1992, Quebec Cambro–Ordovician explorationfocuses
on Wilburton Arbuckle analog: Oil & Gas Journal,February 17, v.
90, no. 7, p. 82–83.
Raymond, P. E., 1913, Ordovician of Montreal and
Ottawa:Commission Geologique du Canada, 12e Congres
GeologiqueInternational, Livret Guide 3, p. 137–160.
Scotese, C. R., and W. S. McKerrow, 1991, Ordovician plate
tec-tonic reconstructions, in C. R. Barnes and S. H. Williams,
eds.,Advances in Ordovician geology: Geological Survey of
Canada,Paper 90-9, p. 271–282.
Shaw, B. R., 1993, Strike-slip interpretation of the
basin-boundingfaults of the St. Lawrence Lowlands basin in the
Quebec Cityarea: AAPG Bulletin, v. 77, no. 5, p. 743–760.
St. Julien, P., and C. Hubert, 1975, Evolution of the Taconian
oro-gen in the Quebec Appalachians: American Journal of Science,v.
275-A, p. 337–362.
St. Julien, P., A. Slivitzky, and T. Feininger, 1983, A deep
structuralprofile across the Appalachians of southern
Quebec:Geological Society of America Memoir 158, p. 103–111.
Stockmal, G. S., S. P. Colman-Sadd, C. E. Keen, S. J. O’Brian,
and G.Quinlan, 1987, Collision along an irregular margin: a
regionalplate tectonic interpretation of the Canadian
Appalachians:Canadian Journal of Earth Sciences, v. 24, p.
1098–1107.
Tremblay, A., 1992, Tectonic and accretionary history of
Taconianoceanic rocks of the Quebec Appalachians: American
Journalof Science, v. 292, p. 229–252.
Wilson, J. L., 1993, The Lower Ordovician of the
southwesternUnited States, in D. R. Keller and C. L. Reed, eds.,
Paleokarst,karst-related diagenesis, reservoir development and
explo-ration concepts: examples from the Paleozoic section ofthe
southern mid-continent: 1993 Annual Fieldtrip Guide-book, Permian
Basin Section, SEPM, Publication No. 93-34,p. 35–43.
530 Beekmantown Group Gas Potential
John C. F. Dykstra
John Dykstra graduated with aB.Sc. degree in geology from
theUniversity of Calgary in 1982. Sincethat time, he has worked as
anexploration geologist in frontierCanada and international areas
formost of his career. His interestsinclude geochemistry, basin
model-ing, and tectonic reconstruction.John is an active member of
AAPGand a professional member of theAssociation of Professional
Geologists, Geophysicistsand Engineers of Alberta (APEGGA).
Mark W. Longman
Mark W. Longman received hisB.A. degree from Albion
College(Michigan) in 1972 and his Ph.D.from the University of Texas
atAustin in 1976. He specializes inthe study of carbonate rocks
andreservoirs and enjoys working withcores, thin sections, wireline
logs,and seismic data to define andinterpret the nature and
produc-tion performance of such reser-voirs. For the past eighteen
years, first while workingwith Cities Service Company, Coastal Oil
and Gas, andButtercup Energy, and later as a consultant, he has
stud-ied a variety of carbonate reservoirs ranging in age
fromOrdovician to Miocene in areas ranging from theWilliston basin
to Texas, as well as southeast Asia,Mexico, the Middle East, and
the Mediterranean region.
ABOUT THE AUTHORS