- Email: [email protected]
The source and evolution of paleofluids responsible for secondary minerals in low-permeability Ordovician limestones of the Michigan Basin
Accepted Manuscript The source and evolution of paleofluids responsible for secondary minerals in lowpermeability Ordovician limestones of the Michigan Basin D.C. Petts, J.K. Saso, L.W. Diamond, L. Aschwanden, T.A. Al, M. Jensen PII:
S0883-2927(17)30282-2
DOI:
10.1016/j.apgeochem.2017.09.011
Reference:
AG 3952
To appear in:
Applied Geochemistry
Received Date: 7 July 2017 Revised Date:
14 September 2017
Accepted Date: 16 September 2017
Please cite this article as: Petts, D.C., Saso, J.K., Diamond, L.W., Aschwanden, L., Al, T.A., Jensen, M., The source and evolution of paleofluids responsible for secondary minerals in lowpermeability Ordovician limestones of the Michigan Basin, Applied Geochemistry (2017), doi: 10.1016/ j.apgeochem.2017.09.011. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Title: The source and evolution of paleofluids responsible for secondary minerals in low-
RI PT
permeability Ordovician limestones of the Michigan Basin
Authors:
a
M AN U
SC
Petts, D.C.a, *, Saso, J.K.b, Diamond, L.W.c, Aschwanden, L.c, Al, T.A.a and Jensen, M.d
Department of Earth and Environmental Sciences, University of Ottawa, Ottawa, ON, K1N
6N5, Canada b
Department of Earth Sciences, University of New Brunswick, Fredericton, New Brunswick,
c
TE D
Canada
Rock-Water Interaction Group, Institute of Geological Sciences, University of Bern,
Baltzerstrasse 3, 3012 Bern, Switzerland
Nuclear Waste Management Organization, Toronto, ON, M4T 2S3, Canada
EP
d
AC C
* Corresponding author: [email protected]
Version: September 19, 2017
1
ACCEPTED MANUSCRIPT
1 2
Abstract In this study we report on the source and evolution of fluids associated with secondary vein and replacement minerals in low-permeability carbonate units in the Michigan Basin. This
4
petrogenetic information was collected using thermometric data from fluid inclusions combined
5
with C-, O- and Sr-isotope data, and focuses on mm- or cm-wide vein and vug minerals from
6
Ordovician limestones of the Trenton and Black River groups and Cambrian sandstones in SW
7
Ontario, Canada.
SC
9
Primary fluid inclusions in dolomite represent fluid stage I and have the highest trapping temperatures (Ttrap) in the sedimentary succession, between 88 and 128 °C. Primary inclusions in
M AN U
8
RI PT
3
calcite (stage II), and celestine and anhydrite (stage III), represent the final stages of secondary
11
mineral formation and have Ttrap values of 54–78 °C. All three stages of vein minerals formed
12
from brines with salinities of 31 to 37 wt.% [CaCl2+NaCl]eq that were saturated in halite and
13
methane gas at the time of mineral growth. Three subsequent stages of secondary fluid
14
inclusions were observed in the samples (stages IV–VI), however, no secondary vein minerals
15
formed during these stages and they are interpreted as re-mobilization and/or minor fluid ingress
16
along grain boundaries and micro-fractures. Notably, stage IV secondary inclusions are gas-
17
undersaturated with salinities of 32 to 34 wt.% [CaCl2+NaCl]eq and minimum Ttrap values of 57–
18
106 °C, and are interpreted to have formed during a Late Devonian–Mississippian regional
19
heating event.
EP
AC C
20
TE D
10
Secondary minerals from the Cambrian units and the Shadow Lake Formation have δ13C
21
values of −6.1 to −2.5‰ (VPDB), δ18O values of +14.6 to +24.2‰(VSMOW) and 87Sr/86Sr of
22
0.70975 to 0.71043. Relative shifts of approximately +2 to +3‰ in δ13C, >+4‰ in δ18O and
23
−0.002 in 87Sr/86Sr are observed upward across the boundary between the Cambrian–Shadow
2
ACCEPTED MANUSCRIPT
Lake units and the overlying Gull River Formation. Samples from the Gull River and Coboconk
25
formations and the Trenton Group (Kirkfield, Sherman Fall and Cobourg formations) have δ13C
26
values of −1.0 to +1.9‰, δ18O values of +18.9 to +28.1‰ and 87Sr/86Sr values of 0.70790 to
27
0.70990.
RI PT
24
The combined microthermometric and isotopic data for secondary minerals in the
29
Cambrian–Shadow Lake units suggest they formed from hydrothermal brine with a geochemical
30
signature obtained by interaction with the underlying Precambrian shield, or shield-derived
31
minerals in the Cambrian sandstones. Previous U-Pb dating of vein calcite and Rb-Sr dating of
32
secondary K-feldspar from the region suggests that brine ingress occurred during the Late
33
Ordovician–Silurian. The O- and Sr-isotope variability in vein samples from the Gull River and
34
Coboconk formations is interpreted as localized mixing of 18O-enriched, connate fluids with
35
hydrothermal brine. In comparison, isotopic data for the Trenton Group indicate secondary
36
mineral formation from connate fluids, sourced from 18O-enriched, evolved seawater and/or
37
modified hydrothermal brine that experienced fluid–rock interaction during transit through the
38
underlying stratigraphy.
M AN U
TE D
EP
39
SC
28
Keywords: Michigan Basin; low-permeability rocks; secondary minerals; fluid–rock interaction;
41
fluid inclusions; strontium isotopes; stable isotopes
AC C
40
3
ACCEPTED MANUSCRIPT
42 43
1. Introduction Low-permeability sedimentary rocks have increasingly become a global focus for a wide range of environmental and geoscience studies, including nuclear waste management (Russell
45
and Gale, 1982; Hendry et al., 2015), CO2 sequestration (Shafeen et al., 2004; Benson and Cole,
46
2008) and extraction of gas from shale (Arthur and Cole, 2014). In the cases of CO2
47
sequestration and extraction of gas with hydraulic fracturing, the low-permeability rocks play a
48
critical role in limiting the upward migration of fluids that are injected or displaced. Low-
49
permeability sedimentary rocks have been of particular interest in countries such as Canada,
50
Switzerland, France, Belgium and Germany for the long-term storage of nuclear waste because
51
solute transport is dominated by diffusion and the clay-rich mineralogy provides large surface
52
areas for solute retention by sorption mechanisms. Internationally, detailed paleofluid studies
53
have provided valuable insight into fluid sources and solute transport mechanisms (Altmann et
54
al., 2012; Clark et al., 2013; Al et al., 2015; Mazurek and de Haller, 2017) that are essential for
55
evaluating the long-term viability of a site for nuclear waste storage.
SC
M AN U
TE D
56
RI PT
44
Low-permeability argillaceous limestones of the Trenton Group (10-16 to 10-12 m s-1; Beauheim et al., 2014), in the Bruce region of the Michigan Basin (southwestern Ontario,
58
Canada) have been identified as a potential host rock for the construction of a deep geologic
59
repository (DGR) for low- and intermediate-level nuclear waste. The Cobourg Formation, a
60
low-permeability, argillaceous limestone at the top of the Trenton Group occurs >650 m beneath
61
the Bruce nuclear site and has been proposed as the repository host rock. These limestones are
62
overlain by up to 200 m of regionally-extensive, low-permeability Ordovician shale (Intera,
63
2011; Beauheim et al., 2014). Detailed studies on the long-term physical and chemical stability
64
of the surrounding rocks have focused on a wide range of stratigraphic, structural, geophysical
AC C
EP
57
4
ACCEPTED MANUSCRIPT
and geochemical methods (Intera, 2011 and references therein). Previous investigations at the
66
Bruce site have contributed valuable information on the residence time and transport
67
mechanisms of fluids in these rocks, which include studies of porewater geochemistry and solute
68
transport (Clark et al., 2013; Al et al., 2015), hydraulic conductivity (Beauheim et al., 2014), and
69
mechanisms responsible for anomalous hydraulic-head profiles observed in the Trenton and
70
Black River group limestones (Normani and Sykes, 2012; Khader and Novakowski, 2014;
71
Neuzil and Provost, 2014; Neuzil, 2015).
SC
72
RI PT
65
Secondary minerals from veins in sedimentary rocks provide direct information on the movement of deep basin fluids. Detailed isotopic and petrographic information of secondary
74
minerals combined with fluid inclusion temperature constraints have been utilized in paleofluid
75
studies to differentiate between fluid sources that relate to either diagenesis/burial (Ayalon and
76
Longstaffe, 1988), or the influx of fluids from allochthonous sources (Spencer, 1980; Davies and
77
Smith, 2006). Detailed petrography (e.g., Taylor and Sibley, 1986) combined with thermometric
78
data from fluid inclusions (e.g., Roedder, 1979; Goldstein 2001) provide an understanding of the
79
relative timing of fluid infiltration and temperature constraints during crystallization. Here, we
80
report on the source and evolution of fluids associated with secondary mineral formation in low-
81
permeability Ordovician limestones and Cambrian siliciclastic units in the Michigan Basin from
82
southwestern Ontario, Canada.
TE D
EP
AC C
83
M AN U
73
Ordovician limestones of the Trenton and Black River groups, which are laterally
84
traceable across much of the Michigan Basin and occur in sections of the Appalachian Basin,
85
have long been a focus of paleofluid studies due to their association with hydrocarbon reservoirs
86
in the Albion-Scipio/Stoney Point fields in south-central Michigan (e.g., Hurley and Budros,
87
1990; Allan and Wiggins, 1993), the Lima-Indiana field in Ohio/Indiana (e.g., Wickstrom et al.,
5
ACCEPTED MANUSCRIPT
1992) and fields of southern Ontario (Powell et al., 1984; Middleton et al., 1993; Coniglio et al.,
89
1994; Obermajer et al., 1999). In these fields, hydrocarbon migration/reservoir formation is
90
closely tied to the formation of hydrothermal dolomite (Allan and Wiggins, 1993), which has
91
been suggested to form during high-temperature, fluid–rock interaction between the Ordovician
92
limestones and upwelling hydrothermal fluids (Davies and Smith, 2006). Furthermore, C-, O-
93
and Sr-isotope compositions of secondary minerals in southern Ontario (Harper et al., 1995;
94
Zeigler and Longstaffe, 2000; Haeri-Ardakani et al., 2013) and central Michigan (Allan and
95
Wiggins, 1993) have provided valuable information on the source of deep-seated fluid
96
migrations in these regions of the basin. In the Bruce region, however, detailed paleofluid
97
studies have largely focused on the geochemical and isotopic evolution of porewaters in the
98
Michigan Basin sedimentary succession (Clark et al., 2010a; 2010b; 2013; 2015; Al et al., 2015).
99
Knowledge gaps remain in the understanding of paleofluid contributions to the formation of
M AN U
SC
RI PT
88
secondary vein minerals in the Ordovician limestones, and in particular, the source and transport
101
mechanisms that are responsible for vein-related fluid migrations in these low-permeability
102
sedimentary rocks.
The objective of this study is to enhance the understanding of fluid residence time and
EP
103
TE D
100
migration that has been developed from previous research at the Bruce site. Detailed
105
petrography and microthermometric data from fluid inclusions are combined with C-, O- and Sr-
106
isotope data to determine the source and relative timing of paleofluid migrations that resulted in
107
the formation of secondary minerals, including vein and vug infill and replacement phases.
108
Samples were obtained from drill cores that section the entire succession of Cambrian to
109
Devonian sedimentary rocks and the underlying Precambrian gneissic basement.
AC C
104
110
6
ACCEPTED MANUSCRIPT
111
2. Geology and Stratigraphy of the Michigan Basin The Michigan Basin represents a broadly circular, ~500 km diameter intracratonic
113
sedimentary basin located in central North America. The study area is located along the eastern
114
margin of the basin, at the Bruce nuclear site (Fig. 1). A detailed summary of the Paleozoic
115
bedrock units in southern Ontario is discussed in Armstrong and Carter (2010) (and references
116
therein), a brief description of these units is provided below. At the Bruce site, the sedimentary
117
succession reaches a maximum thickness of ~ 860 m. The base of this succession comprises ~17
118
m of Cambrian sandstone which unconformably overlies Precambrian gneissic basement. Upper
119
Ordovician rocks disconformably overlie the Cambrian sandstones and are subdivided into the
120
Black River and Trenton group limestones and a thick sequence of shale. The Black River
121
Group consists of a 4–5 m thick basal unit of siltstone and minor dolostone (Shadow Lake
122
Formation) that is overlain by ~ 75 m of variably bioturbated and fossiliferous lithographic
123
limestone (Gull River and Coboconk formations). Overlying the Black River Group are ~ 110 m
124
of mostly argillaceous limestones of the Trenton Group (subdivided into the Kirkfield, Sherman
125
Fall and Cobourg formations). The Cobourg Formation is further subdivided into lower and
126
upper members, which, respectively comprise ~ 28 m of argillaceous limestone, with minor
127
interbeds of calcareous shale, and ~ 7.5 m of calcareous shale. The Cobourg Formation is
128
capped by ~ 200 m of Ordovician shale which is subdivided into the Blue Mountain, Georgian
129
Bay and Queenston formations. Overall, the Ordovician limestone–shale succession has been
130
interpreted as initial deposition in supra-tidal/tidal flat to lagoonal marine or offshore shallow
131
and deep-shelf settings (Black River and Trenton groups), and open marine to shallow intertidal
132
or supratidal settings for deposition of the Blue Mountain–Queenston shales. Overlying the
133
Ordovician shales are 415 m of Silurian and Devonian interlayered dolostone, shale and
AC C
EP
TE D
M AN U
SC
RI PT
112
7
ACCEPTED MANUSCRIPT
134
evaporitic units, which represent deposition in a carbonate-shelf setting with periods of restricted
135
marine circulation.
136
Herein we refer to “Ordovician limestones”, “Trenton Group limestones”, “Black River Group limestones” and “Cambrian sandstones” as generalized nomenclature for each of the
138
respective units, regardless of the minor occurrence of units/formations of differing lithology
139
(e.g., Cambrian dolostone units, Shadow Lake siltstones/dolostones in the Black River Group,
140
Collingwood shale member at the top of the Cobourg Formation).
SC
141
RI PT
137
Hydrogeologic conditions at the Bruce site have been investigated by Intera (2011), Normani and Sykes (2012), Beauheim et al. (2014), Khader and Novakowski (2014), Neuzil and
143
Provost (2014) and Neuzil (2015). The objective of these studies was to understand the
144
hydraulic properties of the sedimentary rocks through modelling (Normani and Sykes, 2012), in
145
situ hydraulic conductivity measurements (Beauheim et al., 2014) and interpretation of observed
146
anomalies in hydraulic-head distributions (Khader and Novakowski, 2014; Neuzil and Provost,
147
2014; Neuzil, 2015). Overall, the hydraulic data indicate that the permeability of the Upper
148
Ordovician limestones and shales is very low, such that solute transport is limited to diffusion.
149
In situ measurements of horizontal hydraulic conductivity (Kh) by Beauheim et al. (2014)
150
indicate that the Ordovician shales and the argillaceous limestones of the Trenton Group
151
represent an aquiclude (Fig. 1), with Kh values ranging from 10-16 to 10-12 m s-1. Limestones of
152
the Black River Group have Kh values of 10-12 to 10-10 m s-1 and define an aquitard (Fig. 1). The
153
Cambrian sandstone unit has a Kh value of 10-6 m s-1 and represents a basal aquifer (Fig. 1).
TE D
EP
AC C
154
M AN U
142
155
3. Methods
156
3.1 Sampling and Petrography
8
ACCEPTED MANUSCRIPT
157
Samples were obtained from six drill cores (DGR-2 to DGR-6; DGR-8) at the Bruce site, and were initially collected on the basis of visible secondary minerals (calcite, dolomite,
159
anhydrite, celestine, pyrite, halite, etc.) in veins, vugs and replacement/alteration zones, largely
160
within the Trenton and Black River group limestones and the Cambrian sandstones. Sample
161
depths are reported here as metres below ground surface (mBGS) relative to drill hole DGR-2.
162
RI PT
158
Polished thin sections of the secondary minerals were prepared at the University of New Brunswick, and characterized using optical microscopy and electron-beam techniques
164
(Microscopy and Microanalysis Facility at the University of New Brunswick). Secondary
165
minerals were imaged with a JEOL 6400 scanning electron microscope (SEM) using a
166
backscattered electron detector (BSE) and identified using SEM-based energy-dispersive X-ray
167
spectrometry (EDS). Quantitative measurements of major and trace element concentrations (Ca,
168
Mg and Sr) were made using wavelength-dispersive X-ray spectrometry on a JEOL 733 Super-
169
probe electron probe microanalyzer (EPMA), operating at 15 kV with a beam current of 15-20
170
nA. Integration times for each element ranged from 30–80 s and the detection limits were
171
generally lower than 160 ppm for Ca and Mg, and 260 ppm for Sr.
174
M AN U
TE D
EP
173
3.2 Fluid Inclusion Microthermometry Fluid inclusion measurements were undertaken in the Institute of Geological Sciences at
AC C
172
SC
163
175
the University of Bern, and follow the general fluid inclusion methodology outlined by Roedder
176
(1984), Goldstein and Reynolds (1994) and Samson et al. (2003). Double-polished thick
177
sections (150 µm) were prepared for each sample to permit characterization by optical
178
microscopy using transmitted, plane-polarized light and reflected, ultraviolet (UV) light. UV
179
light microscopy was utilized to identify fluid inclusions that contain hydrocarbons.
9
ACCEPTED MANUSCRIPT
180
Microthermometric measurements were made using a Linkam MSD-600 heating-cooling stage mounted on an Olympus BX51 polarizing microscope. The stage was calibrated against
182
synthetic CO2–H2O inclusions and synthetic H2O inclusions with critical density, such that the
183
accuracy is ± 0.1 °C for measurements made below room temperature and ± 1.0 °C for
184
measurements above room temperature. For the microthermometric measurements of inclusions
185
in calcite, careful attention was paid to avoid spurious data and/or anomalously high
186
homogenization temperatures that resulted from inclusion “stretching” (i.e. deformation of the
187
inclusion walls during heating/pressurization; Roedder and Bodnar, 1980; Goldstein and
188
Reynolds, 1994). “Stretched” inclusions were identified by increases in the diameter of the
189
contained vapour bubbles after heating to homogenization and cooling down to room
190
temperature, as visible in photographs taken prior to and after heating.
SC
M AN U
191
RI PT
181
Identification of gas and mineral phases trapped within the inclusions was undertaken using laser Raman spectroscopy. Either red (632.8 nm) He-Ne or green (532.12 nm) Nd-YAG
193
lasers were focused onto the sample through an Olympus 100/0.95 UM PlanFI objective lens on
194
a Olympus BX41 microscope, and Raman spectra were collected for 10–20 s using a Horiba
195
Jobin-Yvon LabRam HR-800 spectrometer calibrated against the emission lines of a neon lamp.
198 199
EP
197
AC C
196
TE D
192
3.3 Stable Isotope Analysis
C- and O-isotope measurements were conducted in the G.G. Hatch Stable Isotope
200
Laboratory at the University of Ottawa using isotope ratio mass spectrometry (IRMS). Calcite
201
and dolomite samples were acquired from thin section chips using a Merchantek microdrill (300
202
µm drill bit). Drill locations were pre-selected in portions of veins, vugs and/or
10
ACCEPTED MANUSCRIPT
alteration/replacement zones using a combination of optical microscopy and SEM-BSE imaging
204
of the associated thin section. The powdered calcite and dolomite samples were weighed into
205
glass reaction vessels and flushed with helium. Orthophosphoric acid was introduced into the
206
vials and reacted for 24 hours at 25 °C for calcite and 50 °C for dolomite. The C- and O-isotope
207
ratios of the resulting CO2 gas were measured using a Thermo-Finnigan Delta XP and Gas
208
Bench II in continuous flow mode, and normalized against NBS-18, NBS-19 and LSVEC
209
(carbon only). C- and O-isotope compositions are reported in δ-notation relative to Vienna
210
Peedee belemnite (VPDB) for carbon and Vienna standard mean ocean water (VSMOW) for
211
oxygen, in permil (‰). Analytical uncertainties associated with the δ13C and δ18O data are ± 0.1
212
‰ (2σ).
M AN U
SC
RI PT
203
213
215
3.4 87Sr/86Sr Analysis
Sr-isotope ratios (87Sr/86Sr) for carbonate samples were obtained in the Isotope
TE D
214
Geochemistry and Geochronology Research Centre (IGGRC) at Carleton University using
217
thermal ionization mass spectrometry (TIMS). Calcite, dolomite, anhydrite and celestine
218
samples were collected by micro-drilling following the same methodology described in Section
219
3.3. Sr-concentrations were determined using mean EPMA analyses for the target mineral.
220
Samples were weighed out between 0.5 and 15 mg (depending on Sr-concentration) and
221
dissolved in 2.5 N nitric acid. Micro-chromatography columns were prepared using 100 µL of
222
Eichrom Sr Spec resin (5 to 100 µm) in a 14 mL borosilicate glass column. The resin was
223
washed with 1200 µL of ultrapure water and then conditioned with 400 µL of 7 N nitric acid.
224
The samples were added to the column, then washed using 7 N nitric acid. Sr ions were eluted
225
from the resin using 1600 µL of ultrapure water, and dried down on a hot plate. The dried Sr
AC C
EP
216
11
ACCEPTED MANUSCRIPT
226
samples were loaded onto Ta filaments using 0.3 N orthophosphoric acid and analyzed for
227
87
228
blanks for the procedure are < 250 picograms. The three-year average for analyses of NIST
229
SRM987 is 87Sr/86Sr = 0.710254 ± 0.000019. Analytical uncertainties associated with the Sr-
230
isotope data are reported at 2σ.
RI PT
Sr/86Sr using a Thermo-Finnigan Triton TIMS, at temperatures of 1300-1500 °C. Total Sr
231 4. Results
233
4.1 Paragenetic sequence and chemistry of secondary minerals
Secondary minerals in the Cambrian sandstones and the Black River and Trenton group
M AN U
234
SC
232
limestones occur as matrix replacement and/or veins and vugs (Fig. 2). The mineral assemblages
236
predominantly consist of calcite and dolomite, with varying proportions of anhydrite, celestine,
237
pyrite and quartz (Fig. 3). Varying amounts of clay minerals and halite also occur as secondary
238
minerals (Intera, 2011). Relative proportions of dolomite versus calcite (as wt.% dolomite) are
239
exhibited in Fig. 4A using bulk XRD data from Intera (2011). Elemental concentrations (Ca, Mg
240
and Sr) are shown in Table 1S (Supporting Material). Saddle dolomite is observed as linings in mm- or cm-wide veins and vugs within: (1)
EP
241
TE D
235
stratabound dolostone in the Sherman Fall, Coboconk, Gull River and Cambrian units, (2) Gull
243
River limestone and (3) Shadow Lake siltstone (e.g., Figs. 2C–E; 2G–H; Fig. 1S – Supporting
244
Material). Replacement dolomite occurs as fine-grained and dolomite-rich, mm- or cm-wide
245
zones in limestone within both the Trenton and Black River groups (e.g., Fig. 2B). Uniform
246
MgO abundances of 18−22 wt.% were obtained for saddle/vein dolomite and dolostone matrix
247
(Fig. 4B). Sr abundances are commonly less than 1000 ppm for all three types of dolomite (Fig.
AC C
242
12
ACCEPTED MANUSCRIPT
248
4C). Dolomitic growth zones in calcite crystals (MgO up to 22 wt.%) are identified in
249
Coboconk, Gull River and Shadow Lake veins and vugs (Fig. 4B). Secondary calcite commonly consists of medium- or coarse-grained, euhedral crystals
251
within mm- to cm-sized veins and vugs, and are the principal vein/vug forming mineral in the
252
Trenton and Black River groups and the Cambrian sandstones (Fig. 1S – Supporting Material).
253
Low MgO abundances (< 2 wt.%) are common for most secondary calcite. Sr abundances in
254
vein and vug calcite exhibit a decreasing trend with depth, from as high as 4300 ppm in the
255
Cobourg to <950 ppm in the Cambrian (Fig. 4C).
SC
Anhydrite and celestine are observed as late infill after saddle dolomite and calcite (Figs.
M AN U
256
RI PT
250
2A; 2D; 2E; Saso, 2013; Fig. 1S – Supporting Material). The abundance of Sr in anhydrite is
258
comparable to calcite in the Cobourg and Sherman Fall formations, but it is higher than calcite in
259
the Gull River Formation (Fig. 4C). Authigenic quartz was identified in one sandstone and one
260
dolostone in the Cambrian, as euhedral quartz rims mantling rounded, detrital quartz cores (Fig.
261
2I), and druse quartz lining the wall of a calcite vein (Fig. 2G). At both of these occurrences,
262
authigenic quartz appears to have formed either prior to or synchronous with calcite formation.
263
In DGR-4-844.78, the quartz lining appears to postdate the growth of saddle dolomite (Fig. 2G).
264
Pyrite is observed throughout the succession, commonly as an early matrix mineral in all
265
limestone units (except the Sherman Fall Formation) and stratabound dolostones in the
266
Coboconk, Gull River, and Cambrian units, and as a late vug mineral, post-dating calcite in
267
DGR-2-843.83 (Fig. 2F).
269
EP
AC C
268
TE D
257
4.2 Petrography and microthermometry of fluid inclusions
13
ACCEPTED MANUSCRIPT
270
Six paleofluid stages were identified from the primary inclusions in dolomite (stage I), primary inclusions in calcite (stage II), primary inclusions in celestine and anhydrite (stage III),
272
and secondary inclusions mostly in calcite (stage IV to VI), and are summarized in Figs. 3–8 and
273
Table 1. Photomicrographs of representative primary and secondary inclusions are shown in
274
Figures 5 and 6, respectively. All inclusion assemblages trapped during stages I to III contain
275
variable proportions of liquid and methane vapour (e.g., Figs. 5B, E, H), indicating
276
heterogeneous fluid entrapment. Therefore, trapping temperatures (Ttrap) for these stages are
277
defined by the minimum homogenization temperature (Th) observed in each inclusion
278
assemblage (Diamond et al., 2015). Uniform proportions of liquid and vapour were observed in
279
assemblages of stage IV inclusions (e.g., Figs. 6B–C, E), thus the maximum Th for each
280
assemblage represents a minimum constraint on Ttrap (i.e. Ttrap > Th,max). All assemblages in
281
stages I to IV contain variable proportions of halite, indicating they were saturated with respect
282
to halite at the time of trapping (principles of these interpretations are explained in Diamond,
283
2003). Stage V inclusions consist of saline brine without vapour bubbles at room temperature
284
(e.g., Fig. 6G). In this case, the only constraint on Ttrap is the empirical rule that they were
285
trapped below 70 °C (Goldstein and Reynolds, 1994; Diamond, 2003). The Ttrap results are
286
summarized in Fig. 7 and Table 1 (detailed Th data are shown in Table 2S – Supporting Material,
287
after Diamond et al., 2015). The relative timing of fluid inclusion entrapment associated with
288
stages IV vs. VI was determined by cross-cutting relationships as shown in Figures 6H and 6I.
SC
M AN U
TE D
EP
AC C
289
RI PT
271
Temperatures of final melting of ice, Tm(ice), in the presence of halite + liquid + vapour
290
were measured in assemblages of inclusions in stages I to IV. No observations of eutectic
291
melting could be made but the span of Tm(ice) values between –52.7 and –34°C suggests the
292
principal salts are NaCl and CaCl2. As commonly observed in CaCl2-rich inclusions, nominally
14
ACCEPTED MANUSCRIPT
stable hydrohalite nucleated only rarely upon cooling and so the measured Tm(ice) transitions
294
represent metastable equilibria. To enable salinity determinations from these data, the locus of
295
the metastable cotectic involving ice + halite + liquid + vapour was calculated for this study (red
296
curve Figs. 8A, B) from the stable phase relations for the H2O–NaCl–CaCl2 ternary system given
297
by Oakes et al. (1990) and Steele-MacInnis et al. (2011). Fluid compositions and salinities in
298
terms of the model ternary components were then obtained for each assemblage by constructing
299
a halite-melting path beginning at Tm(ice) on the cotectic and moving towards the NaCl apex of
300
the ternary. The point at which this path intersects Ttrap on the halite liquidus defines the ternary
301
composition and the salinity of the inclusions (i.e. green arrow in Fig. 8A). For the halite-
302
undersaturated inclusions of stage V, salinity could only be broadly constrained to lie along the
303
relevant isotherms of Tm (ice) (dark bands in Fig. 8B; question marks show range of uncertainty
304
in salinity). Table 1 lists the resulting ternary compositions and salinities. Microthermometric
305
data of all individual inclusions measured are given in Diamond et al. (2015).
SC
M AN U
TE D
306
RI PT
293
Primary, stage I inclusions in saddle dolomite (e.g., Figs. 5A–C) from the Coboconk and Gull River formations yielded the highest Ttrap values; 122 to 128 °C (n = 3) and a Ttrap value of
308
88 °C was obtained for primary inclusions in rhombic dolomite from the Cambrian. Stage I
309
inclusions have Tm (ice) values between −44.1 and −41.5 °C, implying that the fluid had a
310
salinity of 33 to 34 wt.% [CaCl2+NaCl]eq at halite saturation (at the time of Ttrap).
AC C
311
EP
307
Primary stage II inclusions in calcite (e.g., Figs. 5D–F) yielded Ttrap values between 54
312
and 78 °C (mean and S.D. of 65 ± 9 °C; n = 7) and Tm (ice) values between −47.9 and −35.9 °C.
313
Inclusions in two Coboconk samples yielded Tm (hydrohalite) values from −22.9 to −16.5 °C.
314
The calculated salinity of the halite-saturated stage II fluid at Ttrap varies between 30.7 and 33.2
315
wt.% [CaCl2+NaCl]eq.
15
ACCEPTED MANUSCRIPT
316
Stage III is represented by primary inclusions in celestine and anhydrite (e.g., Figs. 5G–I) and by early secondary inclusions in stage II calcite. Crystallites of anhydrite were occasionally
318
identified in stage III inclusions in calcite (Diamond et al., 2015). The inclusions in celestine
319
and calcite yielded Ttrap values between 42 and 80 °C and Tm (ice) values from −46.5 to −33.2
320
°C, corresponding to salinities of 32 to 33 wt.% [CaCl2+NaCl]eq for the halite-saturated stage III
321
fluid at Ttrap.
RI PT
317
Secondary, stage IV inclusions in calcite (e.g., Figs. 6A–E) have minimum Ttrap values
323
between 57 and 106 °C. Their Tm (ice) values vary from −52.7 to −26.4 °C, which indicates a
324
narrow range of halite-saturated salinities between 32 and 34 wt.% [CaCl2+NaCl]eq for this gas-
325
undersaturated fluid at minimum Ttrap.
M AN U
326
SC
322
Late secondary, liquid-only (metastable) inclusions associated with stage V (e.g., Figs. 6F–G) formed at temperatures <70 °C (Goldstein and Reynolds, 1994) and were identified in
328
calcite and dolomite throughout the sample suite. Stage V inclusions from the Gull River
329
Formation and the Cambrian show Tm(ice) values of −35.0 to −24.3 °C, indicating salinities of
330
23 to 26 wt.% [CaCl2+NaCl]eq.
Hydrocarbon inclusions (stage VI) in calcite (e.g., Figs. 6H–I) from the Coboconk and
EP
331
TE D
327
Gull River formations contain variable proportions of liquid oil and methane vapour, indicating
333
they were gas-saturated at the time of trapping. Their minimum Th values yield Ttrap values of 48
334
and 59 °C, respectively.
335 336 337 338
AC C
332
4.3 C- and O-isotope Compositions Calcite vein and vug samples from the Cambrian sandstones yielded the most 13Cdepleted C-isotope compositions in the sedimentary succession, with δ13C values of −6.1 to
16
ACCEPTED MANUSCRIPT
−3.5‰ (Fig. 9A; Table 3S – Supporting Material). Cambrian dolomite samples, including one
340
dolostone and one saddle dolomite, have δ13C values of −2.5‰ and −3.8‰, respectively.
341
Similar C-isotope compositions were obtained for two samples from a vug in the Shadow Lake
342
Formation, including δ13C values of −3.4‰ for saddle dolomite and −5.0‰ for calcite (Table 3S
343
– Supporting Material).
344
RI PT
339
A pronounced C-isotope shift of +2 to +3‰ occurs at the boundary between the Shadow Lake and Gull River Formations. Uniform C-isotope compositions are observed for most
346
samples from the base of the Gull River Formation to the Cobourg Formation, with δ13C values
347
of −1.0 to +1.9‰ and no identifiable differences between the Black River and Trenton group
348
samples. Similarly, no differences in δ13C are observed between samples of limestone matrix
349
and calcite veins and vugs, and samples of saddle dolomite, dolostone and matrix replacement
350
dolomite. Although most of the C-isotope data are constant versus depth, three calcite vein
351
samples from a 3 m section at the base of the Kirkfield Formation (DGR-6-864.60; DGR-6-
352
865.20; DGR-6-867.60) vary by 2.7‰ (−0.8 to +1.9‰; Fig. 9A). It should be noted, however,
353
these data fall within the broader range of C-isotope compositions defined by other samples from
354
the Ordovician succession, including a limestone from the Gull River Formation (−1.0‰) and
355
dolostone from the Coboconk Formation (+1.9‰).
M AN U
TE D
EP
AC C
356
SC
345
Calcite vein and vug samples from the Cambrian have δ18O values between +14.6‰ and
357
+24.2‰ and overall represent the most 18O-depleted compositions in the succession, with only
358
one of these Cambrian calcite samples higher than +17.8‰, (Fig 9B; Table 3S – Supporting
359
Material). Two dolomite samples from the Cambrian units have δ18O values of +23.0‰
360
(dolostone) and +23.2‰ (saddle dolomite). Two samples from a vug in the Shadow Lake
361
Formation (DGR-2-843.83) yielded similar O-isotope compositions to the other respective
17
ACCEPTED MANUSCRIPT
362
calcite and dolomite samples in the Cambrian sample suite, and include δ18O values of +15.9‰
363
for calcite and +20.6‰ for saddle dolomite.
364
A pronounced shift of ~ +3 to +7‰ is observed in the δ18O data along the boundary between the Shadow Lake and Gull River formations (Fig. 9B). Calcite vein and vug samples
366
from the Gull River and Coboconk formations have a wide range of δ18O values, from +18.9 to
367
+26.7‰. Notably, a ~ 10 m section at the top of the Coboconk Formation, along the Trenton–
368
Black River Group boundary, exhibits a spread in δ18O values of up to 5.4‰ (+21.3 to +26.7‰).
369
Calcite veins and vug samples from the Trenton Group have δ18O values from +23.0 to +28.1‰,
370
and are broadly consistent with calcite O-isotope data for the Black River Group. Most of the
371
Trenton Group calcite samples vary by up to 3.1‰, with only one exceptional δ18O value higher
372
than +26.1‰ (Fig. 9B). Six saddle dolomite samples from the combined Black River–Trenton
373
group units have uniform δ18O values of +20.3 to +21.3‰, and are > 1–2‰ lower in δ18O than
374
most calcite vein and vug samples in the Ordovician succession. In contrast, one sample of
375
dolostone matrix (DGR-2-778.44) and one replacement dolomite sample (DGR-4-772.10) from
376
the Coboconk Formation yielded δ18O values of +23.6‰ and +22.0‰, respectively. Eleven
377
limestone matrix samples from the Trenton and Black River groups exhibit a subtle trend of
378
increasing δ18O values towards the surface, from +23.8 to +26.8‰, and no identifiable shifts or
379
sections of O-isotope variability along the boundary between the two groups, as observed in the
380
calcite vein and vug data (Fig. 9B).
382 383 384
SC
M AN U
TE D
EP
AC C
381
RI PT
365
4.4 Sr-isotope ratios Secondary calcite and dolomite samples from the Cambrian units have the highest Srisotope ratios in the sedimentary succession. Four calcite vein/vug samples have 87Sr/86Sr values
18
ACCEPTED MANUSCRIPT
from 0.70975 to 0.71039, while one sample of dolostone matrix and one saddle dolomite yielded
386
similar 87Sr/86Sr values of 0.71043 and 0.71027, respectively (Fig 9C; Table 3S – Supporting
387
Material). Furthermore, two calcite/saddle dolomite samples from a vug in the Shadow Lake
388
Formation have 87Sr/86Sr values of 0.71037 and 0.70984, and overlap in Sr-isotope composition
389
with other samples from the Cambrian.
RI PT
385
A shift in 87Sr/86Sr of up to ~ –0.02 is observed across the boundary between the Shadow
390
Lake and the Gull River formations. Calcite vein and vug samples from the Gull River and
392
Coboconk formations have the highest Sr-isotope variability in the succession, with 87Sr/86Sr
393
values ranging from 0.70799 to 0.70990 (Fig. 9C). Similar variability is observed for three
394
samples of saddle dolomite, one dolostone and one sample of anhydrite from the Gull River–
395
Coboconk units, with combined 87Sr/86Sr values from 0.70854 to 0.70939. In contrast, three
396
limestone samples from these Black River Group units (two Gull River, one Coboconk) have
397
uniform 87Sr/86Sr values of 0.70810 to 0.70833, and overlap with the lowest Sr-isotope ratios of
398
calcite veins from these units.
TE D
M AN U
SC
391
399
For the Trenton Group, five calcite vein/vug samples and three limestone matrix samples have uniform 87Sr/86Sr values of 0.70790 to 0.70817 (Fig. 9C), and are consistent with the
401
87
402
values of 0.70818, 0.70808 and 0.70848 are observed for anhydrite samples from the Cobourg
403
and Sherman Fall formations, and a celestine sample from the Kirkfield Formation, respectively
404
(Fig. 9C). One sample of saddle dolomite from a vug in the Sherman Fall Formation (DGR-5-
405
743.10) has an exceptional 87Sr/86Sr value of 0.70856. However, this sample has an associated
406
2σ analytical uncertainty of ± 0.00083, which is 20–40 times larger than other analyses (Table
407
3S – Supporting Material).
EP
400
AC C
Sr/86Sr values of limestone samples from the underlying Black River Group. Similar 87Sr/86Sr
19
ACCEPTED MANUSCRIPT
408 409
5. Discussion
410
5.1 Paragenetic sequence Overall six distinct fluid events were identified in the microthermometric/petrographic
RI PT
411
results (Figs. 3 and 7), with the first two stages responsible for most of the mineral precipitation
413
in veins and vugs and only minor amounts of anhydrite and celestine that formed during stage
414
III. Based on the preservation of well-defined dolomite, calcite, anhydrite and celestine crystal
415
margins from precipitation events during stages I to III (Fig. 2), and on the absence of dissolution
416
features, fluid stages IV–VI are interpreted as migration of relatively minor quantities of fluids
417
along grain boundaries or micro-fractures.
M AN U
418
SC
412
Stage I is represented by formation of Fe- and Mn-enriched saddle dolomite in the Black River Group (Diamond et al., 2015) that resulted from the influx of a halite-saturated brine (34
420
wt.% [CaCl2+NaCl]eq) and free methane gas at temperatures of 122–128 °C. Elsewhere in the
421
Michigan Basin, high temperatures have been determined for fluid inclusions in saddle dolomite
422
from the Trenton–Black River groups, including up to ~160 °C in the Albion-Scipio oil field of
423
south-central Michigan (Allan and Wiggins, 1993) and temperatures as high as ~170 °C and 220
424
°C in dolomite samples from Manitoulin Island and southwestern Ontario, respectively (Coniglio
425
et al., 1994). Elevated temperatures of ~97–144 °C were also obtained by Haeri-Ardakani et al.
426
(2013) for saddle dolomite samples from the Trenton Group in southwestern Ontario. Davies
427
and Smith (2006) suggested that the formation of saddle dolomite in these fault-controlled
428
reservoirs is due to ascending fluid that is hotter than the ambient or burial temperatures of the
429
host formation (i.e. hydrothermal). In our study, a third Fe- and Mn-enriched saddle dolomite
430
sample from the Cambrian yielded a significantly lower Ttrap value of 88 °C, but a similarly high
AC C
EP
TE D
419
20
ACCEPTED MANUSCRIPT
431
salinity of 34 wt.% [CaCl2+NaCl]eq. It is not known whether these saddle dolomite samples are
432
contemporaneous.
433
Stage II primary inclusions in calcite have Ttrap values of 54–78 °C (mean 65 ± 9 °C; n = 7) and represent a halite-saturated brine with a salinity of 31 to 37 wt.% [CaCl2+NaCl]eq and
435
entrained free methane gas. Based on the predominance of calcite as a vein/vug mineral, it can
436
be presumed that stage II was the most significant vein- and vug-forming fluid event. The range
437
of Ttrap values from stage II inclusions is slightly lower than minimum paleotemperature
438
estimates of 80 °C at the base of the Silurian and 90 °C at the base of the Trenton Group reported
439
by Legall et al. (1981). Determinations of the absolute timing of vein and vug calcite formation
440
in the Trenton–Black River groups have been made using U-Pb dating methods with laser
441
ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) and thermal ionization
442
mass spectrometry (TIMS) (Davis et al., 2013). Davis et al. (2013; 2017) reported a LA-ICP-MS
443
U-Pb age of 434 ± 5 Ma, based on a calculated mean of three model age determinations from the
444
Cobourg and Coboconk formations, and a TIMS U-Pb date of 451 ± 38 Ma for a calcite vein
445
from the Cobourg Formation. The Davis et al. (2013; 2017) U-Pb data compare well with a Rb-
446
Sr isochron age of 425 ± 6 Ma (errorchron age of 440 ± 50 Ma) by Harper et al. (1995) and 7
447
combined K-Ar dates ranging from 453 ± 9 Ma to 412 ± 8 Ma by Harper et al. (1995) and
448
Ziegler and Longstaffe (2000) (weighted mean of 442 ± 16 Ma) for secondary K-feldspar
449
obtained along the Precambrian−Paleozoic boundary in southern Ontario. These data suggest the
450
influx of stage II, calcite-forming fluids in the Trenton–Black River groups likely occurred
451
between the Late Ordovician and Silurian.
452 453
AC C
EP
TE D
M AN U
SC
RI PT
434
Stage III fluid inclusions have Ttrap values of 42–80 °C and represent a stage of fluid migration during which anhydrite and celestine were formed. The mean Ttrap value of 62 ± 17 °C
21
ACCEPTED MANUSCRIPT
for four stage III inclusions is similar to the mean Ttrap value (65 ± 9 °C; n = 7) for stage II
455
primary inclusions in calcite. Salinities of stage III inclusions range from 32 to 33 wt.%
456
[CaCl2+NaCl]eq, and are comparable to the salinity estimates for stage I and stage II inclusions.
457
This suggests a paragenetic link between the stage I–III fluids, with saddle dolomite forming
458
during initial brine influx at 122–128 °C (stage I), then subsequent cooling and formation of
459
calcite (stage II) and penecontemporaneous formation of anhydrite and celestine (stage III). Secondary fluid inclusions defining stage IV have minimum Ttrap values between 57 and
SC
460
RI PT
454
106 °C, including one calcite vein from the Sherman Fall Formation (DGR-2-691.70; see Figs.
462
6D–E) that yielded minimum Ttrap values ranging from 77 to 106 °C. The absolute timing and
463
burial depth (i.e. thickness of overburden sediments) at the time of stage IV fluid inclusion
464
entrapment is unclear, therefore, definitive pressure corrections to determine Ttrap (true) cannot
465
be applied to the stage IV data. In order to determine an upper limit for stage IV fluid inclusion
466
entrapment, Ttrap (maximum) values were calculated using pressure corrected data for peak burial
467
conditions for the Ordovician–Cambrian succession. This is a conservative approach because we
468
do not know that stage IV coincides with peak burial, and because all corrections are made to the
469
depth at the base of the Paleozoic succession. After peak burial conditions, it can be inferred
470
from the work of Coniglio and Williams-Jones (1992) that approximately 1000 m of Paleozoic
471
sediments were eroded from the sedimentary succession at the Bruce site (NWMO, 2011).
472
Utilizing an overburden thickness of 1000 m and a current depth of 860 m for the base of the
473
Paleozoic succession, a maximum burial depth of 1860 m was used to calculate an upper limit
474
for stage IV fluid inclusion entrapment. Pressure corrected Ttrap (maximum) values for the stage
475
IV inclusions range from 77–127 °C (Fig. 7; Table 1).
AC C
EP
TE D
M AN U
461
22
ACCEPTED MANUSCRIPT
There are two models that explain the formation of stage IV fluids: (1) influx of
477
hydrothermal brines, or (2) regional heating of the sediments (i.e. through conduction), coupled
478
with fracturing and re-mobilization of fluids. For model 1, significant quantities of fluid and
479
high flow rates along permeable structures would be required to facilitate the transfer of heat, but
480
there is no evidence for fluid movement at this scale (e.g., large fracture/veins systems,
481
dissolution/re-precipitation features, formation of new secondary minerals with primary
482
inclusions analogous to stage IV). Alternatively, a regional heating event (model 2) is consistent
483
with an estimated minimum paleotemperature of 90 °C at the base of the Trenton Group (Legall
484
et al., 1981). Evidence for a regional heating event exists elsewhere in the Michigan Basin.
485
Temperatures of ~70–170 °C and ~40–260 °C have been obtained for secondary fluid inclusions
486
in calcite from Trenton–Black River rocks on Manitoulin Island and in southwestern Ontario,
487
respectively (Coniglio et al., 1994). Haeri-Ardakani et al. (2013) also reported temperatures of
488
~60–131 °C for fluid inclusions in calcite cement samples from the Trenton Group in southwest
489
Ontario. In central Michigan, fluid inclusions in Middle Ordovician St. Peter Formation
490
sandstones yielded temperatures of 76–169 °C for quartz overgrowths, and 94–149 °C for
491
dolomite cements (Girard and Barnes, 1995), and are consistent with temperatures of ~ 100–160
492
°C obtained for fluid inclusions in saddle dolomite from elsewhere in central Michigan (Allan
493
and Wiggins, 1993). Overall, these high temperatures (up to 260 °C) strongly suggest that
494
regional heat sources were present at depth below the Michigan Basin. Reactivation of the mid-
495
continent rift was suggested by Ma et al. (2009) as a source of heat and mantle-derived He and
496
Ne anomalies in the hydrothermal fluids that interacted with the basin rocks. Regional-scale
497
heating has also been linked to the formation of secondary illite at 367–322 Ma in the St. Peter
498
sandstone along the base of the Paleozoic succession (Girard and Barnes, 1995). Reactivation of
AC C
EP
TE D
M AN U
SC
RI PT
476
23
ACCEPTED MANUSCRIPT
499
the mid-continent rift during the Late Devonian–Mississippian could be responsible for initiating
500
regional-scale heating below the Michigan Basin and may be linked to the stage IV fluid
501
migration event observed in the Ordovician carbonates at the Bruce site. Stage V and VI inclusions represent the latest observed records of fluid migration in the
503
sedimentary succession, and for the latter stage VI fluids, include hydrocarbon inclusions in the
504
Coboconk and Gull River formations and the Cambrian units. Samples from the Coboconk and
505
Gull River formations yielded trapping temperatures of 48 and 59 °C (Fig. 7). The occurrence of
506
hydrocarbon fluids is consistent with the presence of bitumen layering or oil in the Coboconk,
507
Gull River and Shadow Lake formations (see Fig. 2.32 in Intera, 2011), suggesting that stage VI
508
represents an important period of hydrocarbon generation and/or migration in the Black River
509
Group.
510 5.2 Source and evolution of fluids
TE D
511
M AN U
SC
RI PT
502
By combining the petrographic and fluid inclusion results with the C- O- and Sr-isotope
513
data, inferences can be made about the sources and paleofluid evolution through the secondary-
514
mineral formation events of stages I to III. Secondary calcite (veins/vugs) and dolomite
515
(dolostone matrix and saddle) samples from the combined Cambrian–Shadow Lake units have
516
δ13C values of ~ −6.1 to −2.5‰ (Fig. 9A). Estimates of the average C-isotope composition of
517
Cambrian seawater include δ13C values between −0.9‰ and +0.0‰ (Fig. 9A; Veizer et al.,
518
1999), suggesting that the calcite and dolomite samples from the Cambrian–Shadow Lake units
519
were not sourced from connate fluids derived from Cambrian seawater. Alternatively, these low
520
δ13C values reflect carbonate mineral formation from fluids that reacted with organic carbon.
521
Harper et al. (1995) reported similar 13C-depleted data (−9.3 to −3.9‰) for secondary calcite and
AC C
EP
512
24
ACCEPTED MANUSCRIPT
522
dolomite associated with hydrothermally altered granitoid rocks along the
523
Precambrian−Paleozoic boundary in southern Ontario.
524
A 2−3‰ shift in δ13C occurs at the boundary between Cambrian−Shadow Lake units and the overlying Gull River Formation. The C-isotope compositions of calcite and dolomite
526
samples show little scatter through the Gull River to Cobourg formations (Fig. 9A) and are
527
consistent with estimates by Veizer et al. (1999) for the average C-isotope composition of
528
seawater during the Late Ordovician (−0.1 to +1.7‰), suggesting a C source dominated by
529
seawater-derived connate fluids or external fluids that experienced high degrees of fluid–rock
530
interaction with the host limestones (Fig. 9A).
SC
M AN U
531
RI PT
525
Calcite samples from the Shadow Lake and Cambrian units generally have the lowest δ18O values (+14.6 to +24.2‰), with most calcite veins and vugs plotting at ~ +16‰ (7 of 8
533
calcite veins form a calculated mean of +15.9 ± 1.0‰; S.D.). A prominent feature of the O-
534
isotope data through the Cambrian−Ordovician succession is the >4‰ shift that occurs along the
535
boundary between the Cambrian–Shadow Lake units and the overlying Gull River Formation
536
(Fig. 9B). Relatively high variability is observed in the δ18O data for the Coboconk Formation
537
just below the Trenton−Black River group boundary (Fig. 9B), with values from calcite vein and
538
vug samples exhibiting a spread of up to 5.4‰. This variability could be explained by: (1)
539
mixing between two or more fluids that had distinct O-isotope compositions (i.e. connate vs.
540
external hydrothermal fluids), or (2) temperature-related differences in ∆18Ocalcite−fluid during the
541
precipitation of calcite (e.g., Friedman and O'Neil, 1977). The fact that large temperature
542
differences are not observed in the primary fluid inclusion data for the Coboconk Formation
543
(Fig. 7) suggests that the δ18O variability may be explained by mixing with external fluids. The
544
Trenton−Black River boundary represents a significant change in hydraulic conductivity, with
AC C
EP
TE D
532
25
ACCEPTED MANUSCRIPT
rocks in the Trenton Group having Kh values up to six orders of magnitude lower than those of
546
the Black River Group (Normani and Sykes, 2012; Beauheim et al., 2014; Khader and
547
Novakowski, 2014). Furthermore, thin dolostone units (< 20 cm) and a bentonite unit (< 10 cm
548
thick) occur as stratabound layers in the Coboconk Formation (~ 790.50 m and 781.00 mBGS,
549
respectively). These units, particularly the dolostone, may have relatively high permeability and
550
porosity compared to the bounding limestones, and as such, could represent pathways of
551
enhanced lateral flow, allowing for ingress of external fluids.
SC
552
RI PT
545
The O-isotope compositions of source fluids responsible for secondary mineral formation were estimated using the microthermometric data and mineral−fluid O-isotope exchange
554
thermometers for calcite (Friedman and O'Neil, 1977; Fig. 10A) and dolomite (Horitia, 2014;
555
Fig. 10B). The O-isotope composition of Late Ordovician seawater is shown in Figure 10, with
556
δ18O values between −4 and −2‰ (Veizer et al., 1999; Shields et al., 2003). Calculated values
557
for the O-isotope composition of fluids in equilibrium with all limestones from the Gull River to
558
Cobourg formations are consistent with Ordovician seawater (Fig. 10A), indicating that there has
559
been limited diagenetic or hydrothermal alteration effects to their primary O-isotope
560
compositions.
TE D
EP
561
M AN U
553
Calcite vein and vug samples from the Cambrian−Shadow Lake units have calculated δ18Ofluid values between ~ −6‰ and +4‰, based on formation temperatures of 70 to 72 °C (Fig.
563
10A). Gull River and Coboconk samples have similar calculated δ18Ofluid values of ~ −6 to
564
+4‰, while samples from the Trenton Group have calculated fluid compositions of ~ +2 to +8‰
565
(Fig. 10A). These fluid compositions are suggestive of calcite precipitation from Late
566
Ordovician seawater or an 18O-enriched, seawater-derived basin fluid. Similar 18O-enrichments
567
relative to seawater have been described for fluids in many sedimentary basins (Holser, 1979;
AC C
562
26
ACCEPTED MANUSCRIPT
Knauth and Beeunas, 1986; Ayalon and Longstaffe, 1988; Kyser and Kerrich, 1990; Wilson and
569
Long, 1993a; 1993b; Hanor, 1994), and are interpreted to have undergone 18O-enrichment by
570
evaporation and/or fluid−rock interaction following burial. Modification of seawater-derived
571
connate fluids by influx of hydrothermal fluids that interacted with, or were derived from the
572
Precambrian shield (Spencer, 1980) could also explain such 18O-enrichments.
573
RI PT
568
Calculated fluid compositions for saddle dolomite have δ18Ofluid values of ~ 0 to +7‰ (Fig. 10B) and overlap with the calculated fluid compositions of calcite, suggesting that dolomite
575
precipitated from 18O-enriched hydrothermal fluids originating from evolved seawater. No
576
microthermometry data could be obtained from fluid inclusions in samples of matrix replacement
577
dolomite and dolostone. However, if the matrix replacement dolomite and dolostone samples
578
formed at temperatures similar to those of calcite veins and vugs (54 to 78 °C) or saddle
579
dolomite (88 to 128 °C), the calculated fluid compositions would be −6 to +1‰ or +1 to +7‰,
580
respectively (Fig. 10B). These calculated fluid compositions are suggestive of matrix
581
replacement dolomite and dolostone formation directly from Late Ordovician seawater or
582
hydrothermal fluids derived from evolved seawater, and are broadly consistent with fluid source
583
determinations for all other secondary minerals.
M AN U
TE D
EP
584
SC
574
Sr-isotope data for many of the samples from the Black River Group, and all samples from the Cambrian, exhibit 87Sr/86Sr enrichment well above the seawater curve of Veizer et al.
586
(1999). Enriched 87Sr/86Sr values (> 0.710) are a common signature of Precambrian brines (e.g.,
587
Frape et al., 1984; McNutt et al. 1984; Bottomley et al., 1999), and the observed enrichment
588
suggests vein and vug formation from shield-derived hydrothermal fluids, and/or fluid–rock
589
interaction with shield-derived detrital minerals in the Cambrian sandstones. In contrast, all vein
590
and vug samples from the Trenton Group, and approximately half of the samples from the Black
AC C
585
27
ACCEPTED MANUSCRIPT
River Group, overlap with the seawater curve of Veizer et al. (1999), suggesting they formed
592
from either connate fluids derived from Late Ordovician seawater and/or external hydrothermal
593
brines that underwent high degrees of fluid–rock interaction and associated modification of their
594
original isotopic signature during transit through the underlying stratigraphy.
RI PT
591
The combined microthermometry and isotopic data are consistent with a model of
596
secondary mineral formation by fluid mixing; with one end member being either a shield-derived
597
hydrothermal fluid, and/or a fluid that was isotopically influenced by detrital minerals in the
598
Cambrian sandstone, and the second end member being a connate pore fluid derived from the
599
Paleozoic carbonates.
M AN U
SC
595
600 601 602
5.3 Implications for the migration of high-temperature fluids in the Ordovician limestones An important feature of our fluid influx model for the Ordovician sedimentary succession is the formation of dolomite in the Black River Group at temperatures of 122–128 °C, which are
604
higher than the nearest available estimate of peak burial temperature, 55–90 °C from Coniglio
605
and Williams-Jones (1992) at Manitoulin Island, providing evidence for migration of
606
hydrothermal fluid.
EP
TE D
603
Hydrothermal dolomite is an important reservoir for oil and gas plays within the
608
Trenton−Black River groups in the Albion-Scipio/Stoney Point fields in south-central Michigan
609
(Hurley and Budros, 1990; Allan and Wiggins, 1993; Davies and Smith, 2006), the Lima-Indiana
610
field in Ohio/Indiana (Wickstrom et al., 1992) and southern Ontario (Powell et al., 1984;
611
Middleton et al., 1993; Coniglio et al., 1994; Obermajer et al., 1999). In these regions,
612
hydrothermal dolomite developed along large-scale extensional faults as a result of high-
613
temperature fluid−rock interaction (up to ~ 260 °C; Coniglio et al., 1994) between limestone and
AC C
607
28
ACCEPTED MANUSCRIPT
hydrothermal fluids that migrated upward from the underlying Cambrian aquifer. Hydrothermal
615
dolomite reservoirs are recognized in 2D seismic data aligned with vertical “sag” or “flower”
616
structures, often 10s of kilometers in strike length and 100s of meters in width (e.g., Davies and
617
Smith, 2006). At that scale, these structures represent regionally extensive zones of
618
dolomitization. At the Bruce site, however, the abundance and scale of hydrothermal dolomite
619
occurrence is much different. Saddle dolomite with associated high fluid trapping temperatures
620
was observed in veins within Ordovician limestone, stratabound dolostone in the Ordovician, and
621
in Cambrian dolostone. These features are typically mm or cm in thickness (e.g., Fig 2B–E; 2G;
622
2H; Fig. 1S – Supporting Material), and as such, do not represent reservoir-scale structures of
623
dolomitization as observed elsewhere in the Michigan Basin.
M AN U
SC
RI PT
614
624 625
5.4 Conceptual model
Models for the source and timing of paleofluid movements in the Cambrian and
627
Ordovician sedimentary units below the Bruce site, until now, have been largely based on direct
628
geochemical and isotopic constraints from porewater extracted from drill cores (Clark et al.,
629
2010a; 2010b; 2013; 2015; Al et al., 2015). In a detailed study, Clark et al. (2013) reported O-
630
and H-isotope compositions and Cl- and Br-concentration data for porewater obtained
631
throughout the Paleozoic sedimentary succession. Downward trends of decreasing Cl− and Br−
632
concentrations, coupled with decreases in δ18O, were observed in the Ordovician limestones
633
(Fig. 2 in Clark et al., 2013). These trends were interpreted to represent a mixing profile
634
between hypersaline evaporated seawater from the overlying Silurian Salina Formation and
635
shield-derived fluids (Clark et al., 2013; Al et al. 2015). The influence of a shield end member is
636
supported by a downward trend of 2H-enrichment – a common feature of shield fluids (Fritz and
AC C
EP
TE D
626
29
ACCEPTED MANUSCRIPT
637
Frape, 1982; Frape et al., 1984; Pearson, 1987; Bottomley et al., 1999, 2004; Douglas et al.,
638
2000; Gascoyne, 2004; Al et al., 2015).
639
Clark et al. (2013) also reported He-isotope ratios for porewater from the Ordovician shales and the Trenton and Black River group limestones. Porewater extracted from the shales
641
and the Cobourg Formation limestone yielded uniform R/RA values (0.020 ± 0.002; where R/RA
642
= (3He/4He)sample/(3He/4He)air, and (3He/4He)air = 1.38 x 10-6), which are consistent with an
643
authigenic origin, and the total He concentrations indicated a minimum period of 260 Ma for He
644
ingrowth. A positive shift in R/RA values was observed from the base of the Cobourg to the
645
upper portion of the Sherman Fall Formation, followed by uniform R/RA values of 0.035 ± 0.002
646
with increasing depth into the Cambrian (Fig. 4; Clark et al., 2013). The elevated R/RA values
647
below the Cobourg Formation limestone were interpreted by Clark et al. (2013) to reflect
648
allochthonous fluids of mixed basin and shield origin.
SC
M AN U
Furthermore, 87Sr/86Sr data were reported by Clark et al. (2010a; 2010b) for porewater
TE D
649
RI PT
640
samples obtained at varying depths in the sedimentary succession. Porewater samples from the
651
Ordovician shales yielded uniform Sr-isotope compositions (mean 87Sr/86Sr value of ~ 0.710)
652
that are enriched in radiogenic Sr relative to the seawater curve of Veizer et al. (1999). Similar
653
to the interpretation of the He isotope data, Clark et al. (2010a; 2010b) considered the 87Sr
654
enrichment to have resulted from in situ ingrowth by decay of 87Rb in the porewater. Porewater
655
samples from the Trenton and Black River group limestones yielded highly variable 87Sr/86Sr
656
values, between 0.7093 and 0.7103, and consistent with the He-isotope data, were interpreted as
657
resulting from mixing between basin- and shield-derived fluids.
658 659
AC C
EP
650
The isotopic and fluid inclusion data from secondary minerals are generally consistent with the conceptual model for porewater evolution presented by Clark et al. (2013) and Al et al.
30
ACCEPTED MANUSCRIPT
(2015), but our new data allow for some refinements to the model. Based on the porewater data,
661
Clark et al. (2013) and Al et al. (2015) suggest that mixing between basin- and shield-derived
662
fluids was driven by diffusion, but the secondary mineral data indicate that advection in small
663
fractures also contributed to upward transport of shield-type, halite-saturated fluids during stages
664
I to III. These events formed the secondary vein minerals and are broadly constrained by U-Pb
665
dating (Davis et al., 2013) to the Late Ordovician or Silurian. The isotopic signature of the
666
shield end member does not extend upward into the Trenton Group, likely because it was
667
eliminated by fluid−rock interaction. The fact that present-day porewater in the Black River
668
Group is below halite saturation is consistent with indications of allochthonous porewater from
669
87
670
fluid migration that are not recorded in the chemical and isotopic composition of the vein
671
minerals.
M AN U
SC
RI PT
660
Sr/86Sr and 3He/4He ratios, and likely reflects the later stages (V to VI) of halite-undersaturated
673 674
Conclusions
TE D
672
Overall, the combined microthermometric and isotopic data allow us to establish a model for the fluid migration history associated with the formation of secondary vein minerals in low-
676
permeability Ordovician limestones of the Michigan Basin. Six stages of fluid migration are
677
suggested by the fluid inclusion data. These comprise fluid stages I through III, which are
678
represented by primary inclusions trapped during the formation of dolomite (stage I: 88–128 °C),
679
calcite (stage II: 54–78 °C) and celestine/anhydrite (stage III: 42–80 °C). The ubiquitous
680
presence of variable amounts of halite in these fluid inclusions indicates that the secondary vein
681
minerals precipitated from halite-saturated brines with salinities of 31–37 wt.% [CaCl2+NaCl]eq,
682
and is suggestive of a paragenetic link between fluid stages I–III.
AC C
EP
675
31
ACCEPTED MANUSCRIPT
683
Secondary, stage IV inclusions in calcite have a similar, halite-saturated fluid composition as stages I–III, but were trapped above 57–106 °C. No new secondary vein
685
minerals formed during this stage and dissolution features are absent from pre-existing secondary
686
minerals. The stage IV inclusions are interpreted as re-mobilization of trapped fluids along grain
687
boundaries or micro-fractures during a Late Devonian–Mississippian regional heating event.
688
Stages V and VI represent the latest recognized fluid migrations in the sample suite, during
689
which a halite-undersaturated brine (23–26 wt.% [CaCl2+NaCl]eq) was trapped at temperatures <
690
70 °C.
Secondary minerals from the Cambrian and Shadow Lake units are characterized by the
M AN U
691
SC
RI PT
684
lowest δ13C and δ18O values in the succession and 87Sr/86Sr ratios that reflect formation from
693
radiogenic Sr sources. The combined Cambrian–Shadow Lake isotopic data are suggestive of
694
secondary mineral formation from hydrothermal brines that experienced high degrees of fluid–
695
rock interaction with the underlying shield, and/or with shield-derived minerals in the Cambrian
696
sandstone. For the Coboconk and Gull River formations, greater variability in the δ18O and
697
87
698
connate fluids and hydrothermal brines, with the latter having a shield signature sourced from
699
below. Uniform C- O- and Sr-isotope compositions for Trenton Group samples are suggestive of
700
secondary mineral formation from either connate fluids, which were sourced directly from Late
701
Ordovician seawater (or 18O-enriched, evolved seawater), or mixed with hydrothermal brines
702
that underwent high degrees of fluid–rock interaction.
703
TE D
692
AC C
EP
Sr/86Sr values of the secondary minerals is interpreted to reflect mixed fluid sources, including
Age constraints on the absolute timing of secondary mineral formation (stages I–III) are
704
gleaned from previous geochronology studies in the region, which indicate that brine influx
705
occurred during the Late Ordovician to Silurian. Secondary inclusions associated with stage IV–
32
ACCEPTED MANUSCRIPT
VI inclusions did not form new secondary vein minerals and these late fluid stages are
707
interpreted as re-mobilization and/or minor fluid migration along grain boundaries and micro-
708
fractures during a regional heating event. The timing of this regional heating event cannot be
709
easily constrained from our data, however, elsewhere in the Michigan Basin it has been
710
suggested that reactivation of the mid-continent rift during the Late Devonian–Mississippian
711
could be responsible for regional-scale heating and associated fluid migration. Despite evidence
712
elsewhere in the Michigan Basin for secondary-mineral formation during the Late Devonian–
713
Mississippian, the Ordovician limestones at the Bruce site show no evidence of large-scale fluid
714
ingress during this time period (e.g., dissolution/re-precipitation features, formation of new
715
secondary vein minerals); which suggests that the low permeability nature of these rocks had
716
been established before this time.
717
719
Acknowledgements
TE D
718
M AN U
SC
RI PT
706
We are grateful to Paul Middlestead, Wendy Abdi and Patricia Wickham for assistance in the G.G. Hatch Stable Isotope Laboratory (uOttawa), and to Douglas Hall and Steven Cogswell
721
for help with EPMA and SEM work at the University of New Brunswick. Shuangquan Zang,
722
E.A. Spencer and Rhea Mitchell are acknowledged for their assistance with Sr-isotope work in
723
the Isotope Geochemistry and Geochronology Research Centre (Carleton). Rita Caldas kindly
724
assisted with initial microthermometry. We gratefully acknowledge Ian Clark and Josué Jautzy
725
(uOttawa) for discussions on fluid migration models in the Michigan Basin. We are also grateful
726
to Don Davis and Chelsea Sutcliffe (U of T) for thoughtful discussions on the absolute timing of
727
calcite vein emplacement in the Ordovician carbonates. An anonymous reviewer and Laura
728
Kennell-Morrison and Richard Crowe (NWMO) are thanked for comments and suggestions on
AC C
EP
720
33
ACCEPTED MANUSCRIPT
729
an earlier version of the manuscript. Financial support for this project was provided by the
730
NWMO and NSERC, through an NSERC Industrial Postgraduate Scholarship to JS and an
731
NSERC/NWMO Collaborative Research and Development grant (477852-14) held by TA.
RI PT
732 733 Tables
735
Table 1. Microthermometric results for fluid inclusions in the secondary minerals.
SC
734
736
M AN U
737 Figure Captions
739
Figure 1. Bedrock geology, stratigraphy and hydrostratigraphy of the Michigan Basin below the
740
Bruce site, Southwestern Ontario, Canada. Also shown here are regional structural features
741
(Algonquin Arch, Chatham Sag) and contours lines for the base of the Paleozoic sediments (i.e.
742
elevation in metres to the top of the Precambrian basement rocks). The geologic map is based on
743
Ontario Geological Survey (1991) and Armstrong and Carter (2010).
EP
744
TE D
738
Figure 2. Photomicrographs of secondary minerals in the Trenton Group (A–B), Black River
746
Group (C–F) and the Cambrian units (G–I). (A) DGR-5-704.44, sharp-edged calcite vein in an
747
argillaceous limestone, with an infill of anhydrite that post-dates calcite formation (SL); (B)
748
DGR-2-691.70, irregular calcite vein in a fossiliferous limestone with zones of fine-grained
749
replacement dolomite (SL); (C) DGR-2-778.44, dolostone with zones of tan saddle dolomite and
750
discrete veins of colourless saddle dolomite (SL); (D) DGR-2-817.41, micritic limestone with a
751
saddle dolomite vug with infill anhydrite. Note the fine-grained dolomite lining the interior of
AC C
745
34
ACCEPTED MANUSCRIPT
vug (PPL); (E) DGR-4-829.53, dolostone with saddle dolomite vug, infill anhydrite and discrete
753
veins of colourless dolomite (SL); (F) DGR-2-843.83, micritic limestone with a sparry calcite
754
vug, infill minerals include rough-edged pyrite partly replaced and followed by crystallites of
755
magnetite suspended in an organic-rich matrix (RL); (G) DGR-4-844.78, irregular vein with
756
saddle dolomite and quartz lining and calcite infill within a fine-grained dolostone (RL); (H)
757
DGR-4-845.20, dolostone with saddle dolomite-lined calcite vug (PPL); (I) DGR-2-847.42,
758
sandstone with a calcite-lined vug. Note the euhedral authigenic quartz rims that mantle detrital
759
quartz grains (XPL). Abbreviations: Anh – anhydrite; Cal – calcite; Dol – dolomite; Mag –
760
magnetite; PPL – plane-polarized light image; Py – pyrite; RL – reflected light image; SDol –
761
saddle dolomite; SL – stereo-light image (combination of both transmitted and reflected light);
762
Qtz – quartz; XPL – cross-polarized light image.
M AN U
SC
RI PT
752
763
Figure 3. Relative timing of secondary mineral formation and trapping of fluid inclusion stages
765
I–VI, based on detailed petrography. The timing of authigenic quartz and pyrite formation
766
relative to fluid stages I–II and III–V, respectively, is uncertain and indicated by the dashed lines.
767
Note: matrix replacement dolomite predates the formation of saddle dolomite.
EP
768
TE D
764
Figure 4. Compositions of the secondary minerals in the DGR samples, including wt.% dolomite
770
relative to calcite (A), which was determined using XRD (Saso, 2013) and MgO (B) and Sr (C)
771
concentrations determined using EPMA. Note, the location of stratabound dolostone layers in
772
the succession are indicated in (B) by block symbols. Boundaries in the sedimentary succession
773
are also shown here (in descending order): Queenston–Collingwood/Cobourg (Trenton Group);
AC C
769
35
ACCEPTED MANUSCRIPT
774
Kirkfield (Trenton Group)–Coboconk (Black River Group); Shadow Lake (Black River Group)–
775
Cambrian; Cambrian–Precambrian.
776 Figure 5. Photomicrographs of fluid inclusions representative of paleofluid stages I–III, all
778
viewed in plane-polarized, transmitted light. (A) Stage I in DGR-4-829.53, saddle dolomite
779
crystal showing curved crystal faces lining the walls of a vug. The core is rich in primary fluid
780
inclusions and is overgrown by a clear, inclusion-free outer rim; (B–C) Details of A showing
781
variable proportions of liquid brine, CH4–CO2–vapour and halite within individual inclusions of
782
the same assemblage, indicating that all three phases were present at mutual saturation during
783
fluid entrapment; (D) Stage II in DGR-3-799.15, calcite crystals in a vug, containing clouds of
784
three-dimensionally distributed primary fluid inclusions. The rims of these crystals contain
785
linear arrays of fluid inclusions that radiate outwards, parallel to the growth direction of the
786
crystal, and are indicative of primary inclusions. All are younger than inclusions in A, because
787
calcite systematically overgrows saddle dolomite (Fig. 2H); (E–F) Details of D showing variable
788
proportions of liquid brine, CH4–vapour and halite within individual inclusions of the same
789
assemblage, indicating that all three phases were present at mutual saturation during fluid
790
entrapment; (G) Stage III in DGR-4-829.53, primary inclusions within fibrous anhydrite. The
791
inclusions are younger than in D, because anhydrite systematically overgrows calcite (Fig. 2A);
792
(H–I) Details of G showing variable proportions of liquid brine, CH4–vapour and halite within
793
individual inclusions of the same assemblage, indicating that all three phases were present at
794
mutual saturation during fluid entrapment.
AC C
EP
TE D
M AN U
SC
RI PT
777
795
36
ACCEPTED MANUSCRIPT
Figure 6. Photomicrographs of fluid inclusions representative of paleofluid stages IV–VI, all
797
viewed in plane-polarized, transmitted light. (A) Stage IV in DGR-2-847.42, planar array of
798
secondary fluid inclusions on a healed fracture in vein calcite; (B–C) Details of A showing
799
uniform proportions of liquid brine and H2O–vapour but variable proportions of halite within
800
individual inclusions of the same assemblage. The inclusions homogenize to liquid, therefore
801
only brine and halite were present (without free vapour) at mutual saturation during fluid
802
entrapment; (D) Stage IV in DGR-2-691.70, planar array of secondary inclusions on a healed
803
fracture in vein calcite, as in A; (E) Detail of D showing the same relations as in (B). This key
804
assemblage homogenizes to liquid at 106 °C; Sample (F) Stage V in DGR-2-844.31, planar
805
array of secondary liquid-only inclusions (metastable stretched water) on a healed fracture in
806
vein calcite. Such assemblages cross-cut stage IV assemblages (not shown); (H) Stages IV and
807
VI in DGR-6-892.99, viewed in simultaneous transmitted and UV epi-illumination.
808
Interpretation of relative timing of secondary fluid inclusion generations based on cross-cutting
809
relations. Traces of two healed fractures are visible, each marked by secondary fluid inclusions.
810
Green: Stage IV aqueous inclusions within a fracture trace that strikes to top-left of image and
811
dips steeply to top-right. Blue: Stage VI blue fluorescent liquid hydrocarbon inclusions with
812
CH4-vapour bubbles in a fracture trace that strikes to top-right and dips shallowly to bottom-
813
right. No inclusion is present at the intersection of the fracture traces (at shallow focus), so the
814
relative ages of the two fractures are ambiguous; (I) Same view as H but at a deeper focus level.
815
A hydrocarbon inclusion is present at the intersection of the fracture traces, indicating that the
816
hydrocarbon inclusions and their host fracture are younger than the aqueous inclusions and their
817
corresponding host fracture.
AC C
EP
TE D
M AN U
SC
RI PT
796
818
37
ACCEPTED MANUSCRIPT
Figure 7. Microthermometric data for assemblages of primary and secondary fluid inclusions in
820
the secondary minerals. For the stage IV inclusions, which have uniform liquid–vapour
821
proportions, each symbol represents a range of estimated temperatures between the minimum
822
Ttrap value (i.e. Th,max) and a pressure corrected Ttrap value, which was calculated for peak burial
823
conditions using an estimated burial depth of 1860 m (see section 5.1). Estimated Ttrap values of
824
30–70 °C are indicated for stage V inclusions. Boundaries in the sedimentary succession are
825
also shown here (in descending order): Queenston–Collingwood/Cobourg (Trenton Group);
826
Kirkfield (Trenton Group)–Coboconk (Black River Group); Shadow Lake (Black River Group)–
827
Cambrian; Cambrian–Precambrian.
M AN U
828
SC
RI PT
819
Figure 8. Salinity estimates for assemblages of primary and secondary fluid inclusions in the
830
secondary minerals, derived from Tm (ice) values for all fluid stages (A–B) shown in a H2O–
831
NaCl–CaCl2 ternary model. Stable phase relations of Oakes et al., (1990) and Steele-MacInnis et
832
al. (2011) are also shown in (A). The metastable cotectic in (A) and (B) was calculated in this
833
study.
EP
834
TE D
829
Figure 9. C-, O- and Sr-isotope results for the secondary minerals in the Ordovician–Cambrian
836
sedimentary succession (A–C, respectively). Average C-isotope compositions of seawater
837
during the Late Ordovician (−0.1 to +1.7‰) and the Cambrian (−0.9 to +0.0‰) are also shown
838
in (A) and are based on marine carbonate rocks from Veizer et al. (1999). Average 87Sr/86Sr
839
values for Late Ordovician seawater (0.70887 to 0.70941) were obtained from Veizer et al.
840
(1999). The 87Sr/86Sr values of Precambrian shield brines from the Sudbury region (0.71037 to
841
0.74009) were reported by McNutt et al. (1984). Boundaries in the sedimentary succession are
AC C
835
38
ACCEPTED MANUSCRIPT
842
also shown here (in descending order): Queenston–Collingwood/Cobourg (Trenton Group);
843
Kirkfield (Trenton Group)–Coboconk (Black River Group); Shadow Lake (Black River Group)–
844
Cambrian; Cambrian–Precambrian.
RI PT
845 846
Figure 10. Estimates of O-isotope compositions of fluids (contours) that precipitated carbonate
848
minerals based on mineral–fluid exchange thermometers of Friedman and O'Neil (1977) for
849
calcite (A) and Horitia (2014) for dolomite (B). The boxed fields represent formation
850
temperature estimates, based largely on fluid inclusion data discussed in Section 5.1, plotted
851
against δ18O values for each calcite/dolomite type. Formation temperatures of limestone matrix
852
samples were estimated at 20–30 °C based on Shields et al. (2003). Cambrian–Shadow Lake
853
calcite samples are plotted at 70–72 °C and are based on Ttrap values for stage II fluid inclusions
854
in these units (see Section 5.2; Table 3S – Supporting Material). Calcite vein and vug samples
855
from the Gull River–Coboconk formations and the Trenton Group are plotted using Ttrap
856
temperatures of 54–61 °C and 65–78 °C, respectively. For saddle and vein dolomite, Ttrap
857
temperatures of 88–128 °C were used. No fluid inclusion data were obtained for dolostone and
858
replacement dolomite samples but fluid estimates at 54–78 °C (based on calcite veins/vugs) and
859
88–128 °C (based on saddle dolomite) are shown by dashed boxes. The O-isotope composition
860
of late-Ordovician seawater is indicated by the shaded bands between −2 and −4‰ (Shields et
861
al., 2003).
AC C
EP
TE D
M AN U
SC
847
862 863
References
864
Al, T.A., Clark, I.D., Kennell, L., Jensen, M., Raven, K.G., 2015. Geochemical evolution and
39
ACCEPTED MANUSCRIPT
865
residence time of porewater in low-permeability rocks of the Michigan Basin, Southwest
866
Ontario. Chem. Geol. 404, 1–17. doi:10.1016/j.chemgeo.2015.03.005
867
Allan, J.R., Wiggins, W.D., 1993. Dolomite reservoirs: geochemical techniques for evaluating origin and distribution. American Association of Petroleum Geologists; Continuing
869
Education Course Note Series, 36.
870
RI PT
868
Altmann, S., Tournassat, C., Goutelard, F., Parneix, J.-C., Gimmi, T., Maes, N., 2012. Diffusiondriven transport in clayrock formations. Appl. Geochem. 27, 463–478.
872
doi:10.1016/j.apgeochem.2011.09.015
875 876 877
M AN U
874
Armstrong, D.K. and Carter, T.R. 2010. The Subsurface Paleozoic Stratigraphy of Southern Ontario. Ontario Geological Survey, Special Volume 7
Arthur, M.A., Cole, D.R., 2014. Unconventional Hydrocarbon Resources: Prospects and Problems. ELEMENTS 10, 257–264. doi:10.2113/gselements.10.4.257 Ayalon, A., Longstaffe, F.J., 1988. Oxygen isotope studies of diagenesis and pore-water
TE D
873
SC
871
evolution in the Western Canada sedimentary basin; evidence from the Upper Cretaceous
879
basal Belly River Sandstone, Alberta. Journal of Sedimentary Research 58, 489–505.
880
doi:10.1306/212F8DCD-2B24-11D7-8648000102C1865D
EP
878
Beauheim, R.L., Roberts, R.M., Avis, J.D., 2014. Hydraulic testing of low-permeability Silurian
882
and Ordovician strata, Michigan Basin, southwestern Ontario. Journal of Hydrology 509,
883
163–178. doi:10.1016/j.jhydrol.2013.11.033
884 885
AC C
881
Benson, S.M., Cole, D.R., 2008. CO2 Sequestration in Deep Sedimentary Formations. ELEMENTS 4, 325–331. doi:10.2113/gselements.4.5.325
886
Bottomley, D.J., Clark, I.D., 2004. Potassium and boron co-depletion in Canadian Shield brines:
887
evidence for diagenetic interactions between marine brines and basin sediments. Chem.
40
ACCEPTED MANUSCRIPT
888
Geol. 203, 225–236. doi:10.1016/j.chemgeo.2003.10.010 Bottomley, D.J., Katz, A., Chan, L.H., Starinsky, A., Douglas, M., Clark, I.D., Raven, K.G.,
890
1999. The origin and evolution of Canadian Shield brines: evaporation or freezing of
891
seawater? New lithium isotope and geochemical evidence from the Slave craton. Chem.
892
Geol. 155, 295–320. doi:10.1016/S0009-2541(98)00166-1
RI PT
889
Clark, I.D., Al, T., Jensen, M., Kennell, L., Mazurek, M., Mohapatra, R., Raven, K.G., 2013.
894
Paleozoic-aged brine and authigenic helium preserved in an Ordovician shale aquiclude.
895
Geology 41, 951–954. doi:10.1130/G34372.1
SC
893
Clark, I.D., Ilin, D., Jackson, R.E., Jensen, M., Kennell, L., Mohammadzadeh, H., Poulain, A.,
897
Xing, Y.P., Raven, K.G., 2015. Paleozoic-aged microbial methane in an Ordovician shale
898
and carbonate aquiclude of the Michigan Basin, southwestern Ontario. Organic
899
Geochemistry 83-84, 118–126. doi:10.1016/j.orggeochem.2015.03.006 Clark, I.D., Mohapatra, R., Mohammadzadeh, H., Kotzer, T., 2010a. Porewater and Gas
TE D
900
M AN U
896
901
Analyses in DGR-1 and DGR-2 Core. NWMO Technical Report TR-07-21. Nuclear Waste
902
Management Organization (NWMO), Toronto, pp. 25. https://www.nwmo.ca/en/Reports Clark, I.D., Liu, I., Mohammadzadeh, H., Zhang, P., Wilk, M., 2010b. Porewater and Gas
EP
903
Analyses in DGR-3 and DGR-4 Core. NWMO Technical Report TR-08-19. Nuclear Waste
905
Management Organization (NWMO), Toronto, pp. 29. https://www.nwmo.ca/en/Reports
906
Coniglio, M., Williams-Jones, A.E., 1992. Diagenesis of Ordovician Carbonates From the North-
AC C
904
907
East Michigan Basin, Manitoulin Island Area, Ontario - Evidence From Petrography, Stable
908
Isotopes and Fluid Inclusions. Sedimentology 39, 813–836. doi:10.1111/j.1365-
909
3091.1992.tb02155.x
910
Coniglio, M., Sherlock, R., Williams Jones, A.E., Middleton, K., Frape, S.K., 1994. Burial and
41
ACCEPTED MANUSCRIPT
Hydrothermal Diagenesis of Ordovician Carbonates from the Michigan Basin, Ontario,
912
Canada, in: Purser, B., Tucker, M., Zenger, D. (Eds.), Dolomites a Volume in Honour of
913
Dolomieu. Special Publications of the International Association of Sedimentologists 21.
914
Blackwell Publishing Ltd., Oxford, UK, pp. 231–254. doi:10.1002/9781444304077.ch14
915
Coplen, T.B., Hopple, J.A., Boehike, J.K., Peiser, H.S., Rieder, S.E., Krouse, H.R., Rosman,
RI PT
911
K.J.R., Ding, T., Vocke, R.D., Jr., Révész, K.M., Lamberty, A., Taylor, P., De Bièvre, P.,
917
2002. Compilation of minimum and maximum isotope ratios of selected elements in
918
naturally occurring terrestrial materials and reagents. U.S. GEOLOGICAL SURVEY Water
919
Resources Investigation Report 01-4222.
M AN U
SC
916
920
Davies, G.R., Smith, L.B., Jr., 2006. Structurally controlled hydrothermal dolomite reservoir
921
facies: An overview. American Association of Petroleum Geologists Bulletin 90, 1641–
922
1690. doi:10.1306/05220605164
Davis, D.W., 2013. Application of U-Pb Geochronology Methods to the Absolute Age
TE D
923
Determination of Secondary Calcite. NWMO Technical Report TR-2013-21. Nuclear Waste
925
Management Organization (NWMO), Toronto, pp. 31. https://www.nwmo.ca/en/Reports
926
Davis, D.W., Sutcliffe, C.N., Smith, P., Zajacz, Z., Thibodeau, A.M., Spalding, J., Schneider, D.,
EP
924
Adams, J., Cruden, A. and Parmenter, A., 2017. Secondary calcite as a useful U-Pb
928
geochronometer: An example from the Paleozoic sedimentary sequence in southern Ontario.
929
GAC-MAC 2017 Joint Annual Meeting, Kingston, ON, Canada, May 14–18. Abstracts
930
Volume.
931
AC C
927
Diamond, L.W., 2003. Systematics of H2O inclusions. In: Samson, I.M., Anderson, A.J.,
932
Marshall, D.D. (Eds.), Short Course 32. Fluid Inclusions: Analysis and Interpretation. Short
933
Course 32. Mineralogical Association of Canada, pp. 55-79.
42
ACCEPTED MANUSCRIPT
934
Diamond, L.W., Aschwanden, L. and Caldas, R., 2015. Fluid inclusion study of core and outcrop samples from the Bruce Nuclear Site, Southern Ontario, Canada. NWMO Technical Report
936
TR-2015-24. Nuclear Waste Management Organization (NWMO), Toronto, pp. 138.
937
https://www.nwmo.ca/en/Reports
938
RI PT
935
Douglas, M., Clark, I.D., Raven, K., Bottomley, D., 2000. Groundwater mixing dynamics at a Canadian Shield mine. Journal of Hydrology 235, 88–103. doi:10.1016/S0022-
940
1694(00)00265-1
941
SC
939
Frape, S.K., Fritz, P., McNutt, R.H., 1984. Water-rock interaction and chemistry of groundwaters from the Canadian Shield. Geochim. Cosmochim. Acta 48, 1617–1627. doi:10.1016/0016-
943
7037(84)90331-4
945 946 947
Friedman, I., O'Neil, J.R., 1977. Compilation of stable isotope fractionation factors of geochemical interest. USGS Professional Paper 440-KK.
Fritz, P., Frape, S.K., 1982. Saline groundwaters in the Canadian Shield — A first overview.
TE D
944
M AN U
942
Chem. Geol. 36, 179–190. doi:10.1016/0009-2541(82)90045-6 Gascoyne, M., 2004. Hydrogeochemistry, groundwater ages and sources of salts in a granitic
949
batholith on the Canadian Shield, southeastern Manitoba. Appl. Geochem. 19, 519–560.
950
doi:10.1016/S0883-2927(03)00155-0 Girard, J.-P., Barnes, D.A., 1995. Illitization and Paleothermal Regimes in the Middle
AC C
951
EP
948
952
Ordovician St. Peter Sandstone, Central Michigan Basin: K-Ar, Oxygen Isotope, and Fluid
953
Inclusion Data. American Association of Petroleum Geologists Bulletin 79, 49–69.
954 955 956
Goldstein, R.H., 2001. Fluid inclusions in sedimentary and diagenetic systems. Lithos 55, 159– 193. doi:10.1016/S0024-4937(00)00044-X Goldstein, R.H., Reynolds, T.J., 1994. Systematics of fluid inclusions in diagenetic minerals.
43
ACCEPTED MANUSCRIPT
957
SEPM (Society for Sedimentary Geology) Short Course 31. Tulsa, U.S.A. Haeri-Ardakani, O., Al-Aasm, I., Coniglio, M., 2013. Fracture mineralization and fluid flow
959
evolution: an example from Ordovician-Devonian carbonates, southwestern Ontario,
960
Canada. Geofluids 13, 1–20. doi:10.1111/gfl.12003
RI PT
958
961
Hanor, J.S., 2001. Reactive transport involving rock-buffered fluids of varying salinity.
962
Geochim. Cosmochim. Acta 65, 3721–3732. doi:10.1016/S0016-7037(01)00703-7
Harper, D.A., Longstaffe, F.J., Wadleigh, M.A., McNutt, R.H., 1995. Secondary K-feldspar at
964
the Precambrian–Paleozoic unconformity, southwestern Ontario. Can. J. Earth Sci. 32,
965
1432–1450. doi:10.1139/e95-116
M AN U
966
SC
963
Hendry, M.J., Solomon, D.K., Person, M., Wassenaar, L.I., Gardner, W.P., Clark, I.D., Mayer, K.U., Kunimaru, T., Nakata, K., Hasegawa, T., 2015. Can argillaceous formations isolate
968
nuclear waste? Insights from isotopic, noble gas, and geochemical profiles. Geofluids 15,
969
381–386. doi:10.1111/gfl.12132
TE D
967
970
Holser, W.T., 1979. Mineralogy of Evaporites, in: Burns, R.G. (Ed.), Marine Minerals. Reviews
971
in Mineralogy and Geochemistry, Vol. 6. Mineralogical Society of America. pp. 295–346. Horita, J., 2014. Oxygen and carbon isotope fractionation in the system dolomite-water-CO2 to
EP
972
elevated temperatures. Geochim. Cosmochim. Acta 129, 111–124.
974
doi:10.1016/j.gca.2013.12.027
975 976 977
AC C
973
Hurley, N.F., Budros, R., 1990. Albion-Scipio and Stoney Point Fields-U.S.A., Michigan Basin. American Association of Petroleum Geologists Special Volumes 23, 1–37. Intera, 2011. Descriptive Geosphere Site Model. NWMO Technical Report TR-2011-24. Nuclear
978
Waste Management Organization (NWMO), Toronto, pp. 426.
979
(http://www.nwmo.ca/dgrgeoscientificsitecharacterization).
44
ACCEPTED MANUSCRIPT
980 981
Khader, O., Novakowski, K., 2014. The study of diffusion dominant solute transport in solid host rock for nuclear waste disposal. Report to: Canadian Nuclear Safety Commission. Knauth, L.P., Beeunas, M.A., 1986. Isotope geochemistry of fluid inclusions in Permian halite
983
with implications for the isotopic history of ocean water and the origin of saline formation
984
waters. Geochim. Cosmochim. Acta 50, 419–433. doi:10.1016/0016-7037(86)90195-X
985
Kurtis Kyser, T., Kerrich, R., 1990. Geochemistry of fluids in tectonically active crustal regions.
RI PT
982
Short Course on Fluids in Tectonically Active Regimes of the Continental Crust,
987
Mineralogical Association of Canada, Short Course Handbook 18, 133–230. Legall, F.D., Barnes, C.R., Macqueen, R.W., 1981. Thermal Maturation, Burial History and
M AN U
988
SC
986
989
Hotspot Development, Paleozoic Strata of Southern Ontario - Quebec, from Conodont and
990
Acritarch Colour Alteration Studies. Bulletin of Canadian Petroleum Geology 29, 492–539. Ma, L., Castro, M.C., Hall, C.M., 2009. Atmospheric noble gas signatures in deep Michigan
992
Basin brines as indicators of a past thermal event. Earth Planet. Sci. Lett. 277, 137–147.
993
doi:10.1016/j.epsl.2008.10.015
994
TE D
991
Mazurek, M., de Haller, A., 2017. Pore-water evolution and solute-transport mechanisms in Opalinus Clay at Mont Terri and Mont Russelin (Canton Jura, Switzerland). Swiss J Geosci
996
110, 129–149. doi:10.1007/s00015-016-0249-9 McNutt, R.H., Frape, S.K., Fritz, P., 1984. Strontium isotopic composition of some brines from
AC C
997
EP
995
998
the Precambrian Shield of Canada. Chem. Geol. 46, 205–215. doi:10.1016/0009-
999
2541(84)90190-6
1000
Middleton, K., Coniglio, M., Frape, S.K., 1993. Dolomitization of Middle Ordovician Carbonate
1001
Reservoirs, Southwestern Ontario. Bulletin of Canadian Petroleum Geology 41, 150–163.
1002
Neuzil, C. E., and A. M. Provost. 2014. Ice sheet load cycling and fluid underpressures in the
45
ACCEPTED MANUSCRIPT
1003
Eastern Michigan Basin, Ontario, Canada, J. Geophys. Res. Solid Earth, 119, 8748–8769,
1004
doi:10.1002/ 2014JB011643.
1006 1007
Neuzil, C.E., 2015. Interpreting fluid pressure anomalies in shallow intraplate argillaceous formations. Geophys. Res. Lett. 42, 4801–4808. doi:10.1002/2015GL064140
RI PT
1005
Normani, S.D., Sykes, J.F., 2012. Paleohydrogeologic simulations of Laurentide ice‐sheet history on groundwater at the eastern flank of the Michigan Basin. Geofluids 12, 97–122.
1009
doi:10.1111/j.1468-8123.2012.00362.x
1011 1012
NWMO, 2011. Geosynthesis. NWMO Technical Report TR-2011-11. Nuclear Waste Management Organization (NWMO), Toronto, pp. 418. (https://www.nwmo.ca/en/Reports).
M AN U
1010
SC
1008
Oakes, C.S., Bodnar, R.J., Simonson, J.M., 1990. The system NaCl-CaCl2-H2O: I. The ice
1013
liquidus at 1 atm total pressure. Geochim. Cosmochim. Acta 54, 603–610.
1014
doi:10.1016/0016-7037(90)90356-P
Obermajer, M., Fowler, M.G., Snowdon, L.R., 1999. Depositional environment and oil
TE D
1015
generation in Ordovician source rocks from southwestern Ontario, Canada: Organic
1017
geochemical and petrological approach. American Association of Petroleum Geologists
1018
Bulletin 83, 1426–1453.
1020 1021 1022 1023 1024 1025
Ontario Geological Survey, 1991. Bedrock Geology of Ontario, Southern Sheet. Ontario Geological Survey, Map 2544, Scale 1:1,000,000.
AC C
1019
EP
1016
Pearson, F.J., Jr, 1987. Models of mineral controls on the composition of saline groundwaters of the Canadian Shield. Geological Association of Canada - Special Paper 33, 39–51. Powell, T.G., Macqueen, R.W., Barker, J.F., Bree, D.G., 1984. Geochemical character and origin of Ontario oils. Bulletin of Canadian Petroleum Geology 32, 289–312. Roedder, E., 1984. Fluid Inclusions, in: Ribbe, P.E. (Ed.), Reviews in Mineralogy and
46
ACCEPTED MANUSCRIPT
1026
Geochemistry, Vol. 12. Mineralogical Society of America. Reviews in Mineralogy and
1027
Geochemistry, Mineral Soc Am.
1028
Roedder, E., 1979. Fluid inclusion evidence on the environments of sedimentary diagenesis, a review. The Society of Economic Paleontologists and Mineralogists, Special Publication 89–
1030
107.
1033 1034 1035 1036
Annu. Rev. Earth Planet. Sci. 8, 263–301.
SC
1032
Roedder, E., Bodnar, R.J., 1980. Geologic Pressure Determinations from Fluid Inclusion Studies.
Russell, D.J., Gale, J.E., 1982. Radioactive waste disposal in the sedimentary rocks of southern Ontario. Geoscience Canada 9, 200–207.
M AN U
1031
RI PT
1029
Samson, I.M., Anderson, A.J., Marshall, D.D., 2003. Fluid inclusions: analysis and interpretation. Mineralogical Association of Canada, Short Course 32. Saso, J., 2013. Paleohydrogeological Investigation of Upper Ordovician and Cambrian Rocks
1038
from the Michigan Basin in Southwestern Ontario. University of New Brunswick. M.Sc.
1039
Thesis.
1040
TE D
1037
Shafeen, A., Croiset, E., Douglas, P. and Chatzis, I. 2004. CO2 sequestration in Ontario, Canada. Part I: Storage evaluation of potential reservoirs. Energy Conversion and Management 45,
1042
2645–2659.
EP
1041
Shields, G.A., Carden, G.A.F., Veizer, J., Meidla, T., Rong, J.-Y., Li, R.-Y., 2003. Sr, C, and O
1044
isotope geochemistry of Ordovician brachiopods: a major isotopic event around the Middle-
1045
Late Ordovician transition. Geochim. Cosmochim. Acta 67, 2005–2025. doi:10.1016/S0016-
1046
7037(02)01116-X
1047 1048
AC C
1043
Spencer, R.J., 1987. Origin of CaCl brines in Devonian formations, western Canada sedimentary basin. Appl. Geochem. 2, 373–384. doi:10.1016/0883-2927(87)90022-9
47
ACCEPTED MANUSCRIPT
1049
Steele-MacInnis, M., Bodnar, R.J., Naden, J., 2011. Numerical model to determine the composition of H2O–NaCl–CaCl2 fluid inclusions based on microthermometric and
1051
microanalytical data. Geochim. Cosmochim. Acta 75, 21–40. doi:10.1016/j.gca.2010.10.002
1052
Taylor, T.R., Sibley, D.F., 1986. Petrographic and geochemical characteristics of dolomite types
RI PT
1050
1053
and the origin of ferroan dolomite in the Trenton Formation, Ordovician, Michigan Basin,
1054
U.S.A. Sedimentology 33, 61–86. doi:10.1111/j.1365-3091.1986.tb00745.x
Veizer, J., Ala, D., Azmy, K., Bruckschen, P., Buhl, D., 1999. 87Sr/86Sr, δ13C and δ18O
SC
1055
evolution of Phanerozoic seawater. Chem. Geol. doi:10.1016/S0009-2541(99)00081-9
1057
Wickstrom, L.H., Gray, J.D., Stieglitz, R.D., 1992. Stratigraphy, structure, and production
M AN U
1056
1058
history of the Trenton Limestone (Ordovician) and adjacent strata in northwestern Ohio.
1059
Rep. of investigations/State of Ohio. Dep. of natural resources, Div. of geol. survey.
1060
doi:10.1111/j.1502-3931.1996.tb01833.x/full
Wilson, T.P., Long, D.T., 1993a. Geochemistry and isotope chemistry of CaNaCl brines in
TE D
1061 1062
Silurian strata, Michigan Basin, U.S.A. Appl. Geochem. 8, 507–524. doi:10.1016/0883-
1063
2927(93)90079-V
Wilson, T.P., Long, D.T., 1993b. Geochemistry and isotope chemistry of Michigan Basin brines:
1065
Devonian formations. Appl. Geochem. 8, 81–100. doi:10.1016/0883-2927(93)90058-O Ziegler, K., Longstaffe, F.J., 2000. Multiple Episodes of Clay Alteration at the
AC C
1066
EP
1064
1067
Precambrian/Paleozoic Unconformity, Appalachian Basin: Isotopic Evidence for Long-
1068
Distance and Local Fluid Migrations. Clays and Clay Minerals 48, 474–493.
1069
doi:10.1021/ac60068a025
1070
48
ACCEPTED MANUSCRIPT
Table 1. Microthermometric results for fluid inclusions in the secondary minerals Sample ID
Formation
mBGS
Matrix
T m(ice) or T m(hydrohalite)d (ºC)
T trap (ºC)
Mineral I
II
IVb
III
V
VI
I
[77–96], [106–127]
<70
[57–77] [80–100] [61–78]
<70 <70 <70
48
II III IV min max min max min max –46.9 –40.9 –45.2 –44.7 –21.5 –18.2 –36.4 –33.2 –40.7 –39.0 –45.0 –34.0 –47.0 –42.0 –44.2 –35.9 –46.5 –35.2 –46.8 –42.1
[86–107] [94–113]
<70
59
–22.9 –18.4 –43.1 –40.2
Sherman Fall Coboconk Coboconk Coboconk Coboconk Coboconk Coboconk Gull River Gull River
691.70 764.65 767.81 769.57 778.29 778.44 799.15 818.36 829.53
cal-py dol dol cal-py dol
cal cal cal cal-cls cal
dol-py cal dol-py
dol (SD) cal cal
122a
dol-py
dol (SD)
125, 128a
65, 78 61 57 54
71
42
SC
DGR-2-691.70 DGR-6-883.75 DGR-6-886.91 DGR-4-769.57 DGR-6-892.99 DGR-2-778.44 DGR-3-799.15 DGR-4-818.36 DGR-4-829.53
max
RI PT
min
Vc min
max
–38.8 –44.1
–44.1 –42.7
M AN U
DGR-2-844.31 Cambrian 844.31 55 <70 dol-py dol-cal [75–96], [77–97] –42.7 –41.5 –40.3 –37.4 –43.9 –42.6 –35.0 –34.4 88a qtz cal [91–112], [94–115] –42.9 –41.8 –52.1 –50.5 DGR-2-847.42 Cambrian 847.42 80 qtz cal [88–109], [91–112] DGR-2-850.01 Cambrian 850.01 72 <70 –42.6 –41.5 –52.1 –49.5 –34.2 –32.1 qtz cal –47.9 –42.6 DGR-2-853.72 Cambrian 853.72 70 [102–123], [104–125] <70 –52.7 –50.3 –25.9 –24.3 Notes: Microthermometric measurements were made in calcite and dolomite (a), and are indicated by the Mineral column. Relative timing of fluid stages I–VI was determined petrographically based cross-cutting relationships between fluid inclusion assemblages. (b) T trap for stage IV can only be bracketed between T h (minimum possible value) and a maximum T calculated from an assumed lithostatic pressure of 1860 m (see text for explanation).
AC C
EP
TE D
T m(ice) values in stages I–IV represent the metastable equilibrium between halite+ice+liquid+vapour. (c) T m(ice) for stage V was obtained by artificially stretching the inclusions. (d) Hydrohalite T m values are shown in italics. Abbreviations: cal – calcite; cls – celestine; dol – dolomite; py – pyrite; SD – saddle dolomite; qtz – quartz.
ACCEPTED MANUSCRIPT
Stratigraphy Pleistocene Lucas Amherstburg
Karstic Aquifer
Bois Blanc
RI PT
100
Devonian
0
Hydrostratigraphy
Bass Islands
Silurian
Silurian Aquitard A1 Aquifer
Guelph Goat Island Gasport/Lions Head/ Fossil Hill Cabot Head Manitoulin Queenston
400
500
Georgian Bay
600 Blue Mountain Collingwood Cobourg Sherman Fall Kirkfield Coboconk
800
Gull River Shadow Lake Cambrian
900
Precambrian
Trenton Group
700
Guelph Aquifer
Ordovician Aquiclude
Black River Group
Depth (mBGS)
Salina
300
Ordovician
AC C
EP
TE D
M AN U
SC
200
Black River Group Aquitard Basal Aquifer
150
Temperature (°C)
125 100 75 50
–2
0
+4
+8
–4 –6
Trenton Group
ACCEPTED MANUSCRIPT
Cambrian–Shadow Lake
–10
–14
RI PT
A
Gull River– Coboconk
SC
25
M AN U
All Limestones
0 10.0
14.0
18.0
22.0
26.0
30.0
B
150
–4
–2
TE D
δ 18OVSMOW(‰) of calcite 0
+4
+8
100
–6
AC C
Temperature (°C)
125
EP
Saddle Dolomite
–10
Matrix Replacement and Dolostone
75 –14
50 25 0 10.0
14.0
18.0
22.0
26.0
δ 18OVSMOW(‰) of dolomite
30.0
A
ACCEPTED MANUSCRIPT
B
C
Dol
RDol Anh
Anh
Cal vein
RI PT
Dol vein SDol
Cal vein
Anh
2 mm
E
SDol
Anh
SDol
1 mm
Py
Organics+Mag
Cal
Dol vein
Cal Dol 500 µm
EP
SDol
AC C
H
Dol
Qtz
F
Dol
1 mm
G
TE D
Dol
4-829.53
SDol
reimage SDol
SC
D
2 mm
M AN U
2 mm
Dol
Cal vug
I
SDol
Cal vug
SDol SDol
Cal vein Dol
Dol 500 µm
SDol
1 mm
1 mm
Qtz
Fluid Stages
I
II
III
IV
ACCEPTED MANUSCRIPT
V
VI
Secondary Minerals Dolomite - saddle and veins
RI PT
Authigenic Quartz Calcite
SC
- veins and vugs
Anhydrite
M AN U
Celestine Pyrite Magnetite
EP
Fluids
TE D
Bitumen
AC C
Halite sat. brine I + CH4
Halite sat. brine II + CH4 Halite sat. brine III + CH4 Halite sat. brine IV Brine V Oil + CH4 early
late
650
A
B
C
ACCEPTED MANUSCRIPT
650
650
Calcite Veins and Vugs
RI PT
700 700
Dolomite Dolostone Matrix Matrix Replacement
M AN U
Anhydrite in Veins and Vugs
750
TE D
Depth (mBGS)
SC
Saddle and Veins
AC C
EP
800 800
850 0
20
40
60
80
Dolomite wt.% (Relative to Calcite)
100
0
4
8
12
MgO (wt.%)
16
20
24
0
1000
2000
3000
Sr (ppm)
4000
5000
ACCEPTED MANUSCRIPT
B
A
C
SDol
B
SDol
RI PT
C
Stage I
SC
D
M AN U
E
h) H
Anh
EP
F
Cal
Stage II
Stage II
H
I
Anh
Stage III
Stage I
Cal
AC C
G
TE D
Cal
Stage II
SDol
Stage I
Anh
Stage III
Stage III
A
ACCEPTED MANUSCRIPT
B
C
Cal
RI PT
B
Cal Stage IV
SC
E
Cal
M AN U
D
Cal Stage IV
Stage IV
F
Cal
E
Cal G
H
Cal
Cal
I
AC C
G
Cal Cal
Stage V
Stage V
Stage IV
EP
Stage IV
TE D
Uniform vapour fractions
ACCEPTED MANUSCRIPT Primary inclusions
650
Stage I (dolomite) Stage II (calcite) Stage III (cls+anh)
Stage IV Stage V Stage VI (oil+gas)
M AN U AC C
800 800
TE D
750
EP
Depth (mBGS)
SC
RI PT
700 700
Secondary inclusions
850 25
50
75
100
Temperature (Ttrap, °C)
125
A Ca Cl 2
H 2O
Ma ss%
10
10 Ice + L + V
te ) i s
Metastable cotectic Ice + Halite + L + V 30
otherms
SC
h a li T m(
Hydrohalite + L + V
RI PT
therms
°C 42 °C 70 °C 6 10 8 °C 12
40
Halite + L + V
M AN U
40
20
Tm(ice) iso
30 Eutectic -52 °C Peritectic -22.4 °C
Cl Na
20
-24 °C -26 °C -32 °C -35 °C
% ss Ma
ACCEPTED MANUSCRIPT
Antarcticite + L + V
CaCl2
TE D
B
NaCl
Ca
ss%
Ma
Ice + L + V
20
Cl
Na
AC C
10
%
10
ss
Ma
EP
Cl
2
H 2O
20
Stage I Stage II Stage III Stage IV Stage V
? ?
30
40
CaCl2
30
Halite + L + V
40
NaCl
650
A
B
C
ACCEPTED MANUSCRIPT
650
Calcite Limestone Matrix Veins and Vugs Dolomite Dolostone Matrix Matrix Replacement
RI PT
700
Saddle and Veins Anhydrite and Celestine in Veins and Vugs
M AN U
Late Ordovician Seawater
750
TE D
Depth (mBGS)
SC
700
AC C
EP
800 800
Precambrian Shield Brines
Late Ordovician Seawater
850
Cambrian Seawater
Cambrian Seawater
-8.0
-6.0
-4.0
-2.0
0.0
δ 13CVPDB(‰)
2.0
4.0
10.0
14.0
18.0
22.0
δ 18OVSMOW(‰)
26.0
30.0 0.707
0.708
0.709 87
Sr/ 86 Sr
0.710
0.711
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Highlights 1. Multi-stage paleofluid history recorded by primary and secondary fluid inclusions 2. Primary stages I to III formed the dolomite and calcite veins 3. Secondary fluid stages IV to VI record trapping of local and re-mobilized fluids 4. Mineral isotope data indicate variable fluid sources for the veins 5. Sources include connate fluids and mixed connate–shield-type brines
S0883-2927(17)30282-2
DOI:
10.1016/j.apgeochem.2017.09.011
Reference:
AG 3952
To appear in:
Applied Geochemistry
Received Date: 7 July 2017 Revised Date:
14 September 2017
Accepted Date: 16 September 2017
Please cite this article as: Petts, D.C., Saso, J.K., Diamond, L.W., Aschwanden, L., Al, T.A., Jensen, M., The source and evolution of paleofluids responsible for secondary minerals in lowpermeability Ordovician limestones of the Michigan Basin, Applied Geochemistry (2017), doi: 10.1016/ j.apgeochem.2017.09.011. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Title: The source and evolution of paleofluids responsible for secondary minerals in low-
RI PT
permeability Ordovician limestones of the Michigan Basin
Authors:
a
M AN U
SC
Petts, D.C.a, *, Saso, J.K.b, Diamond, L.W.c, Aschwanden, L.c, Al, T.A.a and Jensen, M.d
Department of Earth and Environmental Sciences, University of Ottawa, Ottawa, ON, K1N
6N5, Canada b
Department of Earth Sciences, University of New Brunswick, Fredericton, New Brunswick,
c
TE D
Canada
Rock-Water Interaction Group, Institute of Geological Sciences, University of Bern,
Baltzerstrasse 3, 3012 Bern, Switzerland
Nuclear Waste Management Organization, Toronto, ON, M4T 2S3, Canada
EP
d
AC C
* Corresponding author: [email protected]
Version: September 19, 2017
1
ACCEPTED MANUSCRIPT
1 2
Abstract In this study we report on the source and evolution of fluids associated with secondary vein and replacement minerals in low-permeability carbonate units in the Michigan Basin. This
4
petrogenetic information was collected using thermometric data from fluid inclusions combined
5
with C-, O- and Sr-isotope data, and focuses on mm- or cm-wide vein and vug minerals from
6
Ordovician limestones of the Trenton and Black River groups and Cambrian sandstones in SW
7
Ontario, Canada.
SC
9
Primary fluid inclusions in dolomite represent fluid stage I and have the highest trapping temperatures (Ttrap) in the sedimentary succession, between 88 and 128 °C. Primary inclusions in
M AN U
8
RI PT
3
calcite (stage II), and celestine and anhydrite (stage III), represent the final stages of secondary
11
mineral formation and have Ttrap values of 54–78 °C. All three stages of vein minerals formed
12
from brines with salinities of 31 to 37 wt.% [CaCl2+NaCl]eq that were saturated in halite and
13
methane gas at the time of mineral growth. Three subsequent stages of secondary fluid
14
inclusions were observed in the samples (stages IV–VI), however, no secondary vein minerals
15
formed during these stages and they are interpreted as re-mobilization and/or minor fluid ingress
16
along grain boundaries and micro-fractures. Notably, stage IV secondary inclusions are gas-
17
undersaturated with salinities of 32 to 34 wt.% [CaCl2+NaCl]eq and minimum Ttrap values of 57–
18
106 °C, and are interpreted to have formed during a Late Devonian–Mississippian regional
19
heating event.
EP
AC C
20
TE D
10
Secondary minerals from the Cambrian units and the Shadow Lake Formation have δ13C
21
values of −6.1 to −2.5‰ (VPDB), δ18O values of +14.6 to +24.2‰(VSMOW) and 87Sr/86Sr of
22
0.70975 to 0.71043. Relative shifts of approximately +2 to +3‰ in δ13C, >+4‰ in δ18O and
23
−0.002 in 87Sr/86Sr are observed upward across the boundary between the Cambrian–Shadow
2
ACCEPTED MANUSCRIPT
Lake units and the overlying Gull River Formation. Samples from the Gull River and Coboconk
25
formations and the Trenton Group (Kirkfield, Sherman Fall and Cobourg formations) have δ13C
26
values of −1.0 to +1.9‰, δ18O values of +18.9 to +28.1‰ and 87Sr/86Sr values of 0.70790 to
27
0.70990.
RI PT
24
The combined microthermometric and isotopic data for secondary minerals in the
29
Cambrian–Shadow Lake units suggest they formed from hydrothermal brine with a geochemical
30
signature obtained by interaction with the underlying Precambrian shield, or shield-derived
31
minerals in the Cambrian sandstones. Previous U-Pb dating of vein calcite and Rb-Sr dating of
32
secondary K-feldspar from the region suggests that brine ingress occurred during the Late
33
Ordovician–Silurian. The O- and Sr-isotope variability in vein samples from the Gull River and
34
Coboconk formations is interpreted as localized mixing of 18O-enriched, connate fluids with
35
hydrothermal brine. In comparison, isotopic data for the Trenton Group indicate secondary
36
mineral formation from connate fluids, sourced from 18O-enriched, evolved seawater and/or
37
modified hydrothermal brine that experienced fluid–rock interaction during transit through the
38
underlying stratigraphy.
M AN U
TE D
EP
39
SC
28
Keywords: Michigan Basin; low-permeability rocks; secondary minerals; fluid–rock interaction;
41
fluid inclusions; strontium isotopes; stable isotopes
AC C
40
3
ACCEPTED MANUSCRIPT
42 43
1. Introduction Low-permeability sedimentary rocks have increasingly become a global focus for a wide range of environmental and geoscience studies, including nuclear waste management (Russell
45
and Gale, 1982; Hendry et al., 2015), CO2 sequestration (Shafeen et al., 2004; Benson and Cole,
46
2008) and extraction of gas from shale (Arthur and Cole, 2014). In the cases of CO2
47
sequestration and extraction of gas with hydraulic fracturing, the low-permeability rocks play a
48
critical role in limiting the upward migration of fluids that are injected or displaced. Low-
49
permeability sedimentary rocks have been of particular interest in countries such as Canada,
50
Switzerland, France, Belgium and Germany for the long-term storage of nuclear waste because
51
solute transport is dominated by diffusion and the clay-rich mineralogy provides large surface
52
areas for solute retention by sorption mechanisms. Internationally, detailed paleofluid studies
53
have provided valuable insight into fluid sources and solute transport mechanisms (Altmann et
54
al., 2012; Clark et al., 2013; Al et al., 2015; Mazurek and de Haller, 2017) that are essential for
55
evaluating the long-term viability of a site for nuclear waste storage.
SC
M AN U
TE D
56
RI PT
44
Low-permeability argillaceous limestones of the Trenton Group (10-16 to 10-12 m s-1; Beauheim et al., 2014), in the Bruce region of the Michigan Basin (southwestern Ontario,
58
Canada) have been identified as a potential host rock for the construction of a deep geologic
59
repository (DGR) for low- and intermediate-level nuclear waste. The Cobourg Formation, a
60
low-permeability, argillaceous limestone at the top of the Trenton Group occurs >650 m beneath
61
the Bruce nuclear site and has been proposed as the repository host rock. These limestones are
62
overlain by up to 200 m of regionally-extensive, low-permeability Ordovician shale (Intera,
63
2011; Beauheim et al., 2014). Detailed studies on the long-term physical and chemical stability
64
of the surrounding rocks have focused on a wide range of stratigraphic, structural, geophysical
AC C
EP
57
4
ACCEPTED MANUSCRIPT
and geochemical methods (Intera, 2011 and references therein). Previous investigations at the
66
Bruce site have contributed valuable information on the residence time and transport
67
mechanisms of fluids in these rocks, which include studies of porewater geochemistry and solute
68
transport (Clark et al., 2013; Al et al., 2015), hydraulic conductivity (Beauheim et al., 2014), and
69
mechanisms responsible for anomalous hydraulic-head profiles observed in the Trenton and
70
Black River group limestones (Normani and Sykes, 2012; Khader and Novakowski, 2014;
71
Neuzil and Provost, 2014; Neuzil, 2015).
SC
72
RI PT
65
Secondary minerals from veins in sedimentary rocks provide direct information on the movement of deep basin fluids. Detailed isotopic and petrographic information of secondary
74
minerals combined with fluid inclusion temperature constraints have been utilized in paleofluid
75
studies to differentiate between fluid sources that relate to either diagenesis/burial (Ayalon and
76
Longstaffe, 1988), or the influx of fluids from allochthonous sources (Spencer, 1980; Davies and
77
Smith, 2006). Detailed petrography (e.g., Taylor and Sibley, 1986) combined with thermometric
78
data from fluid inclusions (e.g., Roedder, 1979; Goldstein 2001) provide an understanding of the
79
relative timing of fluid infiltration and temperature constraints during crystallization. Here, we
80
report on the source and evolution of fluids associated with secondary mineral formation in low-
81
permeability Ordovician limestones and Cambrian siliciclastic units in the Michigan Basin from
82
southwestern Ontario, Canada.
TE D
EP
AC C
83
M AN U
73
Ordovician limestones of the Trenton and Black River groups, which are laterally
84
traceable across much of the Michigan Basin and occur in sections of the Appalachian Basin,
85
have long been a focus of paleofluid studies due to their association with hydrocarbon reservoirs
86
in the Albion-Scipio/Stoney Point fields in south-central Michigan (e.g., Hurley and Budros,
87
1990; Allan and Wiggins, 1993), the Lima-Indiana field in Ohio/Indiana (e.g., Wickstrom et al.,
5
ACCEPTED MANUSCRIPT
1992) and fields of southern Ontario (Powell et al., 1984; Middleton et al., 1993; Coniglio et al.,
89
1994; Obermajer et al., 1999). In these fields, hydrocarbon migration/reservoir formation is
90
closely tied to the formation of hydrothermal dolomite (Allan and Wiggins, 1993), which has
91
been suggested to form during high-temperature, fluid–rock interaction between the Ordovician
92
limestones and upwelling hydrothermal fluids (Davies and Smith, 2006). Furthermore, C-, O-
93
and Sr-isotope compositions of secondary minerals in southern Ontario (Harper et al., 1995;
94
Zeigler and Longstaffe, 2000; Haeri-Ardakani et al., 2013) and central Michigan (Allan and
95
Wiggins, 1993) have provided valuable information on the source of deep-seated fluid
96
migrations in these regions of the basin. In the Bruce region, however, detailed paleofluid
97
studies have largely focused on the geochemical and isotopic evolution of porewaters in the
98
Michigan Basin sedimentary succession (Clark et al., 2010a; 2010b; 2013; 2015; Al et al., 2015).
99
Knowledge gaps remain in the understanding of paleofluid contributions to the formation of
M AN U
SC
RI PT
88
secondary vein minerals in the Ordovician limestones, and in particular, the source and transport
101
mechanisms that are responsible for vein-related fluid migrations in these low-permeability
102
sedimentary rocks.
The objective of this study is to enhance the understanding of fluid residence time and
EP
103
TE D
100
migration that has been developed from previous research at the Bruce site. Detailed
105
petrography and microthermometric data from fluid inclusions are combined with C-, O- and Sr-
106
isotope data to determine the source and relative timing of paleofluid migrations that resulted in
107
the formation of secondary minerals, including vein and vug infill and replacement phases.
108
Samples were obtained from drill cores that section the entire succession of Cambrian to
109
Devonian sedimentary rocks and the underlying Precambrian gneissic basement.
AC C
104
110
6
ACCEPTED MANUSCRIPT
111
2. Geology and Stratigraphy of the Michigan Basin The Michigan Basin represents a broadly circular, ~500 km diameter intracratonic
113
sedimentary basin located in central North America. The study area is located along the eastern
114
margin of the basin, at the Bruce nuclear site (Fig. 1). A detailed summary of the Paleozoic
115
bedrock units in southern Ontario is discussed in Armstrong and Carter (2010) (and references
116
therein), a brief description of these units is provided below. At the Bruce site, the sedimentary
117
succession reaches a maximum thickness of ~ 860 m. The base of this succession comprises ~17
118
m of Cambrian sandstone which unconformably overlies Precambrian gneissic basement. Upper
119
Ordovician rocks disconformably overlie the Cambrian sandstones and are subdivided into the
120
Black River and Trenton group limestones and a thick sequence of shale. The Black River
121
Group consists of a 4–5 m thick basal unit of siltstone and minor dolostone (Shadow Lake
122
Formation) that is overlain by ~ 75 m of variably bioturbated and fossiliferous lithographic
123
limestone (Gull River and Coboconk formations). Overlying the Black River Group are ~ 110 m
124
of mostly argillaceous limestones of the Trenton Group (subdivided into the Kirkfield, Sherman
125
Fall and Cobourg formations). The Cobourg Formation is further subdivided into lower and
126
upper members, which, respectively comprise ~ 28 m of argillaceous limestone, with minor
127
interbeds of calcareous shale, and ~ 7.5 m of calcareous shale. The Cobourg Formation is
128
capped by ~ 200 m of Ordovician shale which is subdivided into the Blue Mountain, Georgian
129
Bay and Queenston formations. Overall, the Ordovician limestone–shale succession has been
130
interpreted as initial deposition in supra-tidal/tidal flat to lagoonal marine or offshore shallow
131
and deep-shelf settings (Black River and Trenton groups), and open marine to shallow intertidal
132
or supratidal settings for deposition of the Blue Mountain–Queenston shales. Overlying the
133
Ordovician shales are 415 m of Silurian and Devonian interlayered dolostone, shale and
AC C
EP
TE D
M AN U
SC
RI PT
112
7
ACCEPTED MANUSCRIPT
134
evaporitic units, which represent deposition in a carbonate-shelf setting with periods of restricted
135
marine circulation.
136
Herein we refer to “Ordovician limestones”, “Trenton Group limestones”, “Black River Group limestones” and “Cambrian sandstones” as generalized nomenclature for each of the
138
respective units, regardless of the minor occurrence of units/formations of differing lithology
139
(e.g., Cambrian dolostone units, Shadow Lake siltstones/dolostones in the Black River Group,
140
Collingwood shale member at the top of the Cobourg Formation).
SC
141
RI PT
137
Hydrogeologic conditions at the Bruce site have been investigated by Intera (2011), Normani and Sykes (2012), Beauheim et al. (2014), Khader and Novakowski (2014), Neuzil and
143
Provost (2014) and Neuzil (2015). The objective of these studies was to understand the
144
hydraulic properties of the sedimentary rocks through modelling (Normani and Sykes, 2012), in
145
situ hydraulic conductivity measurements (Beauheim et al., 2014) and interpretation of observed
146
anomalies in hydraulic-head distributions (Khader and Novakowski, 2014; Neuzil and Provost,
147
2014; Neuzil, 2015). Overall, the hydraulic data indicate that the permeability of the Upper
148
Ordovician limestones and shales is very low, such that solute transport is limited to diffusion.
149
In situ measurements of horizontal hydraulic conductivity (Kh) by Beauheim et al. (2014)
150
indicate that the Ordovician shales and the argillaceous limestones of the Trenton Group
151
represent an aquiclude (Fig. 1), with Kh values ranging from 10-16 to 10-12 m s-1. Limestones of
152
the Black River Group have Kh values of 10-12 to 10-10 m s-1 and define an aquitard (Fig. 1). The
153
Cambrian sandstone unit has a Kh value of 10-6 m s-1 and represents a basal aquifer (Fig. 1).
TE D
EP
AC C
154
M AN U
142
155
3. Methods
156
3.1 Sampling and Petrography
8
ACCEPTED MANUSCRIPT
157
Samples were obtained from six drill cores (DGR-2 to DGR-6; DGR-8) at the Bruce site, and were initially collected on the basis of visible secondary minerals (calcite, dolomite,
159
anhydrite, celestine, pyrite, halite, etc.) in veins, vugs and replacement/alteration zones, largely
160
within the Trenton and Black River group limestones and the Cambrian sandstones. Sample
161
depths are reported here as metres below ground surface (mBGS) relative to drill hole DGR-2.
162
RI PT
158
Polished thin sections of the secondary minerals were prepared at the University of New Brunswick, and characterized using optical microscopy and electron-beam techniques
164
(Microscopy and Microanalysis Facility at the University of New Brunswick). Secondary
165
minerals were imaged with a JEOL 6400 scanning electron microscope (SEM) using a
166
backscattered electron detector (BSE) and identified using SEM-based energy-dispersive X-ray
167
spectrometry (EDS). Quantitative measurements of major and trace element concentrations (Ca,
168
Mg and Sr) were made using wavelength-dispersive X-ray spectrometry on a JEOL 733 Super-
169
probe electron probe microanalyzer (EPMA), operating at 15 kV with a beam current of 15-20
170
nA. Integration times for each element ranged from 30–80 s and the detection limits were
171
generally lower than 160 ppm for Ca and Mg, and 260 ppm for Sr.
174
M AN U
TE D
EP
173
3.2 Fluid Inclusion Microthermometry Fluid inclusion measurements were undertaken in the Institute of Geological Sciences at
AC C
172
SC
163
175
the University of Bern, and follow the general fluid inclusion methodology outlined by Roedder
176
(1984), Goldstein and Reynolds (1994) and Samson et al. (2003). Double-polished thick
177
sections (150 µm) were prepared for each sample to permit characterization by optical
178
microscopy using transmitted, plane-polarized light and reflected, ultraviolet (UV) light. UV
179
light microscopy was utilized to identify fluid inclusions that contain hydrocarbons.
9
ACCEPTED MANUSCRIPT
180
Microthermometric measurements were made using a Linkam MSD-600 heating-cooling stage mounted on an Olympus BX51 polarizing microscope. The stage was calibrated against
182
synthetic CO2–H2O inclusions and synthetic H2O inclusions with critical density, such that the
183
accuracy is ± 0.1 °C for measurements made below room temperature and ± 1.0 °C for
184
measurements above room temperature. For the microthermometric measurements of inclusions
185
in calcite, careful attention was paid to avoid spurious data and/or anomalously high
186
homogenization temperatures that resulted from inclusion “stretching” (i.e. deformation of the
187
inclusion walls during heating/pressurization; Roedder and Bodnar, 1980; Goldstein and
188
Reynolds, 1994). “Stretched” inclusions were identified by increases in the diameter of the
189
contained vapour bubbles after heating to homogenization and cooling down to room
190
temperature, as visible in photographs taken prior to and after heating.
SC
M AN U
191
RI PT
181
Identification of gas and mineral phases trapped within the inclusions was undertaken using laser Raman spectroscopy. Either red (632.8 nm) He-Ne or green (532.12 nm) Nd-YAG
193
lasers were focused onto the sample through an Olympus 100/0.95 UM PlanFI objective lens on
194
a Olympus BX41 microscope, and Raman spectra were collected for 10–20 s using a Horiba
195
Jobin-Yvon LabRam HR-800 spectrometer calibrated against the emission lines of a neon lamp.
198 199
EP
197
AC C
196
TE D
192
3.3 Stable Isotope Analysis
C- and O-isotope measurements were conducted in the G.G. Hatch Stable Isotope
200
Laboratory at the University of Ottawa using isotope ratio mass spectrometry (IRMS). Calcite
201
and dolomite samples were acquired from thin section chips using a Merchantek microdrill (300
202
µm drill bit). Drill locations were pre-selected in portions of veins, vugs and/or
10
ACCEPTED MANUSCRIPT
alteration/replacement zones using a combination of optical microscopy and SEM-BSE imaging
204
of the associated thin section. The powdered calcite and dolomite samples were weighed into
205
glass reaction vessels and flushed with helium. Orthophosphoric acid was introduced into the
206
vials and reacted for 24 hours at 25 °C for calcite and 50 °C for dolomite. The C- and O-isotope
207
ratios of the resulting CO2 gas were measured using a Thermo-Finnigan Delta XP and Gas
208
Bench II in continuous flow mode, and normalized against NBS-18, NBS-19 and LSVEC
209
(carbon only). C- and O-isotope compositions are reported in δ-notation relative to Vienna
210
Peedee belemnite (VPDB) for carbon and Vienna standard mean ocean water (VSMOW) for
211
oxygen, in permil (‰). Analytical uncertainties associated with the δ13C and δ18O data are ± 0.1
212
‰ (2σ).
M AN U
SC
RI PT
203
213
215
3.4 87Sr/86Sr Analysis
Sr-isotope ratios (87Sr/86Sr) for carbonate samples were obtained in the Isotope
TE D
214
Geochemistry and Geochronology Research Centre (IGGRC) at Carleton University using
217
thermal ionization mass spectrometry (TIMS). Calcite, dolomite, anhydrite and celestine
218
samples were collected by micro-drilling following the same methodology described in Section
219
3.3. Sr-concentrations were determined using mean EPMA analyses for the target mineral.
220
Samples were weighed out between 0.5 and 15 mg (depending on Sr-concentration) and
221
dissolved in 2.5 N nitric acid. Micro-chromatography columns were prepared using 100 µL of
222
Eichrom Sr Spec resin (5 to 100 µm) in a 14 mL borosilicate glass column. The resin was
223
washed with 1200 µL of ultrapure water and then conditioned with 400 µL of 7 N nitric acid.
224
The samples were added to the column, then washed using 7 N nitric acid. Sr ions were eluted
225
from the resin using 1600 µL of ultrapure water, and dried down on a hot plate. The dried Sr
AC C
EP
216
11
ACCEPTED MANUSCRIPT
226
samples were loaded onto Ta filaments using 0.3 N orthophosphoric acid and analyzed for
227
87
228
blanks for the procedure are < 250 picograms. The three-year average for analyses of NIST
229
SRM987 is 87Sr/86Sr = 0.710254 ± 0.000019. Analytical uncertainties associated with the Sr-
230
isotope data are reported at 2σ.
RI PT
Sr/86Sr using a Thermo-Finnigan Triton TIMS, at temperatures of 1300-1500 °C. Total Sr
231 4. Results
233
4.1 Paragenetic sequence and chemistry of secondary minerals
Secondary minerals in the Cambrian sandstones and the Black River and Trenton group
M AN U
234
SC
232
limestones occur as matrix replacement and/or veins and vugs (Fig. 2). The mineral assemblages
236
predominantly consist of calcite and dolomite, with varying proportions of anhydrite, celestine,
237
pyrite and quartz (Fig. 3). Varying amounts of clay minerals and halite also occur as secondary
238
minerals (Intera, 2011). Relative proportions of dolomite versus calcite (as wt.% dolomite) are
239
exhibited in Fig. 4A using bulk XRD data from Intera (2011). Elemental concentrations (Ca, Mg
240
and Sr) are shown in Table 1S (Supporting Material). Saddle dolomite is observed as linings in mm- or cm-wide veins and vugs within: (1)
EP
241
TE D
235
stratabound dolostone in the Sherman Fall, Coboconk, Gull River and Cambrian units, (2) Gull
243
River limestone and (3) Shadow Lake siltstone (e.g., Figs. 2C–E; 2G–H; Fig. 1S – Supporting
244
Material). Replacement dolomite occurs as fine-grained and dolomite-rich, mm- or cm-wide
245
zones in limestone within both the Trenton and Black River groups (e.g., Fig. 2B). Uniform
246
MgO abundances of 18−22 wt.% were obtained for saddle/vein dolomite and dolostone matrix
247
(Fig. 4B). Sr abundances are commonly less than 1000 ppm for all three types of dolomite (Fig.
AC C
242
12
ACCEPTED MANUSCRIPT
248
4C). Dolomitic growth zones in calcite crystals (MgO up to 22 wt.%) are identified in
249
Coboconk, Gull River and Shadow Lake veins and vugs (Fig. 4B). Secondary calcite commonly consists of medium- or coarse-grained, euhedral crystals
251
within mm- to cm-sized veins and vugs, and are the principal vein/vug forming mineral in the
252
Trenton and Black River groups and the Cambrian sandstones (Fig. 1S – Supporting Material).
253
Low MgO abundances (< 2 wt.%) are common for most secondary calcite. Sr abundances in
254
vein and vug calcite exhibit a decreasing trend with depth, from as high as 4300 ppm in the
255
Cobourg to <950 ppm in the Cambrian (Fig. 4C).
SC
Anhydrite and celestine are observed as late infill after saddle dolomite and calcite (Figs.
M AN U
256
RI PT
250
2A; 2D; 2E; Saso, 2013; Fig. 1S – Supporting Material). The abundance of Sr in anhydrite is
258
comparable to calcite in the Cobourg and Sherman Fall formations, but it is higher than calcite in
259
the Gull River Formation (Fig. 4C). Authigenic quartz was identified in one sandstone and one
260
dolostone in the Cambrian, as euhedral quartz rims mantling rounded, detrital quartz cores (Fig.
261
2I), and druse quartz lining the wall of a calcite vein (Fig. 2G). At both of these occurrences,
262
authigenic quartz appears to have formed either prior to or synchronous with calcite formation.
263
In DGR-4-844.78, the quartz lining appears to postdate the growth of saddle dolomite (Fig. 2G).
264
Pyrite is observed throughout the succession, commonly as an early matrix mineral in all
265
limestone units (except the Sherman Fall Formation) and stratabound dolostones in the
266
Coboconk, Gull River, and Cambrian units, and as a late vug mineral, post-dating calcite in
267
DGR-2-843.83 (Fig. 2F).
269
EP
AC C
268
TE D
257
4.2 Petrography and microthermometry of fluid inclusions
13
ACCEPTED MANUSCRIPT
270
Six paleofluid stages were identified from the primary inclusions in dolomite (stage I), primary inclusions in calcite (stage II), primary inclusions in celestine and anhydrite (stage III),
272
and secondary inclusions mostly in calcite (stage IV to VI), and are summarized in Figs. 3–8 and
273
Table 1. Photomicrographs of representative primary and secondary inclusions are shown in
274
Figures 5 and 6, respectively. All inclusion assemblages trapped during stages I to III contain
275
variable proportions of liquid and methane vapour (e.g., Figs. 5B, E, H), indicating
276
heterogeneous fluid entrapment. Therefore, trapping temperatures (Ttrap) for these stages are
277
defined by the minimum homogenization temperature (Th) observed in each inclusion
278
assemblage (Diamond et al., 2015). Uniform proportions of liquid and vapour were observed in
279
assemblages of stage IV inclusions (e.g., Figs. 6B–C, E), thus the maximum Th for each
280
assemblage represents a minimum constraint on Ttrap (i.e. Ttrap > Th,max). All assemblages in
281
stages I to IV contain variable proportions of halite, indicating they were saturated with respect
282
to halite at the time of trapping (principles of these interpretations are explained in Diamond,
283
2003). Stage V inclusions consist of saline brine without vapour bubbles at room temperature
284
(e.g., Fig. 6G). In this case, the only constraint on Ttrap is the empirical rule that they were
285
trapped below 70 °C (Goldstein and Reynolds, 1994; Diamond, 2003). The Ttrap results are
286
summarized in Fig. 7 and Table 1 (detailed Th data are shown in Table 2S – Supporting Material,
287
after Diamond et al., 2015). The relative timing of fluid inclusion entrapment associated with
288
stages IV vs. VI was determined by cross-cutting relationships as shown in Figures 6H and 6I.
SC
M AN U
TE D
EP
AC C
289
RI PT
271
Temperatures of final melting of ice, Tm(ice), in the presence of halite + liquid + vapour
290
were measured in assemblages of inclusions in stages I to IV. No observations of eutectic
291
melting could be made but the span of Tm(ice) values between –52.7 and –34°C suggests the
292
principal salts are NaCl and CaCl2. As commonly observed in CaCl2-rich inclusions, nominally
14
ACCEPTED MANUSCRIPT
stable hydrohalite nucleated only rarely upon cooling and so the measured Tm(ice) transitions
294
represent metastable equilibria. To enable salinity determinations from these data, the locus of
295
the metastable cotectic involving ice + halite + liquid + vapour was calculated for this study (red
296
curve Figs. 8A, B) from the stable phase relations for the H2O–NaCl–CaCl2 ternary system given
297
by Oakes et al. (1990) and Steele-MacInnis et al. (2011). Fluid compositions and salinities in
298
terms of the model ternary components were then obtained for each assemblage by constructing
299
a halite-melting path beginning at Tm(ice) on the cotectic and moving towards the NaCl apex of
300
the ternary. The point at which this path intersects Ttrap on the halite liquidus defines the ternary
301
composition and the salinity of the inclusions (i.e. green arrow in Fig. 8A). For the halite-
302
undersaturated inclusions of stage V, salinity could only be broadly constrained to lie along the
303
relevant isotherms of Tm (ice) (dark bands in Fig. 8B; question marks show range of uncertainty
304
in salinity). Table 1 lists the resulting ternary compositions and salinities. Microthermometric
305
data of all individual inclusions measured are given in Diamond et al. (2015).
SC
M AN U
TE D
306
RI PT
293
Primary, stage I inclusions in saddle dolomite (e.g., Figs. 5A–C) from the Coboconk and Gull River formations yielded the highest Ttrap values; 122 to 128 °C (n = 3) and a Ttrap value of
308
88 °C was obtained for primary inclusions in rhombic dolomite from the Cambrian. Stage I
309
inclusions have Tm (ice) values between −44.1 and −41.5 °C, implying that the fluid had a
310
salinity of 33 to 34 wt.% [CaCl2+NaCl]eq at halite saturation (at the time of Ttrap).
AC C
311
EP
307
Primary stage II inclusions in calcite (e.g., Figs. 5D–F) yielded Ttrap values between 54
312
and 78 °C (mean and S.D. of 65 ± 9 °C; n = 7) and Tm (ice) values between −47.9 and −35.9 °C.
313
Inclusions in two Coboconk samples yielded Tm (hydrohalite) values from −22.9 to −16.5 °C.
314
The calculated salinity of the halite-saturated stage II fluid at Ttrap varies between 30.7 and 33.2
315
wt.% [CaCl2+NaCl]eq.
15
ACCEPTED MANUSCRIPT
316
Stage III is represented by primary inclusions in celestine and anhydrite (e.g., Figs. 5G–I) and by early secondary inclusions in stage II calcite. Crystallites of anhydrite were occasionally
318
identified in stage III inclusions in calcite (Diamond et al., 2015). The inclusions in celestine
319
and calcite yielded Ttrap values between 42 and 80 °C and Tm (ice) values from −46.5 to −33.2
320
°C, corresponding to salinities of 32 to 33 wt.% [CaCl2+NaCl]eq for the halite-saturated stage III
321
fluid at Ttrap.
RI PT
317
Secondary, stage IV inclusions in calcite (e.g., Figs. 6A–E) have minimum Ttrap values
323
between 57 and 106 °C. Their Tm (ice) values vary from −52.7 to −26.4 °C, which indicates a
324
narrow range of halite-saturated salinities between 32 and 34 wt.% [CaCl2+NaCl]eq for this gas-
325
undersaturated fluid at minimum Ttrap.
M AN U
326
SC
322
Late secondary, liquid-only (metastable) inclusions associated with stage V (e.g., Figs. 6F–G) formed at temperatures <70 °C (Goldstein and Reynolds, 1994) and were identified in
328
calcite and dolomite throughout the sample suite. Stage V inclusions from the Gull River
329
Formation and the Cambrian show Tm(ice) values of −35.0 to −24.3 °C, indicating salinities of
330
23 to 26 wt.% [CaCl2+NaCl]eq.
Hydrocarbon inclusions (stage VI) in calcite (e.g., Figs. 6H–I) from the Coboconk and
EP
331
TE D
327
Gull River formations contain variable proportions of liquid oil and methane vapour, indicating
333
they were gas-saturated at the time of trapping. Their minimum Th values yield Ttrap values of 48
334
and 59 °C, respectively.
335 336 337 338
AC C
332
4.3 C- and O-isotope Compositions Calcite vein and vug samples from the Cambrian sandstones yielded the most 13Cdepleted C-isotope compositions in the sedimentary succession, with δ13C values of −6.1 to
16
ACCEPTED MANUSCRIPT
−3.5‰ (Fig. 9A; Table 3S – Supporting Material). Cambrian dolomite samples, including one
340
dolostone and one saddle dolomite, have δ13C values of −2.5‰ and −3.8‰, respectively.
341
Similar C-isotope compositions were obtained for two samples from a vug in the Shadow Lake
342
Formation, including δ13C values of −3.4‰ for saddle dolomite and −5.0‰ for calcite (Table 3S
343
– Supporting Material).
344
RI PT
339
A pronounced C-isotope shift of +2 to +3‰ occurs at the boundary between the Shadow Lake and Gull River Formations. Uniform C-isotope compositions are observed for most
346
samples from the base of the Gull River Formation to the Cobourg Formation, with δ13C values
347
of −1.0 to +1.9‰ and no identifiable differences between the Black River and Trenton group
348
samples. Similarly, no differences in δ13C are observed between samples of limestone matrix
349
and calcite veins and vugs, and samples of saddle dolomite, dolostone and matrix replacement
350
dolomite. Although most of the C-isotope data are constant versus depth, three calcite vein
351
samples from a 3 m section at the base of the Kirkfield Formation (DGR-6-864.60; DGR-6-
352
865.20; DGR-6-867.60) vary by 2.7‰ (−0.8 to +1.9‰; Fig. 9A). It should be noted, however,
353
these data fall within the broader range of C-isotope compositions defined by other samples from
354
the Ordovician succession, including a limestone from the Gull River Formation (−1.0‰) and
355
dolostone from the Coboconk Formation (+1.9‰).
M AN U
TE D
EP
AC C
356
SC
345
Calcite vein and vug samples from the Cambrian have δ18O values between +14.6‰ and
357
+24.2‰ and overall represent the most 18O-depleted compositions in the succession, with only
358
one of these Cambrian calcite samples higher than +17.8‰, (Fig 9B; Table 3S – Supporting
359
Material). Two dolomite samples from the Cambrian units have δ18O values of +23.0‰
360
(dolostone) and +23.2‰ (saddle dolomite). Two samples from a vug in the Shadow Lake
361
Formation (DGR-2-843.83) yielded similar O-isotope compositions to the other respective
17
ACCEPTED MANUSCRIPT
362
calcite and dolomite samples in the Cambrian sample suite, and include δ18O values of +15.9‰
363
for calcite and +20.6‰ for saddle dolomite.
364
A pronounced shift of ~ +3 to +7‰ is observed in the δ18O data along the boundary between the Shadow Lake and Gull River formations (Fig. 9B). Calcite vein and vug samples
366
from the Gull River and Coboconk formations have a wide range of δ18O values, from +18.9 to
367
+26.7‰. Notably, a ~ 10 m section at the top of the Coboconk Formation, along the Trenton–
368
Black River Group boundary, exhibits a spread in δ18O values of up to 5.4‰ (+21.3 to +26.7‰).
369
Calcite veins and vug samples from the Trenton Group have δ18O values from +23.0 to +28.1‰,
370
and are broadly consistent with calcite O-isotope data for the Black River Group. Most of the
371
Trenton Group calcite samples vary by up to 3.1‰, with only one exceptional δ18O value higher
372
than +26.1‰ (Fig. 9B). Six saddle dolomite samples from the combined Black River–Trenton
373
group units have uniform δ18O values of +20.3 to +21.3‰, and are > 1–2‰ lower in δ18O than
374
most calcite vein and vug samples in the Ordovician succession. In contrast, one sample of
375
dolostone matrix (DGR-2-778.44) and one replacement dolomite sample (DGR-4-772.10) from
376
the Coboconk Formation yielded δ18O values of +23.6‰ and +22.0‰, respectively. Eleven
377
limestone matrix samples from the Trenton and Black River groups exhibit a subtle trend of
378
increasing δ18O values towards the surface, from +23.8 to +26.8‰, and no identifiable shifts or
379
sections of O-isotope variability along the boundary between the two groups, as observed in the
380
calcite vein and vug data (Fig. 9B).
382 383 384
SC
M AN U
TE D
EP
AC C
381
RI PT
365
4.4 Sr-isotope ratios Secondary calcite and dolomite samples from the Cambrian units have the highest Srisotope ratios in the sedimentary succession. Four calcite vein/vug samples have 87Sr/86Sr values
18
ACCEPTED MANUSCRIPT
from 0.70975 to 0.71039, while one sample of dolostone matrix and one saddle dolomite yielded
386
similar 87Sr/86Sr values of 0.71043 and 0.71027, respectively (Fig 9C; Table 3S – Supporting
387
Material). Furthermore, two calcite/saddle dolomite samples from a vug in the Shadow Lake
388
Formation have 87Sr/86Sr values of 0.71037 and 0.70984, and overlap in Sr-isotope composition
389
with other samples from the Cambrian.
RI PT
385
A shift in 87Sr/86Sr of up to ~ –0.02 is observed across the boundary between the Shadow
390
Lake and the Gull River formations. Calcite vein and vug samples from the Gull River and
392
Coboconk formations have the highest Sr-isotope variability in the succession, with 87Sr/86Sr
393
values ranging from 0.70799 to 0.70990 (Fig. 9C). Similar variability is observed for three
394
samples of saddle dolomite, one dolostone and one sample of anhydrite from the Gull River–
395
Coboconk units, with combined 87Sr/86Sr values from 0.70854 to 0.70939. In contrast, three
396
limestone samples from these Black River Group units (two Gull River, one Coboconk) have
397
uniform 87Sr/86Sr values of 0.70810 to 0.70833, and overlap with the lowest Sr-isotope ratios of
398
calcite veins from these units.
TE D
M AN U
SC
391
399
For the Trenton Group, five calcite vein/vug samples and three limestone matrix samples have uniform 87Sr/86Sr values of 0.70790 to 0.70817 (Fig. 9C), and are consistent with the
401
87
402
values of 0.70818, 0.70808 and 0.70848 are observed for anhydrite samples from the Cobourg
403
and Sherman Fall formations, and a celestine sample from the Kirkfield Formation, respectively
404
(Fig. 9C). One sample of saddle dolomite from a vug in the Sherman Fall Formation (DGR-5-
405
743.10) has an exceptional 87Sr/86Sr value of 0.70856. However, this sample has an associated
406
2σ analytical uncertainty of ± 0.00083, which is 20–40 times larger than other analyses (Table
407
3S – Supporting Material).
EP
400
AC C
Sr/86Sr values of limestone samples from the underlying Black River Group. Similar 87Sr/86Sr
19
ACCEPTED MANUSCRIPT
408 409
5. Discussion
410
5.1 Paragenetic sequence Overall six distinct fluid events were identified in the microthermometric/petrographic
RI PT
411
results (Figs. 3 and 7), with the first two stages responsible for most of the mineral precipitation
413
in veins and vugs and only minor amounts of anhydrite and celestine that formed during stage
414
III. Based on the preservation of well-defined dolomite, calcite, anhydrite and celestine crystal
415
margins from precipitation events during stages I to III (Fig. 2), and on the absence of dissolution
416
features, fluid stages IV–VI are interpreted as migration of relatively minor quantities of fluids
417
along grain boundaries or micro-fractures.
M AN U
418
SC
412
Stage I is represented by formation of Fe- and Mn-enriched saddle dolomite in the Black River Group (Diamond et al., 2015) that resulted from the influx of a halite-saturated brine (34
420
wt.% [CaCl2+NaCl]eq) and free methane gas at temperatures of 122–128 °C. Elsewhere in the
421
Michigan Basin, high temperatures have been determined for fluid inclusions in saddle dolomite
422
from the Trenton–Black River groups, including up to ~160 °C in the Albion-Scipio oil field of
423
south-central Michigan (Allan and Wiggins, 1993) and temperatures as high as ~170 °C and 220
424
°C in dolomite samples from Manitoulin Island and southwestern Ontario, respectively (Coniglio
425
et al., 1994). Elevated temperatures of ~97–144 °C were also obtained by Haeri-Ardakani et al.
426
(2013) for saddle dolomite samples from the Trenton Group in southwestern Ontario. Davies
427
and Smith (2006) suggested that the formation of saddle dolomite in these fault-controlled
428
reservoirs is due to ascending fluid that is hotter than the ambient or burial temperatures of the
429
host formation (i.e. hydrothermal). In our study, a third Fe- and Mn-enriched saddle dolomite
430
sample from the Cambrian yielded a significantly lower Ttrap value of 88 °C, but a similarly high
AC C
EP
TE D
419
20
ACCEPTED MANUSCRIPT
431
salinity of 34 wt.% [CaCl2+NaCl]eq. It is not known whether these saddle dolomite samples are
432
contemporaneous.
433
Stage II primary inclusions in calcite have Ttrap values of 54–78 °C (mean 65 ± 9 °C; n = 7) and represent a halite-saturated brine with a salinity of 31 to 37 wt.% [CaCl2+NaCl]eq and
435
entrained free methane gas. Based on the predominance of calcite as a vein/vug mineral, it can
436
be presumed that stage II was the most significant vein- and vug-forming fluid event. The range
437
of Ttrap values from stage II inclusions is slightly lower than minimum paleotemperature
438
estimates of 80 °C at the base of the Silurian and 90 °C at the base of the Trenton Group reported
439
by Legall et al. (1981). Determinations of the absolute timing of vein and vug calcite formation
440
in the Trenton–Black River groups have been made using U-Pb dating methods with laser
441
ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) and thermal ionization
442
mass spectrometry (TIMS) (Davis et al., 2013). Davis et al. (2013; 2017) reported a LA-ICP-MS
443
U-Pb age of 434 ± 5 Ma, based on a calculated mean of three model age determinations from the
444
Cobourg and Coboconk formations, and a TIMS U-Pb date of 451 ± 38 Ma for a calcite vein
445
from the Cobourg Formation. The Davis et al. (2013; 2017) U-Pb data compare well with a Rb-
446
Sr isochron age of 425 ± 6 Ma (errorchron age of 440 ± 50 Ma) by Harper et al. (1995) and 7
447
combined K-Ar dates ranging from 453 ± 9 Ma to 412 ± 8 Ma by Harper et al. (1995) and
448
Ziegler and Longstaffe (2000) (weighted mean of 442 ± 16 Ma) for secondary K-feldspar
449
obtained along the Precambrian−Paleozoic boundary in southern Ontario. These data suggest the
450
influx of stage II, calcite-forming fluids in the Trenton–Black River groups likely occurred
451
between the Late Ordovician and Silurian.
452 453
AC C
EP
TE D
M AN U
SC
RI PT
434
Stage III fluid inclusions have Ttrap values of 42–80 °C and represent a stage of fluid migration during which anhydrite and celestine were formed. The mean Ttrap value of 62 ± 17 °C
21
ACCEPTED MANUSCRIPT
for four stage III inclusions is similar to the mean Ttrap value (65 ± 9 °C; n = 7) for stage II
455
primary inclusions in calcite. Salinities of stage III inclusions range from 32 to 33 wt.%
456
[CaCl2+NaCl]eq, and are comparable to the salinity estimates for stage I and stage II inclusions.
457
This suggests a paragenetic link between the stage I–III fluids, with saddle dolomite forming
458
during initial brine influx at 122–128 °C (stage I), then subsequent cooling and formation of
459
calcite (stage II) and penecontemporaneous formation of anhydrite and celestine (stage III). Secondary fluid inclusions defining stage IV have minimum Ttrap values between 57 and
SC
460
RI PT
454
106 °C, including one calcite vein from the Sherman Fall Formation (DGR-2-691.70; see Figs.
462
6D–E) that yielded minimum Ttrap values ranging from 77 to 106 °C. The absolute timing and
463
burial depth (i.e. thickness of overburden sediments) at the time of stage IV fluid inclusion
464
entrapment is unclear, therefore, definitive pressure corrections to determine Ttrap (true) cannot
465
be applied to the stage IV data. In order to determine an upper limit for stage IV fluid inclusion
466
entrapment, Ttrap (maximum) values were calculated using pressure corrected data for peak burial
467
conditions for the Ordovician–Cambrian succession. This is a conservative approach because we
468
do not know that stage IV coincides with peak burial, and because all corrections are made to the
469
depth at the base of the Paleozoic succession. After peak burial conditions, it can be inferred
470
from the work of Coniglio and Williams-Jones (1992) that approximately 1000 m of Paleozoic
471
sediments were eroded from the sedimentary succession at the Bruce site (NWMO, 2011).
472
Utilizing an overburden thickness of 1000 m and a current depth of 860 m for the base of the
473
Paleozoic succession, a maximum burial depth of 1860 m was used to calculate an upper limit
474
for stage IV fluid inclusion entrapment. Pressure corrected Ttrap (maximum) values for the stage
475
IV inclusions range from 77–127 °C (Fig. 7; Table 1).
AC C
EP
TE D
M AN U
461
22
ACCEPTED MANUSCRIPT
There are two models that explain the formation of stage IV fluids: (1) influx of
477
hydrothermal brines, or (2) regional heating of the sediments (i.e. through conduction), coupled
478
with fracturing and re-mobilization of fluids. For model 1, significant quantities of fluid and
479
high flow rates along permeable structures would be required to facilitate the transfer of heat, but
480
there is no evidence for fluid movement at this scale (e.g., large fracture/veins systems,
481
dissolution/re-precipitation features, formation of new secondary minerals with primary
482
inclusions analogous to stage IV). Alternatively, a regional heating event (model 2) is consistent
483
with an estimated minimum paleotemperature of 90 °C at the base of the Trenton Group (Legall
484
et al., 1981). Evidence for a regional heating event exists elsewhere in the Michigan Basin.
485
Temperatures of ~70–170 °C and ~40–260 °C have been obtained for secondary fluid inclusions
486
in calcite from Trenton–Black River rocks on Manitoulin Island and in southwestern Ontario,
487
respectively (Coniglio et al., 1994). Haeri-Ardakani et al. (2013) also reported temperatures of
488
~60–131 °C for fluid inclusions in calcite cement samples from the Trenton Group in southwest
489
Ontario. In central Michigan, fluid inclusions in Middle Ordovician St. Peter Formation
490
sandstones yielded temperatures of 76–169 °C for quartz overgrowths, and 94–149 °C for
491
dolomite cements (Girard and Barnes, 1995), and are consistent with temperatures of ~ 100–160
492
°C obtained for fluid inclusions in saddle dolomite from elsewhere in central Michigan (Allan
493
and Wiggins, 1993). Overall, these high temperatures (up to 260 °C) strongly suggest that
494
regional heat sources were present at depth below the Michigan Basin. Reactivation of the mid-
495
continent rift was suggested by Ma et al. (2009) as a source of heat and mantle-derived He and
496
Ne anomalies in the hydrothermal fluids that interacted with the basin rocks. Regional-scale
497
heating has also been linked to the formation of secondary illite at 367–322 Ma in the St. Peter
498
sandstone along the base of the Paleozoic succession (Girard and Barnes, 1995). Reactivation of
AC C
EP
TE D
M AN U
SC
RI PT
476
23
ACCEPTED MANUSCRIPT
499
the mid-continent rift during the Late Devonian–Mississippian could be responsible for initiating
500
regional-scale heating below the Michigan Basin and may be linked to the stage IV fluid
501
migration event observed in the Ordovician carbonates at the Bruce site. Stage V and VI inclusions represent the latest observed records of fluid migration in the
503
sedimentary succession, and for the latter stage VI fluids, include hydrocarbon inclusions in the
504
Coboconk and Gull River formations and the Cambrian units. Samples from the Coboconk and
505
Gull River formations yielded trapping temperatures of 48 and 59 °C (Fig. 7). The occurrence of
506
hydrocarbon fluids is consistent with the presence of bitumen layering or oil in the Coboconk,
507
Gull River and Shadow Lake formations (see Fig. 2.32 in Intera, 2011), suggesting that stage VI
508
represents an important period of hydrocarbon generation and/or migration in the Black River
509
Group.
510 5.2 Source and evolution of fluids
TE D
511
M AN U
SC
RI PT
502
By combining the petrographic and fluid inclusion results with the C- O- and Sr-isotope
513
data, inferences can be made about the sources and paleofluid evolution through the secondary-
514
mineral formation events of stages I to III. Secondary calcite (veins/vugs) and dolomite
515
(dolostone matrix and saddle) samples from the combined Cambrian–Shadow Lake units have
516
δ13C values of ~ −6.1 to −2.5‰ (Fig. 9A). Estimates of the average C-isotope composition of
517
Cambrian seawater include δ13C values between −0.9‰ and +0.0‰ (Fig. 9A; Veizer et al.,
518
1999), suggesting that the calcite and dolomite samples from the Cambrian–Shadow Lake units
519
were not sourced from connate fluids derived from Cambrian seawater. Alternatively, these low
520
δ13C values reflect carbonate mineral formation from fluids that reacted with organic carbon.
521
Harper et al. (1995) reported similar 13C-depleted data (−9.3 to −3.9‰) for secondary calcite and
AC C
EP
512
24
ACCEPTED MANUSCRIPT
522
dolomite associated with hydrothermally altered granitoid rocks along the
523
Precambrian−Paleozoic boundary in southern Ontario.
524
A 2−3‰ shift in δ13C occurs at the boundary between Cambrian−Shadow Lake units and the overlying Gull River Formation. The C-isotope compositions of calcite and dolomite
526
samples show little scatter through the Gull River to Cobourg formations (Fig. 9A) and are
527
consistent with estimates by Veizer et al. (1999) for the average C-isotope composition of
528
seawater during the Late Ordovician (−0.1 to +1.7‰), suggesting a C source dominated by
529
seawater-derived connate fluids or external fluids that experienced high degrees of fluid–rock
530
interaction with the host limestones (Fig. 9A).
SC
M AN U
531
RI PT
525
Calcite samples from the Shadow Lake and Cambrian units generally have the lowest δ18O values (+14.6 to +24.2‰), with most calcite veins and vugs plotting at ~ +16‰ (7 of 8
533
calcite veins form a calculated mean of +15.9 ± 1.0‰; S.D.). A prominent feature of the O-
534
isotope data through the Cambrian−Ordovician succession is the >4‰ shift that occurs along the
535
boundary between the Cambrian–Shadow Lake units and the overlying Gull River Formation
536
(Fig. 9B). Relatively high variability is observed in the δ18O data for the Coboconk Formation
537
just below the Trenton−Black River group boundary (Fig. 9B), with values from calcite vein and
538
vug samples exhibiting a spread of up to 5.4‰. This variability could be explained by: (1)
539
mixing between two or more fluids that had distinct O-isotope compositions (i.e. connate vs.
540
external hydrothermal fluids), or (2) temperature-related differences in ∆18Ocalcite−fluid during the
541
precipitation of calcite (e.g., Friedman and O'Neil, 1977). The fact that large temperature
542
differences are not observed in the primary fluid inclusion data for the Coboconk Formation
543
(Fig. 7) suggests that the δ18O variability may be explained by mixing with external fluids. The
544
Trenton−Black River boundary represents a significant change in hydraulic conductivity, with
AC C
EP
TE D
532
25
ACCEPTED MANUSCRIPT
rocks in the Trenton Group having Kh values up to six orders of magnitude lower than those of
546
the Black River Group (Normani and Sykes, 2012; Beauheim et al., 2014; Khader and
547
Novakowski, 2014). Furthermore, thin dolostone units (< 20 cm) and a bentonite unit (< 10 cm
548
thick) occur as stratabound layers in the Coboconk Formation (~ 790.50 m and 781.00 mBGS,
549
respectively). These units, particularly the dolostone, may have relatively high permeability and
550
porosity compared to the bounding limestones, and as such, could represent pathways of
551
enhanced lateral flow, allowing for ingress of external fluids.
SC
552
RI PT
545
The O-isotope compositions of source fluids responsible for secondary mineral formation were estimated using the microthermometric data and mineral−fluid O-isotope exchange
554
thermometers for calcite (Friedman and O'Neil, 1977; Fig. 10A) and dolomite (Horitia, 2014;
555
Fig. 10B). The O-isotope composition of Late Ordovician seawater is shown in Figure 10, with
556
δ18O values between −4 and −2‰ (Veizer et al., 1999; Shields et al., 2003). Calculated values
557
for the O-isotope composition of fluids in equilibrium with all limestones from the Gull River to
558
Cobourg formations are consistent with Ordovician seawater (Fig. 10A), indicating that there has
559
been limited diagenetic or hydrothermal alteration effects to their primary O-isotope
560
compositions.
TE D
EP
561
M AN U
553
Calcite vein and vug samples from the Cambrian−Shadow Lake units have calculated δ18Ofluid values between ~ −6‰ and +4‰, based on formation temperatures of 70 to 72 °C (Fig.
563
10A). Gull River and Coboconk samples have similar calculated δ18Ofluid values of ~ −6 to
564
+4‰, while samples from the Trenton Group have calculated fluid compositions of ~ +2 to +8‰
565
(Fig. 10A). These fluid compositions are suggestive of calcite precipitation from Late
566
Ordovician seawater or an 18O-enriched, seawater-derived basin fluid. Similar 18O-enrichments
567
relative to seawater have been described for fluids in many sedimentary basins (Holser, 1979;
AC C
562
26
ACCEPTED MANUSCRIPT
Knauth and Beeunas, 1986; Ayalon and Longstaffe, 1988; Kyser and Kerrich, 1990; Wilson and
569
Long, 1993a; 1993b; Hanor, 1994), and are interpreted to have undergone 18O-enrichment by
570
evaporation and/or fluid−rock interaction following burial. Modification of seawater-derived
571
connate fluids by influx of hydrothermal fluids that interacted with, or were derived from the
572
Precambrian shield (Spencer, 1980) could also explain such 18O-enrichments.
573
RI PT
568
Calculated fluid compositions for saddle dolomite have δ18Ofluid values of ~ 0 to +7‰ (Fig. 10B) and overlap with the calculated fluid compositions of calcite, suggesting that dolomite
575
precipitated from 18O-enriched hydrothermal fluids originating from evolved seawater. No
576
microthermometry data could be obtained from fluid inclusions in samples of matrix replacement
577
dolomite and dolostone. However, if the matrix replacement dolomite and dolostone samples
578
formed at temperatures similar to those of calcite veins and vugs (54 to 78 °C) or saddle
579
dolomite (88 to 128 °C), the calculated fluid compositions would be −6 to +1‰ or +1 to +7‰,
580
respectively (Fig. 10B). These calculated fluid compositions are suggestive of matrix
581
replacement dolomite and dolostone formation directly from Late Ordovician seawater or
582
hydrothermal fluids derived from evolved seawater, and are broadly consistent with fluid source
583
determinations for all other secondary minerals.
M AN U
TE D
EP
584
SC
574
Sr-isotope data for many of the samples from the Black River Group, and all samples from the Cambrian, exhibit 87Sr/86Sr enrichment well above the seawater curve of Veizer et al.
586
(1999). Enriched 87Sr/86Sr values (> 0.710) are a common signature of Precambrian brines (e.g.,
587
Frape et al., 1984; McNutt et al. 1984; Bottomley et al., 1999), and the observed enrichment
588
suggests vein and vug formation from shield-derived hydrothermal fluids, and/or fluid–rock
589
interaction with shield-derived detrital minerals in the Cambrian sandstones. In contrast, all vein
590
and vug samples from the Trenton Group, and approximately half of the samples from the Black
AC C
585
27
ACCEPTED MANUSCRIPT
River Group, overlap with the seawater curve of Veizer et al. (1999), suggesting they formed
592
from either connate fluids derived from Late Ordovician seawater and/or external hydrothermal
593
brines that underwent high degrees of fluid–rock interaction and associated modification of their
594
original isotopic signature during transit through the underlying stratigraphy.
RI PT
591
The combined microthermometry and isotopic data are consistent with a model of
596
secondary mineral formation by fluid mixing; with one end member being either a shield-derived
597
hydrothermal fluid, and/or a fluid that was isotopically influenced by detrital minerals in the
598
Cambrian sandstone, and the second end member being a connate pore fluid derived from the
599
Paleozoic carbonates.
M AN U
SC
595
600 601 602
5.3 Implications for the migration of high-temperature fluids in the Ordovician limestones An important feature of our fluid influx model for the Ordovician sedimentary succession is the formation of dolomite in the Black River Group at temperatures of 122–128 °C, which are
604
higher than the nearest available estimate of peak burial temperature, 55–90 °C from Coniglio
605
and Williams-Jones (1992) at Manitoulin Island, providing evidence for migration of
606
hydrothermal fluid.
EP
TE D
603
Hydrothermal dolomite is an important reservoir for oil and gas plays within the
608
Trenton−Black River groups in the Albion-Scipio/Stoney Point fields in south-central Michigan
609
(Hurley and Budros, 1990; Allan and Wiggins, 1993; Davies and Smith, 2006), the Lima-Indiana
610
field in Ohio/Indiana (Wickstrom et al., 1992) and southern Ontario (Powell et al., 1984;
611
Middleton et al., 1993; Coniglio et al., 1994; Obermajer et al., 1999). In these regions,
612
hydrothermal dolomite developed along large-scale extensional faults as a result of high-
613
temperature fluid−rock interaction (up to ~ 260 °C; Coniglio et al., 1994) between limestone and
AC C
607
28
ACCEPTED MANUSCRIPT
hydrothermal fluids that migrated upward from the underlying Cambrian aquifer. Hydrothermal
615
dolomite reservoirs are recognized in 2D seismic data aligned with vertical “sag” or “flower”
616
structures, often 10s of kilometers in strike length and 100s of meters in width (e.g., Davies and
617
Smith, 2006). At that scale, these structures represent regionally extensive zones of
618
dolomitization. At the Bruce site, however, the abundance and scale of hydrothermal dolomite
619
occurrence is much different. Saddle dolomite with associated high fluid trapping temperatures
620
was observed in veins within Ordovician limestone, stratabound dolostone in the Ordovician, and
621
in Cambrian dolostone. These features are typically mm or cm in thickness (e.g., Fig 2B–E; 2G;
622
2H; Fig. 1S – Supporting Material), and as such, do not represent reservoir-scale structures of
623
dolomitization as observed elsewhere in the Michigan Basin.
M AN U
SC
RI PT
614
624 625
5.4 Conceptual model
Models for the source and timing of paleofluid movements in the Cambrian and
627
Ordovician sedimentary units below the Bruce site, until now, have been largely based on direct
628
geochemical and isotopic constraints from porewater extracted from drill cores (Clark et al.,
629
2010a; 2010b; 2013; 2015; Al et al., 2015). In a detailed study, Clark et al. (2013) reported O-
630
and H-isotope compositions and Cl- and Br-concentration data for porewater obtained
631
throughout the Paleozoic sedimentary succession. Downward trends of decreasing Cl− and Br−
632
concentrations, coupled with decreases in δ18O, were observed in the Ordovician limestones
633
(Fig. 2 in Clark et al., 2013). These trends were interpreted to represent a mixing profile
634
between hypersaline evaporated seawater from the overlying Silurian Salina Formation and
635
shield-derived fluids (Clark et al., 2013; Al et al. 2015). The influence of a shield end member is
636
supported by a downward trend of 2H-enrichment – a common feature of shield fluids (Fritz and
AC C
EP
TE D
626
29
ACCEPTED MANUSCRIPT
637
Frape, 1982; Frape et al., 1984; Pearson, 1987; Bottomley et al., 1999, 2004; Douglas et al.,
638
2000; Gascoyne, 2004; Al et al., 2015).
639
Clark et al. (2013) also reported He-isotope ratios for porewater from the Ordovician shales and the Trenton and Black River group limestones. Porewater extracted from the shales
641
and the Cobourg Formation limestone yielded uniform R/RA values (0.020 ± 0.002; where R/RA
642
= (3He/4He)sample/(3He/4He)air, and (3He/4He)air = 1.38 x 10-6), which are consistent with an
643
authigenic origin, and the total He concentrations indicated a minimum period of 260 Ma for He
644
ingrowth. A positive shift in R/RA values was observed from the base of the Cobourg to the
645
upper portion of the Sherman Fall Formation, followed by uniform R/RA values of 0.035 ± 0.002
646
with increasing depth into the Cambrian (Fig. 4; Clark et al., 2013). The elevated R/RA values
647
below the Cobourg Formation limestone were interpreted by Clark et al. (2013) to reflect
648
allochthonous fluids of mixed basin and shield origin.
SC
M AN U
Furthermore, 87Sr/86Sr data were reported by Clark et al. (2010a; 2010b) for porewater
TE D
649
RI PT
640
samples obtained at varying depths in the sedimentary succession. Porewater samples from the
651
Ordovician shales yielded uniform Sr-isotope compositions (mean 87Sr/86Sr value of ~ 0.710)
652
that are enriched in radiogenic Sr relative to the seawater curve of Veizer et al. (1999). Similar
653
to the interpretation of the He isotope data, Clark et al. (2010a; 2010b) considered the 87Sr
654
enrichment to have resulted from in situ ingrowth by decay of 87Rb in the porewater. Porewater
655
samples from the Trenton and Black River group limestones yielded highly variable 87Sr/86Sr
656
values, between 0.7093 and 0.7103, and consistent with the He-isotope data, were interpreted as
657
resulting from mixing between basin- and shield-derived fluids.
658 659
AC C
EP
650
The isotopic and fluid inclusion data from secondary minerals are generally consistent with the conceptual model for porewater evolution presented by Clark et al. (2013) and Al et al.
30
ACCEPTED MANUSCRIPT
(2015), but our new data allow for some refinements to the model. Based on the porewater data,
661
Clark et al. (2013) and Al et al. (2015) suggest that mixing between basin- and shield-derived
662
fluids was driven by diffusion, but the secondary mineral data indicate that advection in small
663
fractures also contributed to upward transport of shield-type, halite-saturated fluids during stages
664
I to III. These events formed the secondary vein minerals and are broadly constrained by U-Pb
665
dating (Davis et al., 2013) to the Late Ordovician or Silurian. The isotopic signature of the
666
shield end member does not extend upward into the Trenton Group, likely because it was
667
eliminated by fluid−rock interaction. The fact that present-day porewater in the Black River
668
Group is below halite saturation is consistent with indications of allochthonous porewater from
669
87
670
fluid migration that are not recorded in the chemical and isotopic composition of the vein
671
minerals.
M AN U
SC
RI PT
660
Sr/86Sr and 3He/4He ratios, and likely reflects the later stages (V to VI) of halite-undersaturated
673 674
Conclusions
TE D
672
Overall, the combined microthermometric and isotopic data allow us to establish a model for the fluid migration history associated with the formation of secondary vein minerals in low-
676
permeability Ordovician limestones of the Michigan Basin. Six stages of fluid migration are
677
suggested by the fluid inclusion data. These comprise fluid stages I through III, which are
678
represented by primary inclusions trapped during the formation of dolomite (stage I: 88–128 °C),
679
calcite (stage II: 54–78 °C) and celestine/anhydrite (stage III: 42–80 °C). The ubiquitous
680
presence of variable amounts of halite in these fluid inclusions indicates that the secondary vein
681
minerals precipitated from halite-saturated brines with salinities of 31–37 wt.% [CaCl2+NaCl]eq,
682
and is suggestive of a paragenetic link between fluid stages I–III.
AC C
EP
675
31
ACCEPTED MANUSCRIPT
683
Secondary, stage IV inclusions in calcite have a similar, halite-saturated fluid composition as stages I–III, but were trapped above 57–106 °C. No new secondary vein
685
minerals formed during this stage and dissolution features are absent from pre-existing secondary
686
minerals. The stage IV inclusions are interpreted as re-mobilization of trapped fluids along grain
687
boundaries or micro-fractures during a Late Devonian–Mississippian regional heating event.
688
Stages V and VI represent the latest recognized fluid migrations in the sample suite, during
689
which a halite-undersaturated brine (23–26 wt.% [CaCl2+NaCl]eq) was trapped at temperatures <
690
70 °C.
Secondary minerals from the Cambrian and Shadow Lake units are characterized by the
M AN U
691
SC
RI PT
684
lowest δ13C and δ18O values in the succession and 87Sr/86Sr ratios that reflect formation from
693
radiogenic Sr sources. The combined Cambrian–Shadow Lake isotopic data are suggestive of
694
secondary mineral formation from hydrothermal brines that experienced high degrees of fluid–
695
rock interaction with the underlying shield, and/or with shield-derived minerals in the Cambrian
696
sandstone. For the Coboconk and Gull River formations, greater variability in the δ18O and
697
87
698
connate fluids and hydrothermal brines, with the latter having a shield signature sourced from
699
below. Uniform C- O- and Sr-isotope compositions for Trenton Group samples are suggestive of
700
secondary mineral formation from either connate fluids, which were sourced directly from Late
701
Ordovician seawater (or 18O-enriched, evolved seawater), or mixed with hydrothermal brines
702
that underwent high degrees of fluid–rock interaction.
703
TE D
692
AC C
EP
Sr/86Sr values of the secondary minerals is interpreted to reflect mixed fluid sources, including
Age constraints on the absolute timing of secondary mineral formation (stages I–III) are
704
gleaned from previous geochronology studies in the region, which indicate that brine influx
705
occurred during the Late Ordovician to Silurian. Secondary inclusions associated with stage IV–
32
ACCEPTED MANUSCRIPT
VI inclusions did not form new secondary vein minerals and these late fluid stages are
707
interpreted as re-mobilization and/or minor fluid migration along grain boundaries and micro-
708
fractures during a regional heating event. The timing of this regional heating event cannot be
709
easily constrained from our data, however, elsewhere in the Michigan Basin it has been
710
suggested that reactivation of the mid-continent rift during the Late Devonian–Mississippian
711
could be responsible for regional-scale heating and associated fluid migration. Despite evidence
712
elsewhere in the Michigan Basin for secondary-mineral formation during the Late Devonian–
713
Mississippian, the Ordovician limestones at the Bruce site show no evidence of large-scale fluid
714
ingress during this time period (e.g., dissolution/re-precipitation features, formation of new
715
secondary vein minerals); which suggests that the low permeability nature of these rocks had
716
been established before this time.
717
719
Acknowledgements
TE D
718
M AN U
SC
RI PT
706
We are grateful to Paul Middlestead, Wendy Abdi and Patricia Wickham for assistance in the G.G. Hatch Stable Isotope Laboratory (uOttawa), and to Douglas Hall and Steven Cogswell
721
for help with EPMA and SEM work at the University of New Brunswick. Shuangquan Zang,
722
E.A. Spencer and Rhea Mitchell are acknowledged for their assistance with Sr-isotope work in
723
the Isotope Geochemistry and Geochronology Research Centre (Carleton). Rita Caldas kindly
724
assisted with initial microthermometry. We gratefully acknowledge Ian Clark and Josué Jautzy
725
(uOttawa) for discussions on fluid migration models in the Michigan Basin. We are also grateful
726
to Don Davis and Chelsea Sutcliffe (U of T) for thoughtful discussions on the absolute timing of
727
calcite vein emplacement in the Ordovician carbonates. An anonymous reviewer and Laura
728
Kennell-Morrison and Richard Crowe (NWMO) are thanked for comments and suggestions on
AC C
EP
720
33
ACCEPTED MANUSCRIPT
729
an earlier version of the manuscript. Financial support for this project was provided by the
730
NWMO and NSERC, through an NSERC Industrial Postgraduate Scholarship to JS and an
731
NSERC/NWMO Collaborative Research and Development grant (477852-14) held by TA.
RI PT
732 733 Tables
735
Table 1. Microthermometric results for fluid inclusions in the secondary minerals.
SC
734
736
M AN U
737 Figure Captions
739
Figure 1. Bedrock geology, stratigraphy and hydrostratigraphy of the Michigan Basin below the
740
Bruce site, Southwestern Ontario, Canada. Also shown here are regional structural features
741
(Algonquin Arch, Chatham Sag) and contours lines for the base of the Paleozoic sediments (i.e.
742
elevation in metres to the top of the Precambrian basement rocks). The geologic map is based on
743
Ontario Geological Survey (1991) and Armstrong and Carter (2010).
EP
744
TE D
738
Figure 2. Photomicrographs of secondary minerals in the Trenton Group (A–B), Black River
746
Group (C–F) and the Cambrian units (G–I). (A) DGR-5-704.44, sharp-edged calcite vein in an
747
argillaceous limestone, with an infill of anhydrite that post-dates calcite formation (SL); (B)
748
DGR-2-691.70, irregular calcite vein in a fossiliferous limestone with zones of fine-grained
749
replacement dolomite (SL); (C) DGR-2-778.44, dolostone with zones of tan saddle dolomite and
750
discrete veins of colourless saddle dolomite (SL); (D) DGR-2-817.41, micritic limestone with a
751
saddle dolomite vug with infill anhydrite. Note the fine-grained dolomite lining the interior of
AC C
745
34
ACCEPTED MANUSCRIPT
vug (PPL); (E) DGR-4-829.53, dolostone with saddle dolomite vug, infill anhydrite and discrete
753
veins of colourless dolomite (SL); (F) DGR-2-843.83, micritic limestone with a sparry calcite
754
vug, infill minerals include rough-edged pyrite partly replaced and followed by crystallites of
755
magnetite suspended in an organic-rich matrix (RL); (G) DGR-4-844.78, irregular vein with
756
saddle dolomite and quartz lining and calcite infill within a fine-grained dolostone (RL); (H)
757
DGR-4-845.20, dolostone with saddle dolomite-lined calcite vug (PPL); (I) DGR-2-847.42,
758
sandstone with a calcite-lined vug. Note the euhedral authigenic quartz rims that mantle detrital
759
quartz grains (XPL). Abbreviations: Anh – anhydrite; Cal – calcite; Dol – dolomite; Mag –
760
magnetite; PPL – plane-polarized light image; Py – pyrite; RL – reflected light image; SDol –
761
saddle dolomite; SL – stereo-light image (combination of both transmitted and reflected light);
762
Qtz – quartz; XPL – cross-polarized light image.
M AN U
SC
RI PT
752
763
Figure 3. Relative timing of secondary mineral formation and trapping of fluid inclusion stages
765
I–VI, based on detailed petrography. The timing of authigenic quartz and pyrite formation
766
relative to fluid stages I–II and III–V, respectively, is uncertain and indicated by the dashed lines.
767
Note: matrix replacement dolomite predates the formation of saddle dolomite.
EP
768
TE D
764
Figure 4. Compositions of the secondary minerals in the DGR samples, including wt.% dolomite
770
relative to calcite (A), which was determined using XRD (Saso, 2013) and MgO (B) and Sr (C)
771
concentrations determined using EPMA. Note, the location of stratabound dolostone layers in
772
the succession are indicated in (B) by block symbols. Boundaries in the sedimentary succession
773
are also shown here (in descending order): Queenston–Collingwood/Cobourg (Trenton Group);
AC C
769
35
ACCEPTED MANUSCRIPT
774
Kirkfield (Trenton Group)–Coboconk (Black River Group); Shadow Lake (Black River Group)–
775
Cambrian; Cambrian–Precambrian.
776 Figure 5. Photomicrographs of fluid inclusions representative of paleofluid stages I–III, all
778
viewed in plane-polarized, transmitted light. (A) Stage I in DGR-4-829.53, saddle dolomite
779
crystal showing curved crystal faces lining the walls of a vug. The core is rich in primary fluid
780
inclusions and is overgrown by a clear, inclusion-free outer rim; (B–C) Details of A showing
781
variable proportions of liquid brine, CH4–CO2–vapour and halite within individual inclusions of
782
the same assemblage, indicating that all three phases were present at mutual saturation during
783
fluid entrapment; (D) Stage II in DGR-3-799.15, calcite crystals in a vug, containing clouds of
784
three-dimensionally distributed primary fluid inclusions. The rims of these crystals contain
785
linear arrays of fluid inclusions that radiate outwards, parallel to the growth direction of the
786
crystal, and are indicative of primary inclusions. All are younger than inclusions in A, because
787
calcite systematically overgrows saddle dolomite (Fig. 2H); (E–F) Details of D showing variable
788
proportions of liquid brine, CH4–vapour and halite within individual inclusions of the same
789
assemblage, indicating that all three phases were present at mutual saturation during fluid
790
entrapment; (G) Stage III in DGR-4-829.53, primary inclusions within fibrous anhydrite. The
791
inclusions are younger than in D, because anhydrite systematically overgrows calcite (Fig. 2A);
792
(H–I) Details of G showing variable proportions of liquid brine, CH4–vapour and halite within
793
individual inclusions of the same assemblage, indicating that all three phases were present at
794
mutual saturation during fluid entrapment.
AC C
EP
TE D
M AN U
SC
RI PT
777
795
36
ACCEPTED MANUSCRIPT
Figure 6. Photomicrographs of fluid inclusions representative of paleofluid stages IV–VI, all
797
viewed in plane-polarized, transmitted light. (A) Stage IV in DGR-2-847.42, planar array of
798
secondary fluid inclusions on a healed fracture in vein calcite; (B–C) Details of A showing
799
uniform proportions of liquid brine and H2O–vapour but variable proportions of halite within
800
individual inclusions of the same assemblage. The inclusions homogenize to liquid, therefore
801
only brine and halite were present (without free vapour) at mutual saturation during fluid
802
entrapment; (D) Stage IV in DGR-2-691.70, planar array of secondary inclusions on a healed
803
fracture in vein calcite, as in A; (E) Detail of D showing the same relations as in (B). This key
804
assemblage homogenizes to liquid at 106 °C; Sample (F) Stage V in DGR-2-844.31, planar
805
array of secondary liquid-only inclusions (metastable stretched water) on a healed fracture in
806
vein calcite. Such assemblages cross-cut stage IV assemblages (not shown); (H) Stages IV and
807
VI in DGR-6-892.99, viewed in simultaneous transmitted and UV epi-illumination.
808
Interpretation of relative timing of secondary fluid inclusion generations based on cross-cutting
809
relations. Traces of two healed fractures are visible, each marked by secondary fluid inclusions.
810
Green: Stage IV aqueous inclusions within a fracture trace that strikes to top-left of image and
811
dips steeply to top-right. Blue: Stage VI blue fluorescent liquid hydrocarbon inclusions with
812
CH4-vapour bubbles in a fracture trace that strikes to top-right and dips shallowly to bottom-
813
right. No inclusion is present at the intersection of the fracture traces (at shallow focus), so the
814
relative ages of the two fractures are ambiguous; (I) Same view as H but at a deeper focus level.
815
A hydrocarbon inclusion is present at the intersection of the fracture traces, indicating that the
816
hydrocarbon inclusions and their host fracture are younger than the aqueous inclusions and their
817
corresponding host fracture.
AC C
EP
TE D
M AN U
SC
RI PT
796
818
37
ACCEPTED MANUSCRIPT
Figure 7. Microthermometric data for assemblages of primary and secondary fluid inclusions in
820
the secondary minerals. For the stage IV inclusions, which have uniform liquid–vapour
821
proportions, each symbol represents a range of estimated temperatures between the minimum
822
Ttrap value (i.e. Th,max) and a pressure corrected Ttrap value, which was calculated for peak burial
823
conditions using an estimated burial depth of 1860 m (see section 5.1). Estimated Ttrap values of
824
30–70 °C are indicated for stage V inclusions. Boundaries in the sedimentary succession are
825
also shown here (in descending order): Queenston–Collingwood/Cobourg (Trenton Group);
826
Kirkfield (Trenton Group)–Coboconk (Black River Group); Shadow Lake (Black River Group)–
827
Cambrian; Cambrian–Precambrian.
M AN U
828
SC
RI PT
819
Figure 8. Salinity estimates for assemblages of primary and secondary fluid inclusions in the
830
secondary minerals, derived from Tm (ice) values for all fluid stages (A–B) shown in a H2O–
831
NaCl–CaCl2 ternary model. Stable phase relations of Oakes et al., (1990) and Steele-MacInnis et
832
al. (2011) are also shown in (A). The metastable cotectic in (A) and (B) was calculated in this
833
study.
EP
834
TE D
829
Figure 9. C-, O- and Sr-isotope results for the secondary minerals in the Ordovician–Cambrian
836
sedimentary succession (A–C, respectively). Average C-isotope compositions of seawater
837
during the Late Ordovician (−0.1 to +1.7‰) and the Cambrian (−0.9 to +0.0‰) are also shown
838
in (A) and are based on marine carbonate rocks from Veizer et al. (1999). Average 87Sr/86Sr
839
values for Late Ordovician seawater (0.70887 to 0.70941) were obtained from Veizer et al.
840
(1999). The 87Sr/86Sr values of Precambrian shield brines from the Sudbury region (0.71037 to
841
0.74009) were reported by McNutt et al. (1984). Boundaries in the sedimentary succession are
AC C
835
38
ACCEPTED MANUSCRIPT
842
also shown here (in descending order): Queenston–Collingwood/Cobourg (Trenton Group);
843
Kirkfield (Trenton Group)–Coboconk (Black River Group); Shadow Lake (Black River Group)–
844
Cambrian; Cambrian–Precambrian.
RI PT
845 846
Figure 10. Estimates of O-isotope compositions of fluids (contours) that precipitated carbonate
848
minerals based on mineral–fluid exchange thermometers of Friedman and O'Neil (1977) for
849
calcite (A) and Horitia (2014) for dolomite (B). The boxed fields represent formation
850
temperature estimates, based largely on fluid inclusion data discussed in Section 5.1, plotted
851
against δ18O values for each calcite/dolomite type. Formation temperatures of limestone matrix
852
samples were estimated at 20–30 °C based on Shields et al. (2003). Cambrian–Shadow Lake
853
calcite samples are plotted at 70–72 °C and are based on Ttrap values for stage II fluid inclusions
854
in these units (see Section 5.2; Table 3S – Supporting Material). Calcite vein and vug samples
855
from the Gull River–Coboconk formations and the Trenton Group are plotted using Ttrap
856
temperatures of 54–61 °C and 65–78 °C, respectively. For saddle and vein dolomite, Ttrap
857
temperatures of 88–128 °C were used. No fluid inclusion data were obtained for dolostone and
858
replacement dolomite samples but fluid estimates at 54–78 °C (based on calcite veins/vugs) and
859
88–128 °C (based on saddle dolomite) are shown by dashed boxes. The O-isotope composition
860
of late-Ordovician seawater is indicated by the shaded bands between −2 and −4‰ (Shields et
861
al., 2003).
AC C
EP
TE D
M AN U
SC
847
862 863
References
864
Al, T.A., Clark, I.D., Kennell, L., Jensen, M., Raven, K.G., 2015. Geochemical evolution and
39
ACCEPTED MANUSCRIPT
865
residence time of porewater in low-permeability rocks of the Michigan Basin, Southwest
866
Ontario. Chem. Geol. 404, 1–17. doi:10.1016/j.chemgeo.2015.03.005
867
Allan, J.R., Wiggins, W.D., 1993. Dolomite reservoirs: geochemical techniques for evaluating origin and distribution. American Association of Petroleum Geologists; Continuing
869
Education Course Note Series, 36.
870
RI PT
868
Altmann, S., Tournassat, C., Goutelard, F., Parneix, J.-C., Gimmi, T., Maes, N., 2012. Diffusiondriven transport in clayrock formations. Appl. Geochem. 27, 463–478.
872
doi:10.1016/j.apgeochem.2011.09.015
875 876 877
M AN U
874
Armstrong, D.K. and Carter, T.R. 2010. The Subsurface Paleozoic Stratigraphy of Southern Ontario. Ontario Geological Survey, Special Volume 7
Arthur, M.A., Cole, D.R., 2014. Unconventional Hydrocarbon Resources: Prospects and Problems. ELEMENTS 10, 257–264. doi:10.2113/gselements.10.4.257 Ayalon, A., Longstaffe, F.J., 1988. Oxygen isotope studies of diagenesis and pore-water
TE D
873
SC
871
evolution in the Western Canada sedimentary basin; evidence from the Upper Cretaceous
879
basal Belly River Sandstone, Alberta. Journal of Sedimentary Research 58, 489–505.
880
doi:10.1306/212F8DCD-2B24-11D7-8648000102C1865D
EP
878
Beauheim, R.L., Roberts, R.M., Avis, J.D., 2014. Hydraulic testing of low-permeability Silurian
882
and Ordovician strata, Michigan Basin, southwestern Ontario. Journal of Hydrology 509,
883
163–178. doi:10.1016/j.jhydrol.2013.11.033
884 885
AC C
881
Benson, S.M., Cole, D.R., 2008. CO2 Sequestration in Deep Sedimentary Formations. ELEMENTS 4, 325–331. doi:10.2113/gselements.4.5.325
886
Bottomley, D.J., Clark, I.D., 2004. Potassium and boron co-depletion in Canadian Shield brines:
887
evidence for diagenetic interactions between marine brines and basin sediments. Chem.
40
ACCEPTED MANUSCRIPT
888
Geol. 203, 225–236. doi:10.1016/j.chemgeo.2003.10.010 Bottomley, D.J., Katz, A., Chan, L.H., Starinsky, A., Douglas, M., Clark, I.D., Raven, K.G.,
890
1999. The origin and evolution of Canadian Shield brines: evaporation or freezing of
891
seawater? New lithium isotope and geochemical evidence from the Slave craton. Chem.
892
Geol. 155, 295–320. doi:10.1016/S0009-2541(98)00166-1
RI PT
889
Clark, I.D., Al, T., Jensen, M., Kennell, L., Mazurek, M., Mohapatra, R., Raven, K.G., 2013.
894
Paleozoic-aged brine and authigenic helium preserved in an Ordovician shale aquiclude.
895
Geology 41, 951–954. doi:10.1130/G34372.1
SC
893
Clark, I.D., Ilin, D., Jackson, R.E., Jensen, M., Kennell, L., Mohammadzadeh, H., Poulain, A.,
897
Xing, Y.P., Raven, K.G., 2015. Paleozoic-aged microbial methane in an Ordovician shale
898
and carbonate aquiclude of the Michigan Basin, southwestern Ontario. Organic
899
Geochemistry 83-84, 118–126. doi:10.1016/j.orggeochem.2015.03.006 Clark, I.D., Mohapatra, R., Mohammadzadeh, H., Kotzer, T., 2010a. Porewater and Gas
TE D
900
M AN U
896
901
Analyses in DGR-1 and DGR-2 Core. NWMO Technical Report TR-07-21. Nuclear Waste
902
Management Organization (NWMO), Toronto, pp. 25. https://www.nwmo.ca/en/Reports Clark, I.D., Liu, I., Mohammadzadeh, H., Zhang, P., Wilk, M., 2010b. Porewater and Gas
EP
903
Analyses in DGR-3 and DGR-4 Core. NWMO Technical Report TR-08-19. Nuclear Waste
905
Management Organization (NWMO), Toronto, pp. 29. https://www.nwmo.ca/en/Reports
906
Coniglio, M., Williams-Jones, A.E., 1992. Diagenesis of Ordovician Carbonates From the North-
AC C
904
907
East Michigan Basin, Manitoulin Island Area, Ontario - Evidence From Petrography, Stable
908
Isotopes and Fluid Inclusions. Sedimentology 39, 813–836. doi:10.1111/j.1365-
909
3091.1992.tb02155.x
910
Coniglio, M., Sherlock, R., Williams Jones, A.E., Middleton, K., Frape, S.K., 1994. Burial and
41
ACCEPTED MANUSCRIPT
Hydrothermal Diagenesis of Ordovician Carbonates from the Michigan Basin, Ontario,
912
Canada, in: Purser, B., Tucker, M., Zenger, D. (Eds.), Dolomites a Volume in Honour of
913
Dolomieu. Special Publications of the International Association of Sedimentologists 21.
914
Blackwell Publishing Ltd., Oxford, UK, pp. 231–254. doi:10.1002/9781444304077.ch14
915
Coplen, T.B., Hopple, J.A., Boehike, J.K., Peiser, H.S., Rieder, S.E., Krouse, H.R., Rosman,
RI PT
911
K.J.R., Ding, T., Vocke, R.D., Jr., Révész, K.M., Lamberty, A., Taylor, P., De Bièvre, P.,
917
2002. Compilation of minimum and maximum isotope ratios of selected elements in
918
naturally occurring terrestrial materials and reagents. U.S. GEOLOGICAL SURVEY Water
919
Resources Investigation Report 01-4222.
M AN U
SC
916
920
Davies, G.R., Smith, L.B., Jr., 2006. Structurally controlled hydrothermal dolomite reservoir
921
facies: An overview. American Association of Petroleum Geologists Bulletin 90, 1641–
922
1690. doi:10.1306/05220605164
Davis, D.W., 2013. Application of U-Pb Geochronology Methods to the Absolute Age
TE D
923
Determination of Secondary Calcite. NWMO Technical Report TR-2013-21. Nuclear Waste
925
Management Organization (NWMO), Toronto, pp. 31. https://www.nwmo.ca/en/Reports
926
Davis, D.W., Sutcliffe, C.N., Smith, P., Zajacz, Z., Thibodeau, A.M., Spalding, J., Schneider, D.,
EP
924
Adams, J., Cruden, A. and Parmenter, A., 2017. Secondary calcite as a useful U-Pb
928
geochronometer: An example from the Paleozoic sedimentary sequence in southern Ontario.
929
GAC-MAC 2017 Joint Annual Meeting, Kingston, ON, Canada, May 14–18. Abstracts
930
Volume.
931
AC C
927
Diamond, L.W., 2003. Systematics of H2O inclusions. In: Samson, I.M., Anderson, A.J.,
932
Marshall, D.D. (Eds.), Short Course 32. Fluid Inclusions: Analysis and Interpretation. Short
933
Course 32. Mineralogical Association of Canada, pp. 55-79.
42
ACCEPTED MANUSCRIPT
934
Diamond, L.W., Aschwanden, L. and Caldas, R., 2015. Fluid inclusion study of core and outcrop samples from the Bruce Nuclear Site, Southern Ontario, Canada. NWMO Technical Report
936
TR-2015-24. Nuclear Waste Management Organization (NWMO), Toronto, pp. 138.
937
https://www.nwmo.ca/en/Reports
938
RI PT
935
Douglas, M., Clark, I.D., Raven, K., Bottomley, D., 2000. Groundwater mixing dynamics at a Canadian Shield mine. Journal of Hydrology 235, 88–103. doi:10.1016/S0022-
940
1694(00)00265-1
941
SC
939
Frape, S.K., Fritz, P., McNutt, R.H., 1984. Water-rock interaction and chemistry of groundwaters from the Canadian Shield. Geochim. Cosmochim. Acta 48, 1617–1627. doi:10.1016/0016-
943
7037(84)90331-4
945 946 947
Friedman, I., O'Neil, J.R., 1977. Compilation of stable isotope fractionation factors of geochemical interest. USGS Professional Paper 440-KK.
Fritz, P., Frape, S.K., 1982. Saline groundwaters in the Canadian Shield — A first overview.
TE D
944
M AN U
942
Chem. Geol. 36, 179–190. doi:10.1016/0009-2541(82)90045-6 Gascoyne, M., 2004. Hydrogeochemistry, groundwater ages and sources of salts in a granitic
949
batholith on the Canadian Shield, southeastern Manitoba. Appl. Geochem. 19, 519–560.
950
doi:10.1016/S0883-2927(03)00155-0 Girard, J.-P., Barnes, D.A., 1995. Illitization and Paleothermal Regimes in the Middle
AC C
951
EP
948
952
Ordovician St. Peter Sandstone, Central Michigan Basin: K-Ar, Oxygen Isotope, and Fluid
953
Inclusion Data. American Association of Petroleum Geologists Bulletin 79, 49–69.
954 955 956
Goldstein, R.H., 2001. Fluid inclusions in sedimentary and diagenetic systems. Lithos 55, 159– 193. doi:10.1016/S0024-4937(00)00044-X Goldstein, R.H., Reynolds, T.J., 1994. Systematics of fluid inclusions in diagenetic minerals.
43
ACCEPTED MANUSCRIPT
957
SEPM (Society for Sedimentary Geology) Short Course 31. Tulsa, U.S.A. Haeri-Ardakani, O., Al-Aasm, I., Coniglio, M., 2013. Fracture mineralization and fluid flow
959
evolution: an example from Ordovician-Devonian carbonates, southwestern Ontario,
960
Canada. Geofluids 13, 1–20. doi:10.1111/gfl.12003
RI PT
958
961
Hanor, J.S., 2001. Reactive transport involving rock-buffered fluids of varying salinity.
962
Geochim. Cosmochim. Acta 65, 3721–3732. doi:10.1016/S0016-7037(01)00703-7
Harper, D.A., Longstaffe, F.J., Wadleigh, M.A., McNutt, R.H., 1995. Secondary K-feldspar at
964
the Precambrian–Paleozoic unconformity, southwestern Ontario. Can. J. Earth Sci. 32,
965
1432–1450. doi:10.1139/e95-116
M AN U
966
SC
963
Hendry, M.J., Solomon, D.K., Person, M., Wassenaar, L.I., Gardner, W.P., Clark, I.D., Mayer, K.U., Kunimaru, T., Nakata, K., Hasegawa, T., 2015. Can argillaceous formations isolate
968
nuclear waste? Insights from isotopic, noble gas, and geochemical profiles. Geofluids 15,
969
381–386. doi:10.1111/gfl.12132
TE D
967
970
Holser, W.T., 1979. Mineralogy of Evaporites, in: Burns, R.G. (Ed.), Marine Minerals. Reviews
971
in Mineralogy and Geochemistry, Vol. 6. Mineralogical Society of America. pp. 295–346. Horita, J., 2014. Oxygen and carbon isotope fractionation in the system dolomite-water-CO2 to
EP
972
elevated temperatures. Geochim. Cosmochim. Acta 129, 111–124.
974
doi:10.1016/j.gca.2013.12.027
975 976 977
AC C
973
Hurley, N.F., Budros, R., 1990. Albion-Scipio and Stoney Point Fields-U.S.A., Michigan Basin. American Association of Petroleum Geologists Special Volumes 23, 1–37. Intera, 2011. Descriptive Geosphere Site Model. NWMO Technical Report TR-2011-24. Nuclear
978
Waste Management Organization (NWMO), Toronto, pp. 426.
979
(http://www.nwmo.ca/dgrgeoscientificsitecharacterization).
44
ACCEPTED MANUSCRIPT
980 981
Khader, O., Novakowski, K., 2014. The study of diffusion dominant solute transport in solid host rock for nuclear waste disposal. Report to: Canadian Nuclear Safety Commission. Knauth, L.P., Beeunas, M.A., 1986. Isotope geochemistry of fluid inclusions in Permian halite
983
with implications for the isotopic history of ocean water and the origin of saline formation
984
waters. Geochim. Cosmochim. Acta 50, 419–433. doi:10.1016/0016-7037(86)90195-X
985
Kurtis Kyser, T., Kerrich, R., 1990. Geochemistry of fluids in tectonically active crustal regions.
RI PT
982
Short Course on Fluids in Tectonically Active Regimes of the Continental Crust,
987
Mineralogical Association of Canada, Short Course Handbook 18, 133–230. Legall, F.D., Barnes, C.R., Macqueen, R.W., 1981. Thermal Maturation, Burial History and
M AN U
988
SC
986
989
Hotspot Development, Paleozoic Strata of Southern Ontario - Quebec, from Conodont and
990
Acritarch Colour Alteration Studies. Bulletin of Canadian Petroleum Geology 29, 492–539. Ma, L., Castro, M.C., Hall, C.M., 2009. Atmospheric noble gas signatures in deep Michigan
992
Basin brines as indicators of a past thermal event. Earth Planet. Sci. Lett. 277, 137–147.
993
doi:10.1016/j.epsl.2008.10.015
994
TE D
991
Mazurek, M., de Haller, A., 2017. Pore-water evolution and solute-transport mechanisms in Opalinus Clay at Mont Terri and Mont Russelin (Canton Jura, Switzerland). Swiss J Geosci
996
110, 129–149. doi:10.1007/s00015-016-0249-9 McNutt, R.H., Frape, S.K., Fritz, P., 1984. Strontium isotopic composition of some brines from
AC C
997
EP
995
998
the Precambrian Shield of Canada. Chem. Geol. 46, 205–215. doi:10.1016/0009-
999
2541(84)90190-6
1000
Middleton, K., Coniglio, M., Frape, S.K., 1993. Dolomitization of Middle Ordovician Carbonate
1001
Reservoirs, Southwestern Ontario. Bulletin of Canadian Petroleum Geology 41, 150–163.
1002
Neuzil, C. E., and A. M. Provost. 2014. Ice sheet load cycling and fluid underpressures in the
45
ACCEPTED MANUSCRIPT
1003
Eastern Michigan Basin, Ontario, Canada, J. Geophys. Res. Solid Earth, 119, 8748–8769,
1004
doi:10.1002/ 2014JB011643.
1006 1007
Neuzil, C.E., 2015. Interpreting fluid pressure anomalies in shallow intraplate argillaceous formations. Geophys. Res. Lett. 42, 4801–4808. doi:10.1002/2015GL064140
RI PT
1005
Normani, S.D., Sykes, J.F., 2012. Paleohydrogeologic simulations of Laurentide ice‐sheet history on groundwater at the eastern flank of the Michigan Basin. Geofluids 12, 97–122.
1009
doi:10.1111/j.1468-8123.2012.00362.x
1011 1012
NWMO, 2011. Geosynthesis. NWMO Technical Report TR-2011-11. Nuclear Waste Management Organization (NWMO), Toronto, pp. 418. (https://www.nwmo.ca/en/Reports).
M AN U
1010
SC
1008
Oakes, C.S., Bodnar, R.J., Simonson, J.M., 1990. The system NaCl-CaCl2-H2O: I. The ice
1013
liquidus at 1 atm total pressure. Geochim. Cosmochim. Acta 54, 603–610.
1014
doi:10.1016/0016-7037(90)90356-P
Obermajer, M., Fowler, M.G., Snowdon, L.R., 1999. Depositional environment and oil
TE D
1015
generation in Ordovician source rocks from southwestern Ontario, Canada: Organic
1017
geochemical and petrological approach. American Association of Petroleum Geologists
1018
Bulletin 83, 1426–1453.
1020 1021 1022 1023 1024 1025
Ontario Geological Survey, 1991. Bedrock Geology of Ontario, Southern Sheet. Ontario Geological Survey, Map 2544, Scale 1:1,000,000.
AC C
1019
EP
1016
Pearson, F.J., Jr, 1987. Models of mineral controls on the composition of saline groundwaters of the Canadian Shield. Geological Association of Canada - Special Paper 33, 39–51. Powell, T.G., Macqueen, R.W., Barker, J.F., Bree, D.G., 1984. Geochemical character and origin of Ontario oils. Bulletin of Canadian Petroleum Geology 32, 289–312. Roedder, E., 1984. Fluid Inclusions, in: Ribbe, P.E. (Ed.), Reviews in Mineralogy and
46
ACCEPTED MANUSCRIPT
1026
Geochemistry, Vol. 12. Mineralogical Society of America. Reviews in Mineralogy and
1027
Geochemistry, Mineral Soc Am.
1028
Roedder, E., 1979. Fluid inclusion evidence on the environments of sedimentary diagenesis, a review. The Society of Economic Paleontologists and Mineralogists, Special Publication 89–
1030
107.
1033 1034 1035 1036
Annu. Rev. Earth Planet. Sci. 8, 263–301.
SC
1032
Roedder, E., Bodnar, R.J., 1980. Geologic Pressure Determinations from Fluid Inclusion Studies.
Russell, D.J., Gale, J.E., 1982. Radioactive waste disposal in the sedimentary rocks of southern Ontario. Geoscience Canada 9, 200–207.
M AN U
1031
RI PT
1029
Samson, I.M., Anderson, A.J., Marshall, D.D., 2003. Fluid inclusions: analysis and interpretation. Mineralogical Association of Canada, Short Course 32. Saso, J., 2013. Paleohydrogeological Investigation of Upper Ordovician and Cambrian Rocks
1038
from the Michigan Basin in Southwestern Ontario. University of New Brunswick. M.Sc.
1039
Thesis.
1040
TE D
1037
Shafeen, A., Croiset, E., Douglas, P. and Chatzis, I. 2004. CO2 sequestration in Ontario, Canada. Part I: Storage evaluation of potential reservoirs. Energy Conversion and Management 45,
1042
2645–2659.
EP
1041
Shields, G.A., Carden, G.A.F., Veizer, J., Meidla, T., Rong, J.-Y., Li, R.-Y., 2003. Sr, C, and O
1044
isotope geochemistry of Ordovician brachiopods: a major isotopic event around the Middle-
1045
Late Ordovician transition. Geochim. Cosmochim. Acta 67, 2005–2025. doi:10.1016/S0016-
1046
7037(02)01116-X
1047 1048
AC C
1043
Spencer, R.J., 1987. Origin of CaCl brines in Devonian formations, western Canada sedimentary basin. Appl. Geochem. 2, 373–384. doi:10.1016/0883-2927(87)90022-9
47
ACCEPTED MANUSCRIPT
1049
Steele-MacInnis, M., Bodnar, R.J., Naden, J., 2011. Numerical model to determine the composition of H2O–NaCl–CaCl2 fluid inclusions based on microthermometric and
1051
microanalytical data. Geochim. Cosmochim. Acta 75, 21–40. doi:10.1016/j.gca.2010.10.002
1052
Taylor, T.R., Sibley, D.F., 1986. Petrographic and geochemical characteristics of dolomite types
RI PT
1050
1053
and the origin of ferroan dolomite in the Trenton Formation, Ordovician, Michigan Basin,
1054
U.S.A. Sedimentology 33, 61–86. doi:10.1111/j.1365-3091.1986.tb00745.x
Veizer, J., Ala, D., Azmy, K., Bruckschen, P., Buhl, D., 1999. 87Sr/86Sr, δ13C and δ18O
SC
1055
evolution of Phanerozoic seawater. Chem. Geol. doi:10.1016/S0009-2541(99)00081-9
1057
Wickstrom, L.H., Gray, J.D., Stieglitz, R.D., 1992. Stratigraphy, structure, and production
M AN U
1056
1058
history of the Trenton Limestone (Ordovician) and adjacent strata in northwestern Ohio.
1059
Rep. of investigations/State of Ohio. Dep. of natural resources, Div. of geol. survey.
1060
doi:10.1111/j.1502-3931.1996.tb01833.x/full
Wilson, T.P., Long, D.T., 1993a. Geochemistry and isotope chemistry of CaNaCl brines in
TE D
1061 1062
Silurian strata, Michigan Basin, U.S.A. Appl. Geochem. 8, 507–524. doi:10.1016/0883-
1063
2927(93)90079-V
Wilson, T.P., Long, D.T., 1993b. Geochemistry and isotope chemistry of Michigan Basin brines:
1065
Devonian formations. Appl. Geochem. 8, 81–100. doi:10.1016/0883-2927(93)90058-O Ziegler, K., Longstaffe, F.J., 2000. Multiple Episodes of Clay Alteration at the
AC C
1066
EP
1064
1067
Precambrian/Paleozoic Unconformity, Appalachian Basin: Isotopic Evidence for Long-
1068
Distance and Local Fluid Migrations. Clays and Clay Minerals 48, 474–493.
1069
doi:10.1021/ac60068a025
1070
48
ACCEPTED MANUSCRIPT
Table 1. Microthermometric results for fluid inclusions in the secondary minerals Sample ID
Formation
mBGS
Matrix
T m(ice) or T m(hydrohalite)d (ºC)
T trap (ºC)
Mineral I
II
IVb
III
V
VI
I
[77–96], [106–127]
<70
[57–77] [80–100] [61–78]
<70 <70 <70
48
II III IV min max min max min max –46.9 –40.9 –45.2 –44.7 –21.5 –18.2 –36.4 –33.2 –40.7 –39.0 –45.0 –34.0 –47.0 –42.0 –44.2 –35.9 –46.5 –35.2 –46.8 –42.1
[86–107] [94–113]
<70
59
–22.9 –18.4 –43.1 –40.2
Sherman Fall Coboconk Coboconk Coboconk Coboconk Coboconk Coboconk Gull River Gull River
691.70 764.65 767.81 769.57 778.29 778.44 799.15 818.36 829.53
cal-py dol dol cal-py dol
cal cal cal cal-cls cal
dol-py cal dol-py
dol (SD) cal cal
122a
dol-py
dol (SD)
125, 128a
65, 78 61 57 54
71
42
SC
DGR-2-691.70 DGR-6-883.75 DGR-6-886.91 DGR-4-769.57 DGR-6-892.99 DGR-2-778.44 DGR-3-799.15 DGR-4-818.36 DGR-4-829.53
max
RI PT
min
Vc min
max
–38.8 –44.1
–44.1 –42.7
M AN U
DGR-2-844.31 Cambrian 844.31 55 <70 dol-py dol-cal [75–96], [77–97] –42.7 –41.5 –40.3 –37.4 –43.9 –42.6 –35.0 –34.4 88a qtz cal [91–112], [94–115] –42.9 –41.8 –52.1 –50.5 DGR-2-847.42 Cambrian 847.42 80 qtz cal [88–109], [91–112] DGR-2-850.01 Cambrian 850.01 72 <70 –42.6 –41.5 –52.1 –49.5 –34.2 –32.1 qtz cal –47.9 –42.6 DGR-2-853.72 Cambrian 853.72 70 [102–123], [104–125] <70 –52.7 –50.3 –25.9 –24.3 Notes: Microthermometric measurements were made in calcite and dolomite (a), and are indicated by the Mineral column. Relative timing of fluid stages I–VI was determined petrographically based cross-cutting relationships between fluid inclusion assemblages. (b) T trap for stage IV can only be bracketed between T h (minimum possible value) and a maximum T calculated from an assumed lithostatic pressure of 1860 m (see text for explanation).
AC C
EP
TE D
T m(ice) values in stages I–IV represent the metastable equilibrium between halite+ice+liquid+vapour. (c) T m(ice) for stage V was obtained by artificially stretching the inclusions. (d) Hydrohalite T m values are shown in italics. Abbreviations: cal – calcite; cls – celestine; dol – dolomite; py – pyrite; SD – saddle dolomite; qtz – quartz.
ACCEPTED MANUSCRIPT
Stratigraphy Pleistocene Lucas Amherstburg
Karstic Aquifer
Bois Blanc
RI PT
100
Devonian
0
Hydrostratigraphy
Bass Islands
Silurian
Silurian Aquitard A1 Aquifer
Guelph Goat Island Gasport/Lions Head/ Fossil Hill Cabot Head Manitoulin Queenston
400
500
Georgian Bay
600 Blue Mountain Collingwood Cobourg Sherman Fall Kirkfield Coboconk
800
Gull River Shadow Lake Cambrian
900
Precambrian
Trenton Group
700
Guelph Aquifer
Ordovician Aquiclude
Black River Group
Depth (mBGS)
Salina
300
Ordovician
AC C
EP
TE D
M AN U
SC
200
Black River Group Aquitard Basal Aquifer
150
Temperature (°C)
125 100 75 50
–2
0
+4
+8
–4 –6
Trenton Group
ACCEPTED MANUSCRIPT
Cambrian–Shadow Lake
–10
–14
RI PT
A
Gull River– Coboconk
SC
25
M AN U
All Limestones
0 10.0
14.0
18.0
22.0
26.0
30.0
B
150
–4
–2
TE D
δ 18OVSMOW(‰) of calcite 0
+4
+8
100
–6
AC C
Temperature (°C)
125
EP
Saddle Dolomite
–10
Matrix Replacement and Dolostone
75 –14
50 25 0 10.0
14.0
18.0
22.0
26.0
δ 18OVSMOW(‰) of dolomite
30.0
A
ACCEPTED MANUSCRIPT
B
C
Dol
RDol Anh
Anh
Cal vein
RI PT
Dol vein SDol
Cal vein
Anh
2 mm
E
SDol
Anh
SDol
1 mm
Py
Organics+Mag
Cal
Dol vein
Cal Dol 500 µm
EP
SDol
AC C
H
Dol
Qtz
F
Dol
1 mm
G
TE D
Dol
4-829.53
SDol
reimage SDol
SC
D
2 mm
M AN U
2 mm
Dol
Cal vug
I
SDol
Cal vug
SDol SDol
Cal vein Dol
Dol 500 µm
SDol
1 mm
1 mm
Qtz
Fluid Stages
I
II
III
IV
ACCEPTED MANUSCRIPT
V
VI
Secondary Minerals Dolomite - saddle and veins
RI PT
Authigenic Quartz Calcite
SC
- veins and vugs
Anhydrite
M AN U
Celestine Pyrite Magnetite
EP
Fluids
TE D
Bitumen
AC C
Halite sat. brine I + CH4
Halite sat. brine II + CH4 Halite sat. brine III + CH4 Halite sat. brine IV Brine V Oil + CH4 early
late
650
A
B
C
ACCEPTED MANUSCRIPT
650
650
Calcite Veins and Vugs
RI PT
700 700
Dolomite Dolostone Matrix Matrix Replacement
M AN U
Anhydrite in Veins and Vugs
750
TE D
Depth (mBGS)
SC
Saddle and Veins
AC C
EP
800 800
850 0
20
40
60
80
Dolomite wt.% (Relative to Calcite)
100
0
4
8
12
MgO (wt.%)
16
20
24
0
1000
2000
3000
Sr (ppm)
4000
5000
ACCEPTED MANUSCRIPT
B
A
C
SDol
B
SDol
RI PT
C
Stage I
SC
D
M AN U
E
h) H
Anh
EP
F
Cal
Stage II
Stage II
H
I
Anh
Stage III
Stage I
Cal
AC C
G
TE D
Cal
Stage II
SDol
Stage I
Anh
Stage III
Stage III
A
ACCEPTED MANUSCRIPT
B
C
Cal
RI PT
B
Cal Stage IV
SC
E
Cal
M AN U
D
Cal Stage IV
Stage IV
F
Cal
E
Cal G
H
Cal
Cal
I
AC C
G
Cal Cal
Stage V
Stage V
Stage IV
EP
Stage IV
TE D
Uniform vapour fractions
ACCEPTED MANUSCRIPT Primary inclusions
650
Stage I (dolomite) Stage II (calcite) Stage III (cls+anh)
Stage IV Stage V Stage VI (oil+gas)
M AN U AC C
800 800
TE D
750
EP
Depth (mBGS)
SC
RI PT
700 700
Secondary inclusions
850 25
50
75
100
Temperature (Ttrap, °C)
125
A Ca Cl 2
H 2O
Ma ss%
10
10 Ice + L + V
te ) i s
Metastable cotectic Ice + Halite + L + V 30
otherms
SC
h a li T m(
Hydrohalite + L + V
RI PT
therms
°C 42 °C 70 °C 6 10 8 °C 12
40
Halite + L + V
M AN U
40
20
Tm(ice) iso
30 Eutectic -52 °C Peritectic -22.4 °C
Cl Na
20
-24 °C -26 °C -32 °C -35 °C
% ss Ma
ACCEPTED MANUSCRIPT
Antarcticite + L + V
CaCl2
TE D
B
NaCl
Ca
ss%
Ma
Ice + L + V
20
Cl
Na
AC C
10
%
10
ss
Ma
EP
Cl
2
H 2O
20
Stage I Stage II Stage III Stage IV Stage V
? ?
30
40
CaCl2
30
Halite + L + V
40
NaCl
650
A
B
C
ACCEPTED MANUSCRIPT
650
Calcite Limestone Matrix Veins and Vugs Dolomite Dolostone Matrix Matrix Replacement
RI PT
700
Saddle and Veins Anhydrite and Celestine in Veins and Vugs
M AN U
Late Ordovician Seawater
750
TE D
Depth (mBGS)
SC
700
AC C
EP
800 800
Precambrian Shield Brines
Late Ordovician Seawater
850
Cambrian Seawater
Cambrian Seawater
-8.0
-6.0
-4.0
-2.0
0.0
δ 13CVPDB(‰)
2.0
4.0
10.0
14.0
18.0
22.0
δ 18OVSMOW(‰)
26.0
30.0 0.707
0.708
0.709 87
Sr/ 86 Sr
0.710
0.711
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Highlights 1. Multi-stage paleofluid history recorded by primary and secondary fluid inclusions 2. Primary stages I to III formed the dolomite and calcite veins 3. Secondary fluid stages IV to VI record trapping of local and re-mobilized fluids 4. Mineral isotope data indicate variable fluid sources for the veins 5. Sources include connate fluids and mixed connate–shield-type brines