Chemical Geology 287 (2011) 27–40

Contents lists available at ScienceDirect

Chemical Geology

j ourna l homepage: www.e lsev ie r.com/ locate /chemgeo

Research papers

Constraining the timing of microbial methane generation in an organic-rich shaleusing noble gases, Illinois Basin, USA

Melissa E. Schlegel a,⁎, Zheng Zhou b, Jennifer C. McIntosh a, Chris J. Ballentine b, Mark A. Person c

a Department of Hydrology and Water Resources, University of Arizona, Tucson, AZ 85721, USAb Department of Earth Sciences, University of Manchester, Oxford Road, Manchester, M13 9PL, United Kingdomc Department of Earth and Environmental Science, New Mexico Tech, 801 Leroy Place, Socorro, NM 87801, USA

⁎ Corresponding author. Tel.: +1 801 380 5145; fax:E-mail addresses: meschleg@email.arizona.edu (M.E

Zheng.Zhou@manchester.ac.uk (Z. Zhou), mcintosh@hwChris.Ballentine@manchester.ac.uk (C.J. Ballentine), mar(M.A. Person).

0009-2541/$ – see front matter © 2011 Elsevier B.V. Aldoi:10.1016/j.chemgeo.2011.04.019

a b s t r a c t

a r t i c l e i n f o

Article history:Received 13 December 2010Received in revised form 21 April 2011Accepted 30 April 2011Available online 14 May 2011

Editor: J.D. Blum

Keywords:Microbial methanogenesisNoble gasesPleistoceneGroundwater rechargeBasinal brineSedimentary basins

At least 20% of the world's natural gas originates from methanogens subsisting on organic-rich coals andshales; however in-situ microbial methane production rates are unknown. Methanogens in the UpperDevonian New Albany Shale in the Illinois Basin extract hydrogen from low salinity formation water to formeconomic quantities of natural gas. Because of this association, constraining the source and timing ofgroundwater recharge will enable estimation of minimum in-situ metabolic rates. Thirty-four formationwater and gas samples were analyzed for stable isotopes (oxygen and hydrogen), chloride, tritium, 14C, andnoble gases. Chloride and δ18O spatial patterns reveal a plume of water with low salinity (0.7 to 2154 mM)and δ18O values (−0.14 to −7.25‰) penetrating ~1 km depth into evapo-concentrated brines parallel toterminal moraines of the Laurentide Ice Sheet, suggesting glacial mediated recharge. However, isotopicmixing trends indicate that the recharge endmember (~−7‰ δ18O) is higher than the assumed bulk ice sheetvalue (b−15‰ δ18O), and similar to modern local precipitation (−7.5 to −4.5‰ δ18O). Continentalpaleoprecipitation records reveal that throughout the Pleistocene, δ18O of precipitation in the region rangedfrom −10 to −5‰, suggesting that the dilute groundwater was primarily sourced from paleoprecipitationwith minor contributions from glacial meltwater.For the majority of samples the atmosphere derived 4He contribution is negligible, and the 4He is dominatedby a crustal radiogenic source, with near complete transfer of dissolved noble gases to the gas phase. Inaddition, mantle derived helium is negligible for all samples (b1%). Helium-4 ages of formation watersassociated with natural gas accumulations range from 0.082 to 1.2 Ma. Thermogenic methane is associatedwith older fluids (average 1.0 Ma), as compared to microbial methane (average 0.33 Ma), consistent withchloride and δ18O data. However, all groundwater in the study area was influenced by Pleistocene recharge.Estimated in-situ microbial methane production rates range from 10 to 1000 TCF/Ma — ~104 to 106 timesslower than average laboratory rates from coals. Findings from this study have implications for targetingundeveloped microbial gas accumulations, improving natural gas reservoir estimates, the potential of in-situmethanogen stimulation, and understanding biologic cycling of carbon in subsurface reservoirs.

+1 520 621 1422.. Schlegel),r.arizona.edu (J.C. McIntosh),kaustinperson@gmail.com

l rights reserved.

© 2011 Elsevier B.V. All rights reserved.

1. Introduction

Methanogens have generated economic quantities of natural gas insubsurface organic-rich units such as coals and black shales. Currentlythis microbial methane supports at least 5% of United States' andCanada's energy needs (Rice, 1992; OEE, 2008; EIA, 2010), howeverlittle is known about the origins, residence times, and productionrates of in-situ microbial communities (e.g. Schlegel et al., 2011).These knowledge gaps are critical in investigating the potential of

in-situ methanogen stimulation and carbon cycling in biologically-active reservoirs.

In the Illinois Basin, economic reservoirs of microbial methaneoccur in the Upper Devonian New Albany Shale in association withanomalously low salinity formation water (McIntosh et al., 2002). Alinear correlation between the δD values of the water and the co-produced methane demonstrates that methanogens extract hydrogenfrom the low salinity water to form methane (Schoell, 1980; Martiniet al., 1998). The anomalously low salinity water extends from thebasin margin to ~1 km depth in the shale and has a pattern divergentfrom simple infiltration from the basin margin subcrop (McIntoshet al., 2002). Rather, the orientation of the plume corresponds toPleistocene ice lobes that advanced and retreated into the basinmultiple times (Killey, 1998; McIntosh et al., 2002), suggesting glacialmeltwater recharge from either subglacial recharge or proglacial

28 M.E. Schlegel et al. / Chemical Geology 287 (2011) 27–40

lakes. However the δ18O values associated with the dilute plume(−7.5 to +1.1‰) are not as low as would be expected for glacialrecharge (b−15‰ δ18O), and mixing trends indicate a relatively highδ18O value (~−7‰) for the freshwater endmember (McIntosh et al.,2002).

Isotopically-enriched recharge waters may have been sourcedfrom: 1) modern precipitation (−11.0 to −4.5‰ δ18O), 2) a mixtureof isotopically depleted Pleistocene-aged recharge and formationalbrines, or 3) an isotopically enriched Pleistocene aged source. Thisstudy constrains the source and timing of recharge in the New AlbanyShale by comparing calculated 4He groundwater ages derived fromproduced natural gas samples with stable isotopes (oxygen andhydrogen), chloride, tritium, and 14C values from the produced water.This data set is enhanced by including published paleoclimate data forthe Illinois Basin area. Results demonstrate that climate change canhave profound effects on basin-scale groundwater recharge, drivingfreshwater recharge deep into sedimentary basins, which may alsostimulate microbial methane generation in the relatively recentgeologic past (b2 Ma), enabling the estimation of in-situ microbialmethane production rates.

2. Geologic setting

The Illinois Basin, located in the midcontinent of the United States,is an interior sedimentary basin centered in southern Illinois (Fig. 1).The basin is oblong NW–SE, ~620 km long, ~375 km wide, over

Fig. 1. Location map of the Illinois Basin, with New Albany Shale subcrop (black) and terminastructural features and sample locations for this study. The cross-section through the Illino

4.5 km in depth, and filled with shallowly dipping (6 to 14 m/km)Paleozoic sediments (Zuppann et al., 1988). The basal sedimentaryunits are a thick sequence of Cambrian–Ordovician sandstones andcarbonates overlying the Precambrian basement (Swann, 1968). LateOrdovician to Mississippian units are composed primarily of carbon-ates with interbedded sandstones and shales (Swann, 1968). LateMississippian to Pennsylvanian units are generally comprised ofinterbedded shale, carbonates, and coal (Swann, 1968). Topping thesedimentary sequence are the surficial Quaternary glacial sedimentsthat form the basin topography, with less than 300 m of relief (Swann,1968; Zuppann et al., 1988). The basin is bounded to the north by theKankakee Arch andWisconsin Arch, to the east by the Cincinnati Arch,to the southwest by the Ozark Dome and Pascola Arch, and to thewestby the Mississippi River Arch (Buschbach and Kolata, 1990). TheIllinois Basin was deformed along its southern margin during the latePaleozoic through Mesozoic by tectonic uplift of the Pascola Arch anddownwarping of the Mississippian Embayment (Buschbach andKolata, 1990). The Cottage Grove and Rough Creek–ShawneetownFault System, which trends east–west through the south-centralbasin, was active throughout the Paleozoic, and extends from thePrecambrian basement to the surface (Buschbach and Kolata, 1990).Hydrocarbons in the basin are dominated by thermogenic processesin the deeper central portion of the basin in western Kentucky andsouthern Indiana, while microbial methane is predominant inDevonian shales and Pennsylvanian coals closer to the basin margins(Figs. 2 and 3) (McIntosh et al., 2002; Strąpoć et al., 2007).

l moraines forWisconsin and Illinoian glaciations (dotted lines). Also marked are majoris Basin identifies important hydrologic units.

Fig. 2. Conceptual model of possible effects of ice sheet loading on groundwater recharge (white arrows) as compared to topographic recharge (black arrows). In addition, areas ofpredominantly microbial and thermogenic methane are denoted in relation to meteoric recharge.

29M.E. Schlegel et al. / Chemical Geology 287 (2011) 27–40

2.1. Regional fluid flow

Based on similar hydrochemistry and hydrogeologic properties,geologic units in the Illinois Basin have been grouped into regionalaquifers and confining units by Lloyd and Lyke (1995) (Figs. 1 and2). This study focuses on the Upper Devonian New Albany Shale,which is part of the larger Silurian–Devonian carbonate aquifer.Additional regional aquifers are present in the highly permeablebasal Cambrian–Ordovician aquifer, theMississippian–Pennsylvaniancoals and sedimentary units, and the surficial Quaternary glacialdeposits. Formation waters in the four major aquifer systems (Quater-nary, Mississippian–Pennsylvanian, Silurian–Devonian, and Cambrian–Ordovician) have distinct geochemical signatures, suggesting that therehas been little cross-formational fluid mixing in the Illinois Basin,impeded by shales (Stueber and Walter, 1991, 1994; Stueber et al.,1993).

Several researchers have suggested that brines and hydrocarbonsmigrated through the Illinois Basin over 600 km overmillions of years,driven by tectonic uplift of the Ouachitas and Pascola Arch to thesouth and sediment compaction (Bethke, 1985; Bethke et al., 1991).Alternatively, Garven et al. (1999) suggest that topographically-driven flow events persisted for only short periods of geologic time(b2 Ma), long enough to transport metals, hydrocarbons, anddiagenetic fluids, but brief enough to prevent flushing of remnantbrines. Rowan and deMarsily (2001) proposed infiltration of brines inthe central Illinois Basin rift system coupled with topographically-driven flow through the basal Mt. Simon Aquifer during the Permianas a possible explanation for regional-scale fluid flow and world-classlead–zinc ore deposition, while maintaining high salinities. Mem-brane filtration is an unlikely mechanism for brine generation in theIllinois Basin, as observed vertical head differences are not largeenough to cause upward flow through basal confining units (Bond,1972), and observed salinity patterns are incompatible with salinitypatterns predicted by membrane filtration. Also, evaporite dissolutionby meteoric recharge is unlikely for the regional Illinois Basin aquifers(with the exception of theMississippian aquifer), as the stable isotopecomposition and Na–Ca–Cl–Br relations of basin formation waters areindicative of evapo-concentrated seawater, suggesting that thesePaleozoic-age fluids are relatively stagnant (Stueber and Walter,1994).

2.2. Glacial-mediated fluid flow

Continental glaciations, occurring on relatively short geologic timescales (b3 Ma) can provide sufficient hydraulic head to force

freshwater recharge deep into sedimentary basins against strongsalinity gradients (Grasby et al., 2000; Breemer et al., 2002; McIntoshet al., 2002; Person et al., 2003, 2007; Hoaglund et al., 2004; Ma et al.,2004; Grasby and Chen, 2005; McIntosh and Walter, 2006; McIntoshet al., 2011). Kilometer-thick ice sheets advanced and retreatedmultiple times over the Illinois Basin during the last 3 million years,increasing the hydraulic gradients and dilating natural fracturesystems, which may have enhanced permeability and depth ofrecharge. Concomitantly, permafrost in front of and beneath the icesheet may have inhibited groundwater recharge (Lively et al., 1993;Edmunds and Milne, 2001), promoting high heads in ice free areas,lowering effective stress (Boulton and Caban, 1995), and generatingblowout structures (Christiansen et al., 1982; Christiansen and Sauer,1998) and other ice sheet tectonic features (Pederson, 2005; Person etal., 2007). Permafrost existed around and beneath the southernmargin of the Laurentide Ice Sheet for thousands of years, during theLast Glacial Maximum (Cutler et al., 2000; Larson et al., 2003; Tarasovand Peltier, 2007). In the Illinois Basin between 21 and 16 ka,permafrost reached a southern limit of 38° 30′ N latitude, lasted up toabout 1 ka in any given location, andwas restricted to about 200 km infront of the glacier (Johnson, 1990).

As ice sheets override sedimentary basins, permeability of aquitardunits, such as shale, can be modified by hydrofracturing, which altersflow patterns, facilitating meteoric recharge (Boulton and Caban,1995; Lerche et al., 1997; Bekele et al., 2003). Much of thegroundwater infiltration occurs as the ice sheet moves across thelandscape, withmore recharge during transgressive phases (Person etal., 2007; Lemieux et al., 2008), suggesting that the number of timesthe ice sheet advanced is an important control on deep basin recharge.In the Illinois Basin a minimum estimate of 7 to 15 Pre-Illinoian (2.7 to0.5 Ma) ice sheet advances are recorded (Killey, 1998). During theIllinoian glaciation (300 to 125 ka) ice sheets advanced at least 3 timesinto the Illinois Basin, of which the second advance reached thefarthest south. These advances are generally associated with thinglacial sediment (b50 ft thick) and small end moraines, suggestingthat the ice surged and then melted in place. During the Wisconsinglaciation, ice entered the Illinois Basin approximately 25 ka, andreached the southern extent by about 20 ka. After approximately 6major advances and retreats, the ice had completely retreated fromthe basin by 14 to 13 ka (Killey, 1998; Patterson et al., 2003).

In the Illinois Basin, groundwater with anomalously low δ18Ovalues and salinity concentrations as deep as ~1 km has beensuggested to be remnant Pleistocene ice sheet meltwater fromsubglacial recharge or complete melting of the ice sheet, as well aspaleoprecipitation (Clayton et al., 1966; Siegel and Mandle, 1984;

Fig. 3. Contour maps of (a) chloride concentrations and (b) δ18O values for New AlbanyShale formation waters collected as part of this study and previous studies (McIntosh etal., 2002; Schlegel et al., 2011). The areas of predominantly thermogenic and microbialmethane are also highlighted.

30 M.E. Schlegel et al. / Chemical Geology 287 (2011) 27–40

Siegel, 1991; McIntosh et al., 2002). Groundwater ages of 17 to 50 kawith δ18O values as low as −15.3‰ in shallow Silurian–Devoniancarbonates have been reported byMcIntosh andWalter (2006). Theseobserved dilute groundwaters form plumes along the northernmargin of the basin that correspond to the location of the LaurentideIce Sheet (Fig. 3).

However, some of the dilute groundwaters have δ18O values thatare more positive than would be expected from glacial recharge(e.g. Siegel, 1991, 1992; McIntosh et al., 2002). Siegel (1991) observedhigh δ18O and δD values (−10.3 to −7.2‰, and −77 to −50‰,respectively) in dilute waters along the northwestern margin of theIllinois Basin. Klump et al. (2008) observed relatively high δ18O values(−12.4 to −8.6‰) in Pleistocene-aged groundwater in southeasternWisconsin. Ma et al. (2004) observed high δ18O and δD values(~−8.6‰, and ~−56‰, respectively) in association with cooltemperatures (~1 °C) at the end of the Last Glacial Maximum(~17.2 ka) in southern Michigan. Siegel (1991,1992), Klump et al.(2008), and Ma et al. (2004) suggest that groundwater recharge was

significantly affected by an enrichedmoisture source for precipitation,possibly from the Gulf of Mexico, over the past 30 ka.

Surface water flow systems were also profoundly altered duringPleistocene glaciation. As the ice sheets grew they destroyed previousdrainage systems, and stored large volumes of freshwater on thecontinents. The Laurentide Ice Sheet likely contained greater than20×106 km3 of water (Teller, 1990). As the ice sheets melted,freshwater was discharged into proglacial lakes, and formed newriver systems. Several large lakes formed behind glacial endmorainesas ice sheets retreated, such as the predecessors of the Great Lakes andIllinoian Glacial Lake Quincy in central Indiana, which was over180 km2 and N20 m in depth (e.g. Killey, 1998; Wood et al., 2010).Additionally, large lakes repeatedly formed and catastrophicallydrained as major drainage systems, such as the Mississippian River,were blocked by advancing glaciers (e.g. Killey, 1998). Moreover,riverine fluxes were 2 to 3 times greater than modern run-off(Marshall and Clarke, 1999), and may have significantly contributedto groundwater recharge.

3. Methods

To investigate the source and timing of fresh groundwaterrecharge in the New Albany Shale, 34 groundwater samples werecollected for anions (including Cl− and Br−) and stable isotopeanalyses (18O and D), with 29 produced from the shale and 5produced from the carbonate aquifers. Of the wells produced from theshale, 1 is a domestic drinking water well (MM0), and 21 are naturalgas wells in predominantly microbial methane areas (MM1–MM21),and 7 are natural gas wells in predominantly thermogenic methaneareas (TM1–TM7). Of the wells produced from the carbonate aquifers,1 is in theMississippian carbonate aquifer above the shale (M1), and 4are in the Devonian carbonate aquifer below the shale (D1–4). Asubset of these samples was selected for age tracer analysis, including16 samples for 14C and 24 for noble gas concentrations. Wells frompredominantly thermogenic methane areas did not produce anywater, so only gas analyses were performed. To supplement oursampling data we also incorporated previously published Cl− andstable isotope data for the New Albany Shale (McIntosh et al., 2002;Schlegel et al., 2011).

Allwater andgas sampleswere collecteddirectly from thewell head.Water samples for chloride, bromide, and stable isotopes were filteredwith a 0.45 μm Micropore filter. Chloride and bromide samples werekept in high-density polypropylene (HDPE) bottles, and oxygen andhydrogen isotope samples were kept in glass scintillation vials. Watersamples for 14Cwere collected directly into amber glass bottles withoutfiltration to minimize oxidation affects due to air contamination. Allsample containers were filled with no headspace, immediately placedon ice and refrigerated until analyzed. Noble gas sampleswere collectedfrom produced gases following the method described in Zhou et al.(2005). Gas was collected in copper tubes (refrigeration grade coppertubing 60 cm long by 10 mm diameter) that were flushed withproduced gas for over 10 min prior to being sealed with stainless steelclamps.

Chloride and bromide were analyzed on a Dionex ICS-3000 IonChromatograph using an AS23 column (precision, ±2%). Oxygen andhydrogen stable isotopes (δ18O and δD) of water were measured on aFinnigan Delta S gas-source isotope ratio mass spectrometer, andreported relative to Vienna Standard Mean Ocean Water (VSMOW;precision, ±0.9‰ or better for hydrogen and ±0.08‰ or better foroxygen). After conversion of sample DIC to CO2 and graphitization, 14Cwas measured in an accelerator mass-spectrometer (AMS) in the NSFlab at the University of Arizona Accelerator Mass Spectrometry (AMS)Laboratory (analytical error b0.2 percent modern carbon (pmc)).Noble gas isotope concentrations were measured at the University ofManchester for He, Ne, Ar, Kr and Xe as described by Gilfillan et al.(2008) and Ballentine and Sherwood Lollar (2002). An atmospheric

31M.E. Schlegel et al. / Chemical Geology 287 (2011) 27–40

standardwas run following each sample. Errors presented are 1 σ, andinclude analytical, interference, standard reproducibility, and expan-sion volume uncertainty.

4. Results and discussion

4.1. Chloride patterns

Chloride concentrations in the New Albany Shale range from 0.7 to2154 mM, and average 1242 mM, with the drinking water well(MM0) containing the lowest concentration (Table 1). For thecarbonate aquifers, chloride concentrations range from 92 to1426 mM, and average 402 mM (Table 1). Interestingly, the lowestchloride concentration in the Devonian carbonate aquifer (92 mM)occurs only 33 km to the north of the most saline concentration(1426 mM). Also, chloride concentrations from Mississippian andDevonian aquifers from the same area are comparable (154 and161 mM, respectively), even though they are separated by ~260 mdepth. Chloride concentrations in the Devonian carbonate aquifer andthe New Albany Shale are comparable (1426 mM; 600–1777 mM,respectively) in Sullivan Co., however to the east in Greene Co.chloride concentrations are lower in the Devonian carbonate aquifer(161 mM) as compared to the shale (647 and 799 mM). Data from thisstudy are consistent with previously observed chloride patterns forthe Devonian shale and carbonate aquifers, and the Mississippiancarbonate aquifers (McIntosh et al., 2002; Schlegel et al., 2011).Moreover, the increased resolution from additional data shows

Table 1Well location, anion, and stable isotope data for New Albany Shale and carbonate aquifer fo

ID County State Date Latitude Longitude

New Albany Shale — microbial methaneMM0 Johnson IN 6/10/2008 39.48577 −86.10317MM1 Lawrence IL 1/11/2008 38.80663 −87.59403MM2 Sullivan IN 1/7/2008 38.99722 −87.43775MM3 Sullivan IN 6/11/2008 39.03270 −87.38730MM4 Sullivan IN 7/24/2008 38.99058 −87.38870MM5 Sullivan IN 7/20/2009 39.03985 −87.34112MM6 Greene IN 6/16/2008 39.05813 −87.17053MM7 Greene IN 6/16/2008 39.01330 −87.07337MM8 Jackson IN 6/16/2008 38.80368 −86.17697MM9 Knox IN 5/13/2008 38.60960 −87.26280MM10 Knox IN 7/23/2009 38.84927 −87.18818MM11 Knox IN 7/24/2009 38.84533 −87.29835MM12 Knox IN 7/24/2009 38.85995 −87.33127MM13 Daviess IN 7/22/2008 38.79643 −87.19978MM14 Daviess IN 7/22/2008 38.78663 −87.08445MM15 Daviess IN 7/22/2008 38.80143 −87.10442MM16 Harrison IN 1/9/2008 38.16943 −86.05365MM17 Harrison IN 1/9/2008 38.27033 −86.14218MM18 Harrison IN 6/14/2008 38.21668 −86.17015MM19 Harrison IN 7/22/2009 38.12593 −86.00627MM20 Harrison IN 7/22/2009 38.17898 −86.07378MM21 Meade KY 1/9/2008 38.00955 −86.14582

New Albany Shale — thermogenic methanea

TM1 Christian KY 7/21/2009 37.09985 −87.36493TM2 Christian KY 7/21/2009 37.08467 −87.33920TM3 Christian KY 7/21/2009 37.07038 −87.32985TM4 Christian KY 7/21/2009 37.09007 −87.31628TM5 Christian KY 7/21/2009 37.08343 −87.30585TM6 Christian KY 7/21/2009 37.06717 −87.30533TM7 Christian KY 7/21/2009 37.06773 −87.27803

Carbonate aquifersD1 Clark IL 7/25/2008 39.37293 −87.54643D2 Vigo IN 7/23/2008 39.44033 −87.36192D3 Sullivan IN 7/23/2008 39.15810 −87.50677D4 Greene IN 7/24/2008 38.98133 −87.00743M1 Greene IN 7/24/2008 38.97878 −86.99515

a Thermogenic methane wells did not produce water.

steeper salinity gradients in the shale than previously observed(Fig. 3; McIntosh et al., 2002; Schlegel et al., 2011).

4.2. Oxygen and hydrogen isotopes

The δ18O values of groundwater from the New Albany Shale rangefrom−0.14 to−7.25‰, with an average of−2.85‰ (Table 1). The δDvalues of groundwater from the shale range from −5.9 to −45.7‰,with an average of−17.7‰ (Table 1). For the carbonate aquifers, δ18Ovalues of groundwater range from−2.50 to−6.99‰, with an averageof −5.9‰ (Table 1). The δD values of groundwater for the carbonateaquifers range from −23.9 to −43.8‰, and average −38.9‰(Table 1). For the shale and carbonate aquifers, the lowest isotopevalues correspond with the least saline samples (Fig. 4b).

Contour maps of δ18O values show a pattern of recharge similar tothat exhibited by Cl− concentrationmaps (Fig. 3). A plot of δ18O valuesversus Cl− concentrations in combinationwith a plot of δ18O versus δDvalues show that the recharge endmember in each unit is freshmeteoric water that mixed with evaporative basinal brines (Fig. 4). Inaddition, the oxygen isotope value of the freshwater endmember isapproximately −7‰, which is within the range of modern precipita-tion in the region (−7.5 to −4.5‰; McIntosh et al., 2002). However,groundwaters in confined Silurian–Devonian aquifers, updip of thedeep shale, have δ18O values ranging from −15 to −6‰ andradiocarbon ages ranging from 16 to N30 ka (McIntosh and Walter,2006). Thesefindings indicate that Pleistocene groundwater is presentin the basin, and that deep fluids have diluted basinal brines with an

rmation water samples.

Depthm

Cl−

mMBr−

mMδ18O‰ (VSMOW)

δD‰ (VSMOW)

113 0.70 0.003 −7.25 −45.70857 2154.74 3.18 −1.78 −15.00745 1576.52 2.90 −2.12 −19.06711 600.19 2.03 −4.70 −31.33738 1776.97 2.83 −1.85 −10.22697 1426.70 2.58 −3.24 −17.13564 435.10 1.46 −5.00 −28.88472 646.68 2.02 −4.43 −21.00116 798.80 2.19 −4.59 −26.06789 1949.24 2.30 −3.50 −9.71628 1508.79 2.73 −2.22 −8.00736 1163.85 2.25 −3.02 −14.63743 1533.21 2.64 −1.61 −5.90652 1351.38 2.74 −2.28 −15.15580 1072.79 2.21 −3.18 −20.29582 1182.82 2.63 −3.07 −19.03290 1880.45 2.98 −0.92 −13.10411 1641.84 2.78 −2.29 −14.13265 1465.43 2.89 −0.14 −7.86310 1761.41 3.21 −0.41 −11.13309 1798.00 3.90 −0.44 −11.89233 849.08 1.59 −4.59 −23.01

761805759821759822734

610 181.24 0.37 −6.93 −42.70503 92.30 0.17 −6.99 −43.77725 1426.20 3.31 −2.50 −23.93488 161.23 0.38 −6.59 −40.97229 153.66 0.31 −6.36 −39.56

Fig. 4. Plots of (a) δD and (b) Cl− versus δ18O reveal a general mixing trend between adilute meteoric water endmember with stable isotope values similar to modern meanannual precipitation for the study area, and a basinal brine endmember.

32 M.E. Schlegel et al. / Chemical Geology 287 (2011) 27–40

‘old’ (Pleistocene) freshwater endmember that has a modern stableisotopic signature.

Potential sources of enriched Pleistocene-aged δ18O values forgroundwater recharge include glacial meltwater in either subglacialrecharge or proglacial lakes enriched due to evaporation, mineralinteraction, preferential melting of portions of the glacier with higherδ18O values (such as the basal layer), and seasonal/annual variations inthe δ18O value of glacial meltwater that supplies either subglacialrecharge or proglacial lakes (e.g. Gooseff et al., 2003; Yde andKnudsen,2004). In addition, regional paleoprecipitation may have had δ18Ovalues similar to modern precipitation, significantly enriching themeteoric water available for groundwater recharge (e.g. Siegel, 1991,1992; Klump et al., 2008).

4.2.1. Isotopic enrichment of glacial meltwaterSamples from dilute groundwater in the Illinois Basin deviate from

the meteoric water line around −7‰ δ18O and −46‰ δD. Thoughmeltwater–mineral interactions can increase δ18O values by as muchas 12‰ (Knight, 1987, 1989; Souchez et al., 1990; Gooseff et al., 2003),isotopic trends would produce a δ18O/δD correlation less than the

slope of the meteoric water line (~8) (Souchez et al., 1990), andwould deviate from the meteoric water line near expected bulk icevalues (b−15‰ δ18O and b−110‰ δD). A similar isotopic trendwould be observed for evaporation of glacial surface layers, ormeltwater in proglacial lakes. As this trend is not seen in the shalesamples, the glacier meltwater–mineral interactions or evaporationare not likely to be the source of high δ18O and δD values in the deepIllinois Basin groundwater. Temporal variation in recharge due todifferential melting of the ice sheetmay also increase the δ18O value inthe groundwater relative to the bulk ice value (Jouzel and Souchez,1982; Grootes and Stuiver, 1997; Stichler and Schotterer, 2000;Gooseff et al., 2003, 2006; Yde and Knudsen, 2004), but would not berecorded in our data due to the averaging effects of dispersion ingroundwater recharge. Moreover, preferential melting of the glaciermay not have been the source of the relatively high δ18O values of thefresh groundwater recharge, as the Laurentide Ice Sheet completelymelted, suggesting that the average meltwater for recharge shouldhave a composition that approached the bulk ice value.

Subglacial recharge may have been a source of significantgroundwater recharge (i.e. Lemieux et al., 2008), which may havebeen dominated by basal melting (isotopically enriched due tomeltwater–mineral interactions and partial refreezing of subglacialmeltwater and shallow groundwater; Jouzel and Souchez, 1982;Shouchez et al., 1990) and surface melting (isotopically enriched dueto evaporation/sublimation effects; Taylor et al., 2001; Stichler et al.,2001; Yde and Knudsen, 2004), as observed in an Icelandic glacier(Boulton et al., 2007). However, subglacially precipitated calcite inIowa records an isotopic composition within the range of localmodern precipitation (−9‰ δ18O), suggesting that subglacialrecharge in the Illinois Basin may have been heavily influenced byprecipitation (Siegel, 1991).

4.2.2. Isotopic composition of regional paleoprecipitationPaleoprecipitation with a relatively enriched isotopic composition

may have mixed with glacial meltwater (either subglacially or inproglacial lakes) with a depleted isotopic composition, resulting in agroundwater recharge endmember significantly more positive thanglacial meltwater (Siegel, 1991, 1992; Ma et al., 2004; Klump et al.,2008). This suggests that though the Laurentide Ice Sheet may haveprovided the mechanism of deep groundwater recharge in the NewAlbany Shale and Devonian carbonate reservoirs, the source of rechargewas likely dominated by paleoprecipitation, with possibly some minorcontribution from glacial meltwater. Continental paleoclimate recordsfrom speleothems, ostracods and wood cellulose near the Illinois Basin(Fig. 5) reveal that throughout most of the Pleistocene the estimatedδ18O of precipitation ranged from−10 to−5‰, similar to the modernprecipitation range (Harmon and Schwarcz, 1981; Edwards andMcAndrews, 1989; Dorale et al., 1992, 1998; Rea et al., 1994; Dettmanet al., 1995; Denniston et al., 2001, 2007). Notable exceptions areevident in ostracod records in the northernGreat Lakes during lake levellow stands,when the influence of icemelt due to retreating glacierswasat a maximum (Rea et al., 1994; Dettman et al., 1995). However, theostracod record in Lake Michigan reveals that generally the southernlake contained relatively enriched δ18O values (−8 to−4.5‰), andwasinfluenced primarily by local runoff and precipitation (Rea et al., 1994).Lake Michigan borders the northern margin of the Illinois Basin,indicating that near the recharge area for the New Albany Shale andDevonian carbonate therewas a large volumeofwaterwith a δ18O valuethat approximates modern precipitation values. Interestingly, therecord shows slightly depressed δ18O values during glacial periods,however these values do not approach the expected bulk ice value ofb−15‰ δ18O (Fig. 5). Age tracers were used to constrain the timing ofrecharge to further distinguish recharge source and estimate themaximum age of the onset of microbial methanogenesis, andminimumin-situ metabolic rates.

Fig. 5. Plot of δ18O values versus years before present for precipitation as calculated from speleothem records from Missouri, Iowa, and Kentucky, USA (black diamonds); ostracodrecords from the Great Lakes, USA and Canada (gray squares); and wood cellulose records from southern Ontario, Canada (black triangles) (Harmon and Schwarcz, 1981; Edwardsand McAndrews, 1989; Dorale et al., 1992, 1998; Rea et al., 1994; Dettman et al., 1995; Denniston et al., 2001, 2007). In addition, time ranges when ice sheets were present in theIllinois Basin during Wisconsin, Illinoian, and Pre-Illinoian periods are indicated in gray boxes along the x-axis (with no relation to δ18O values). Note δ18O values ofpaleoprecipitation are only slightly depressed compared to modern mean annual precipitation, during glacial periods.

33M.E. Schlegel et al. / Chemical Geology 287 (2011) 27–40

4.3. Tritium and carbon-14

Carbon-14 is a useful age tracer for groundwaters less than ~30 ka,but can be detected inwaters as old as ~50 ka, whereas tritium is usefulin detecting modern meteoric water (b60 years) addition to deepgroundwaters (Clark and Fritz, 1997). Sixteen samples were analyzedfor 14C, with two duplicates (Table 2). Eleven results are from wellsproducing from the shale (113 to 857 m), and are low (0.23 to1.49 pmc), with the exception of wells MM7 and MM8 (15.08 and22.61 pmc, respectively). Four results are fromwells producing from theDevonian carbonate (D1–4) and are also low (0.41 to 1.70 pmc), andone is from a well producing from the Mississippian carbonate (M1,0.23 pmc). The majority of the 14C results range from 0.23 to 1.49 pmc,and are near the limit of detection (b0.2 pmc) (Linick et al., 1986). The‘live’ carbon in the system canbediluted by the addition of ‘dead’ carbonto solution, and can come from two sources: inorganic and organiccarbon dissolution. In the shale, carbonate dissolution aswell as organiccarbon inputs from microbial degradation of fossil carbon have addedsubstantial amounts of ‘dead’ carbon to solution (Geyh and Kuenzl,1981).However the results are above theanalytical error, demonstratingthat there is 14C in the sample (Clark and Fritz, 1997).

Five tritium samples were analyzed from various depths andaquifers to determine if the deep groundwater samples had beencontaminated with surface fluids (Table 2). Results ranged fromb0.8 TU in sample M8 to 2.8 TU in sample M7. Detectable tritiumwithlow 14C values indicate that either 1) relatively large amounts of U, Th,and Li in the subsurface have resulted in elevated lithogenic tritiumlevels in ‘old’ water; and/or 2) a small fraction of water that hasrecently been exposed to the atmosphere has mixed with a largefraction of ‘old’ water at depth (Clark and Fritz, 1997). Lithogenictritium in most aquifers is close to 0.1 TU (Clark and Fritz, 1997),suggesting that lithogenic tritium should not be significant in the NewAlbany Shale. Poor well completion, remnant drilling fluids notflushed during well development, hydraulic fracturing fluids, andnaturally occurring highly permeable zones are all avenues for surfacewater contamination of deep fluids. Unfortunately, these possibilitiesare difficult to assess, and remain likely sources of ‘modern’ watercontamination.

Despite the presence of detectable tritium, 14C results couldrepresent 1) the average age of the groundwater (b50 ka), 2) an ageolder than the average groundwater age due to the input of ‘dead’organic carbon from microbial activity (≪50 ka), and/or 3) contam-

ination of deep groundwater with surface materials (e.g. surfacewater, cement casing, etc.) or brief exposure to atmosphere duringsampling (N50 ka). To obtainmore accurate estimates of groundwaterages, noble gas techniques were used.

4.4. Noble gases

The accumulation of radiogenic 4He in aquifers provides a tool toconstrain the age of the groundwaters beyond the ~50 ka limit of 14C(ka to Ma timescales) and has the added advantage of beinginsensitive to small additions of modern surface water (Torgersenand Clarke, 1985). Interaction between the groundwater and any gasor oil phase results in the accumulated 4He, together with dissolvedatmosphere-derived noble gases gained by the groundwater atrecharge, being acquired by the oil or gas phase (e.g. Ballentineet al., 1991, 1996). The combination of noble gases in the produced oilor gas phase can be used to reconstruct the 4He concentration of thewater prior to phase fractionation and gain a 4He age for theassociated water phase (Zhou et al., 2005; Zhou and Ballentine, 2006).

The He, Ne, Ar, Kr and Xe isotopic compositions of the producedgases are presented in Table 2 and supplementary Table S1. Of the 24noble gas samples from the shale, 20 samples have 4He/20Ne ratiosranging from 36 to 32,400, much higher than the atmospheric value of0.288 (e.g. Kipfer et al., 2002). However, 4 samples (MM18–21) havemuch lower 4He/20Ne ratios (0.4 to 2.56), indicating significantcontamination with air He. The 3He/4He ratio (R), normalized to theatmospheric ratio (Ra=1.4×10−6) in the former samples rangesfrom 0.07 to 0.18, which is near the upper crust value of 0.01 to 0.02Ra (Fig. 6a) (Ballentine and Burnard, 2002). This result indicates thatthe atmosphere derived 4He contribution is negligible in these 20samples, and that they are dominated by a crustal radiogenic source.Furthermore, by defining R/Ramantle as 8 and R/Racrust as 0.05,calculated percentages of crustal contribution for samples not contam-inated with air range from 99.0 to 99.8% (Ballentine et al., 1991; Elliotet al., 1993), indicating onlyminor (b1%) contributions from themantle.

The 4 samples showing He that is significantly contaminated withair also have 20Ne/36Ar ratios close to the air value (0.52). An additional6 samples (MM7, MM14–16, TM1, TM3) also have 20Ne/36Ar valuessimilar to air or greater (N0.45) (Fig. 6b). These values could indicatefractionation due to solubility related water–gas–oil phase interactions,however for these 6 samples the 84Kr/36Ar ratios (0.0179–0.0229)and 132Xe/36Ar ratios (4.05–6.38×10−4) are similar to air (0.02 and

Table 2Tritium, 14C, and noble gas data for samples from the New Albany Shale and carbonate aquifers. Also, calculated 4He ages using representative values for in-situ flux, as well as in-situ flux with external flux.

ID Tritium TU Err 14C pmc Err 3He cc/cc STP(×10−11)

% Err 4He cc/cc STP(×10−5)

% Err 20Ne cc/cc STP(×10−8)

% Err 40Ar cc/cc STP(×10−5)

% Err 36Ar cc/cc STP(×10−8)

% Err 4He in-situ agesyear(×106)

4He ages withexternal fluxyear(×105)

New Albany Shale — microbial methaneMM0 0.7 0.29 0.98 0.11MM1 1.45 0.12MM2 1.4 0.40 0.27 0.11 1.64 7.07 10.76 0.09 0.62 3.63 3.71 3.25 3.56 9.05 46.2 7.9MM3 1.67 7.07 9.93 0.11 1.13 0.30 4.49 1.72 7.50 2.56 20.2 3.4MM4 3.91 7.07 15.55 0.07 0.60 4.10 3.63 3.77MM5 0.54 4.24 3.47 0.35 0.71 1.47 3.93 0.47 6.39 0.44 8.3 1.4MM6 0.23 0.11MM6a b0.36MM7 2.8 0.38 15.08 0.16 1.54 7.07 13.70 0.44 59.05 1.55 45.93 1.38 136.52 1.92MM8 b0.8 22.61 0.19MM9 1.49 0.12MM10 0.94 3.74 5.27 0.16 4.42 6.38 11.67 1.16 19.37 2.22 4.2 0.8MM11 1.41 4.46 7.69 0.11 1.79 0.91 4.34 0.51 6.90 0.52 17.0 3.4MM12 0.74 4.24 4.26 0.17 1.73 0.85 3.74 0.47 5.21 0.39 12.5 2.5MM13 5.27 0.11 0.26 7.80 3.24 3.92 3.45 9.82 23.3 4.0MM14 1.36 7.07 7.63 0.10 212.12 0.31 80.43 1.69 266.89 4.16MM15 0.33 0.10 0.96 7.07 6.46 0.11 7.24 0.26 6.07 1.64 14.44 2.59MM15a 0.56 0.10MM16 1.21 0.10 6.61 0.09 20.14 0.32 13.17 1.93 29.29 3.39MM17 0.83 0.10 3.12 7.07 20.44 0.09 1.86 1.12 8.64 2.17 10.26 3.01 30.4 3.1MM18 0.57 4.24 0.59 0.46 1687.88 0.75 1013.65 0.40 3490.96 0.66MM19 0.31 3.84 1.16 0.17 453.47 0.63 279.33 0.45 927.49 0.59MM20 0.70 8.57 1.49 0.25 1103.28 0.81 617.30 0.35 2135.84 0.40MM21 0.50 0.10 1.07 4.24 0.97 0.21 1813.41 0.82 1086.73 0.47 3722.57 0.34

New Albany Shale — thermogenic methaneTM1 5.13 4.24 33.79 0.10 835.67 0.83 451.30 0.50 1557.85 0.35TM2 4.63 4.24 43.97 0.19 1.36 1.12 9.68 0.45 14.70 0.28 45.7 10.3TM3 5.09 4.13 54.17 0.31 50.30 0.65 41.62 0.40 112.87 0.50TM4 6.16 4.24 63.22 0.29 2.60 0.92 16.37 0.38 21.06 0.55 45.8 10.3TM5 5.98 4.24 60.88 0.21 2.81 1.26 11.79 0.48 18.21 0.75 51.0 11.5TM6 5.27 3.82 55.96 0.25 2.00 1.09 10.73 0.51 17.95 0.56 47.6 10.7TM7 4.42 4.24 40.08 0.12 1.78 0.86 10.04 0.36 17.32 0.34 35.3 8.0

Carbonate aquifersD1 0.68 0.10D2 0.41 0.10D3 1.3 0.30 1.70 0.11D4 1.04 0.11M1 0.23 0.10

a Duplicate 14C analysis.

34M.E.Schlegel

etal./

Chemical

Geology

287(2011)

27–40

Fig. 6. (a) Plot of 3He/4He (R/Ra) versus 1/4He shows that 4 New Albany Shale sampleshave significant air contamination, whereas the remaining samples are similar to theupper crust value. The two different symbols denote samples collected from areas ofdominantly microbial methane versus thermogenic methane (Schlegel et al., 2011).(b) Plot of 20Ne/36Ar versus 1/36Ar shows that 10 samples have significant aircontamination, whereas 13 samples are near the air value, indicating near completedegassing of the water to the gas phase. (c) Plot of 40Ar/36Ar versus 1/36Ar shows twomixing trends between an Ar air endmember and Ar endmembers in the north andsouth of the study area.

35M.E. Schlegel et al. / Chemical Geology 287 (2011) 27–40

7.27×10−4, respectively) and distinct from air saturated water (0.045and 30.4×10−4, respectively) (e.g. Kipfer et al., 2002), suggestingsignificant air contamination.

The remaining samples (14) have 20Ne/36Ar ratios (0.08 to 0.33) thatare similar to the range found in air-equilibrated water (0.14–0.18)(Fig. 6b). While there can be significant sources of isotopically air-like Krand Xe in hydrocarbon bearing systems (Torgersen and Kennedy, 1999;Zhou et al., 2005), the 84Kr/36Ar and 132Xe/36Ar ratios in these remainingsamples (0.0348–0.0556 and 13.8–35.0×10−4, respectively) are alsosimilar to the range found in air saturated water. In these samples weassume that there has beenminimal fractionation of the inert gases, withnear complete transfer of the noble gas in the water to the gas phase.These14samplesareused indetermining 4Heages for groundwater in theNew Albany Shale; however 36Ar for MM4 was not measured, reducingthe number of samples used to calculate 4He ages to 13.

Interestingly, the 24 New Albany Shale samples reveal two distincttrends on a plot of 1/36Ar vs. 40Ar/36Ar (Fig. 6c). One trend is formedby samples from the south (western Kentucky and south-eastIndiana), in contrast to the other trend, formed by samples from themore northern areas (west-central Indiana) (Fig. 6c). These twotrends appear to have been formed by mixing of a common 40Ar/36Arratio similar to air, and 2 different argon endmembers resulting fromeither a change in 36Ar or 40Ar. Following the arguments for anexternal flux of radiogenic 4He into the aquifer (Section 4.5), theendmember difference may best be explained by the thermogenicgases being associated with older fluids, which have accumulatedmore 40Ar due to 40K decay, as compared to the microbial gases. Onemicrobial gas sample (MM10) falls in the thermogenic ‘southern Arendmember trend’ (Fig. 6c) and has the youngest 4He groundwaterage, suggesting that this sample is anomalous, with no clear salinity orspatial relation to the thermogenic gas samples.

4.5. 4He groundwater age calculation

Helium-4 ages were calculated following the method of Zhou andBallentine (2006). As mantle helium is negligible in this system (b1%),the measured 4He is comprised of the 4He obtained from when thewater was last in contact with the atmosphere air saturated water(ASW), 4He obtained from radioactive decay of primarily U and Th inthe aquifer matrix (in-situ production), and 4He obtained from oldercrustal fluids that has migrated into the aquifer (external flux). Boththe in-situ production and external flux of 4He contain a timecomponent, which is used to calculate the 4He age. Because the4He/20Ne ratio in the 13 samples being considered is much higherthan either typical air equilibrated water or air, the contribution of4HeASW is considered to be insignificant and the equation can besimplified to:

4Hemeasured=4Hein�situ production þ4 Heexternal flux:

To estimate a groundwater age we have to consider the 4Hein-situproduction+4Heexternal flux concentration measured in the gas phase interms of the 4He concentration in the groundwater before it contactedthe gas phase. The lack of fractionation of 20Ne/36Ar (previouslydiscussed) occurs if there has been quantitative and closed systemtransfer of the dissolved gases into the methane gas (Zhou andBallentine, 2006). Therefore the concentration of 4He in the gas phasecan be related to the concentration of 4He in the original water phasefrom the gas phase 4He/36Ar ratio and the concentration of 36Ar in thewater phase before degassing (known) using:

4Heconcentration in water=4He=36ArmeasuredT

36ArASW:

The amount of 36Ar in ASW will be 1/295.5 of the total argondissolved in thewaterwhich is dependenton thepressure, temperature,

Table 3Minimum, maximum, and representative values for estimated parameters used to calculate 4He ages.

Parameter Units Minimum value Maximum value Representative value

U — New Albany Shale ppm 1 140 36.41Th — New Albany Shale ppm 3 21 10.04Density — New Albany Shale g/cm3 2.4 2.8 2.7Porosity — New Albany Shale % 10 14 14Thickness — New Albany Shale, Harrison Co cm 2743Thickness — New Albany Shale, Sullivan Co and Daviess Co cm 4572Thickness — New Albany Shale, Knox Co cm 5334Thickness — New Albany Shale, Christian Co cm 6096U — upper crust ppm 2.5 3.1 2.8Th — upper crust ppm 7.8 10.9 10.7Density — upper crust g/cm3 2.5 2.8 2.6Thickness — upper crust km 9.3 15 12.3U — lower crust ppm 0.0001 0.91 0.28Th — lower crust ppm 0.0001 3.5 1.06Density — lower crust g/cm3 3 3.3 3.3Thickness — lower crust km 26.6 37 36.9Initial ArASW concentrations cm3 STPg−1 2.47E−04 4.47E−04 3.47E−04

36 M.E. Schlegel et al. / Chemical Geology 287 (2011) 27–40

and salinity of the water at the surface. Given periglacial recharge orquick transport of precipitation below the ice sheet for subglacialrecharge, a pressure of 1 atm is assumed for when the water wasequilibrated with the atmosphere to estimate the total Ar dissolved inthe water (Table 3). A minimum and maximum value of ArASW wasused: 2.47×10−4 cm3 STPg−1 for water with 35‰ salinity and 20 °C,and 4.47×10−4 cm3 STPg−1 for water with 0.1‰ salinity and 4 °C.Additionally a representative valueof 3.47×10−4 cm3 STPg−1 (averageof minimum and maximum values) was used.

The in-situ production of 4He can be calculated by (Torgersen,1980):

4Hein�situ production = ρaquiferλ J4 1−φð Þt� �

=φ

where ρaquifer = density of aquifer (g/cm3), λ = transfer efficiencyfrom mineral to water, φ = rock porosity, t = time, and J4 is definedby:

J4 = 0:2355T10−12T U½ �T 1 + 0:123T Th½ �= U½ �–4ð Þð Þ

Fig. 7. Groundwater 4He ages are shown for dominantly microbial and thermogenic methaperiods. The error bars shown give the maximum (2 σ) uncertainty in 4He age, estimatedproduction, external 4He flux (Table 3) and analytical error. The error term is dominated by tsamples the uncertainty is much smaller, assuming a near uniform external 4He flux, andmethane is significantly younger than the groundwater associated with the thermogenic ga

where [U] and [Th] are the U and Th concentrations in the rock in ppm(Craig and Lupton, 1976; Torgersen, 1980).

4Heexternal flux can be calculated by:

4Heexternal flux = J4ρcrustHtð Þ= φhð Þ

where ρcrust = the density of the crust, H = thickness (cm) of crust,and h = thickness (cm) of aquifer. For most parameters a minimum,maximum, and representative value was found in the literature andused to calculate minimum, maximum, and representative ages(Table 3; discussed in Appendix A). Crustal flux of 4He containsmany uncertainties including errors inherent in the estimates ofparameters for the aquifer and the bulk crust.

Calculated 4He ages for representative values are presented inTable 2 and Fig. 7. 4He ages are presented for two end-memberscenarios. The first is a closed system environment in which onlyin-situ radiogenic noble gases are present. The calculated 4He averagegroundwater ages using representative values for only in-situproduction range from 4.2 to 51 Ma (average 30 Ma), which isyounger than the age of the formation (Devonian–Mississippian

ne areas and compared to the presence of ice sheets in the Illinois Basin during glacialusing the minimum and maximum parameter values used to calculate the in situ 4Hehe uncertainty in external 4He flux. When considering relative age differences betweenwould plot within the symbols presented. Groundwater associated with the microbialses.

37M.E. Schlegel et al. / Chemical Geology 287 (2011) 27–40

boundary is at 360 Ma). Though these ages cannot be discarded, themajority of sedimentary basins require an external crustal flux of 4Heto obtain meaningful groundwater ages (e.g. Torgersen and Clarke,1985; Zhou and Ballentine, 2006). When an external flux of 4He fromthe upper and lower crust is invoked, the average groundwater agerange decreases to 0.082–1.2 Ma (average 0.59 Ma) for representativevalues, with age ranges as old as 0.14–1.9 Ma (average 1.0 Ma, usingmaximum values with the minimum value of porosity) and as youngas 0.039–0.54 Ma (average 0.28 Ma, using minimum values with themaximum value of porosity).

Two key assumptions made in determining the groundwater 4Heage are that: 1) an average crustal 4He flux over the lifetime of thegroundwater age is valid; and 2) the 4He flux is uniform across theIllinois Basin. The variance of 4He flux is smallest in regions that can beconsidered a ‘uniform, stable’ shield,where anapproximate steady stateof 4He crustal degassing is obtained (Torgersen, 2010). The Illinois Basinoverlies a central stable platform adjacent to the Canadian Shield thathas seen no major structural deformation since the Precambrian in thenorth-central portion of the basin, though in the southern portion of thebasin deformation has occurred as late as the Mesozoic and the RoughCreek fault systemhas been active throughout thePaleozoic (Buschbachand Kolata, 1990). In this work we have calculated ages based on theaverage crustal 4He flux to be 2.42×1010 atoms 4Hem−2 s−1 using therepresentative values in Table 3. This compares with the time- andarea-weighted arithmetic mean (standard deviation) 4He crustaldegassing flux of 3.32(±0.45)×10104He atoms m−2 s−1 (Torgersen,2010), which remains the greatest uncertainty in absolute, rather thanrelative, age determination.

Furthermore, the variance of 4He flux may be controlled by localstructural features and episodic events, as observed by Kulongoski et al.(2005), though these effects result in an average 4He crustal flux forlarge spatial and temporal scales. In contrast, other studies have founddiffuse crustal 4He input is consistentwith a coherent spatial increase in4He groundwater age from recharge (e.g. Zhou et al., 2005). Within theIllinois Basin there is a clear spatial coherence between the differentsample populations, the microbial gases which are closer to rechargezones have a younger 4He groundwater age than the deeperthermogenic methane which has an older 4He associated groundwaterage. The clumping of the 4He ages for the thermogenic gases suggeststhat on this spatial scale the input of 4He is relatively uniform.Within theregion of the Illinois Basin studied here there is nomajor structural faultor feature that divides the thermogenic from the microbial methaneproducing areas — the principal spatial difference simply being depthand distance from recharge. The assumption that the 4He flux into thesystem associatedwithmicrobial gas production is the same as that intothe thermogenic area would then also seem reasonable.

4.6. Deep groundwater recharge

All calculated 4He ages indicate that the groundwater recharge isolder than the Wisconsin glaciation (25 to 13 ka in the Illinois Basin),whereas some of the groundwater may have recharged in associationwith Illinoian (300 to 125 ka) and Pre-Illinoian (2.7 to 0.5 Ma)glaciations (Fig. 7). These results support the hypothesis that thedetected tritium and 14C is likely from contamination and do notrepresent the groundwater age, which is N50 ka. These results areconsistent with findings fromMcIntosh andWalter (2006) and Ebertsand George (2000) who found groundwater ages of 17 to 50 ka inSilurian–Devonian carbonates up-dip of the deep natural gas NewAlbany Shale wells. Furthermore, calculated 4He ages for groundwaterfrom the predominantly thermogenic methane area (0.80 to 1.2 Ma,average 1.0 Ma) are older than groundwater in the predominantlymicrobial methane area (0.082 to 0.79 Ma, average 0.33 Ma),demonstrating that more groundwater flushing has occurred nearthe basin margins as compared to the basin center. Similarly, in themore shallow Pennsylvanian coals, thermogenic methane is primarily

found near the center of the basin in areas of limited meteoricrecharge (Strąpoć et al., 2007). These results are consistent withchloride and δ18O data from the New Albany Shale, which showelevated values toward the basin center, indicating limited flushing byfresh water. Though the deep saline brines were probably formedfrom evapo-concentrated seawater, suggesting that the basin fluidsare fairly stagnant (e.g. Stueber andWalter, 1994), all groundwater inthe study area may have been influenced by Pleistocene recharge.Using these age estimates and distances from the Devonian carbonatesubcrop, flow rates range from 0.19 to 1.8 m/year for the microbialmethane area (~150 km distant from subcrop) and 0.28 to 0.41 m/yearfor the thermogenic methane area (~330 km distant from subcrop).Eberts and George (2000) and McIntosh and Walter (2006) foundminimal flushing of groundwater N40 m depth since the Wisconsinglaciation in the Devonian aquifer in the northern portion of the IllinoisBasin, suggesting that modern groundwater recharge deep into thebasin is negligible.

Meteoric recharge deep into the Illinois Basin may have occurredwhile the subcrops where covered with kilometer-thick ice sheets, orwhen the subcrops were ice free, however the current data set cannotdifferentiate between subglacial or periglacial groundwater recharge.If groundwater recharge occurred beneath the ice sheets, then aconduit for rapid transport of paleoprecipitation from the ice surfaceor proglacial lakes to the subglacial environment is required tomaintain a relatively enriched isotopic signature, and groundwaterrecharge would be limited to when glaciers were present in theIllinois Basin. If groundwater recharged aquifers in the absence ofglaciers, then the effects of ice sheet loadingmust propagate beyond theglacialmargin,where significantamounts of paleoprecipitation collectedin proglacial lakes and river systems (Rea et al., 1994; Killey, 1998;Marksamer et al., 2007). Future research, coupling hydrologic modelingwith age tracer results from this study, is targeted at elucidating thepossible mechanisms of groundwater recharge.

4.7. In-situ microbial methane production rates

If fresh meteoric water also transported methanogens deep intothe organic-rich New Albany Shale as suggested by Schlegel et al.(2011), then economic accumulations of microbial gas may haveformed in as little as 82 ka (using best estimates of 4He ages),assuming piston flow with no dispersion or diffusion. This timescalecan be compared to the 4He age of 22 ka derived for microbialproduction of commercial quantities of coalbed methane gas in theSan Juan Basin (Zhou and Ballentine, 2006). Total natural gasestimates for the New Albany Shale range from 86 to 160 TCF (Curtis,2002), of which a portion is thermogenic methane (McIntosh et al.,2002). Minimum in-situ methane production rates are calculatedassuming a linear rate of accumulation, 20–80% of the methane asmicrobial in origin, and the range of 4He ages for predominantlymicrobial methane calculated with representative values. Magnitudeof production rates range from 10 to 100 TCF/Ma for 20% microbialmethane to 100–1000 TCF/Ma for 80% microbial methane. Thesemethane production rates are much higher than those observed indeep quartzite and metamorphosed volcanic units in South Africa(0.01–100 nM/year which is approximately 10−5–10−1 scf/Ma;Onstott et al., 2006). However, these rates are significantly less thanestimated microbial methane production rates from laboratoryexperiments using coal. Jones et al. (2008) applied a bioassay to 16different coal samples from 4 basins for at least 145 days, resultingin a range of methane production rates (0–56 scf/ton; average5.6 scf/ton). Applying the average value to the volume of the NewAlbany Shale, the magnitude of methane production is 10 TCF/year.The difference in magnitude of methane production rates (~104 to106) indicates that either laboratory experiments are fundamentallydifferent than in-situ conditions (e.g. increased substrate surface areaand nutrients), or that the deep microbial methane systems may be

38 M.E. Schlegel et al. / Chemical Geology 287 (2011) 27–40

more open than assumed. Slow rates of in-situ microbial productionmay be due to limited nutrients and substrate, transport processes, and/or other toxic environmental factors such ashigh concentrationsof tracemetals (e.g. Megonigal et al., 2005; Karri et al., 2006).

5. Conclusions

Meteoric water penetrated up to ~1 km depth into saline brines,which stimulated microbial methanogenesis in the organic-rich NewAlbany Shale. The dilute groundwater parallels terminal moraines ofthe Laurentide Ice Sheet, suggesting glacial mediated recharge.However, isotopicmixing trends indicate that the recharge endmember(~−7‰ δ18O) is higher than the assumed bulk ice sheet value (b−15‰δ18O). Paleoprecipitation records near the Illinois Basin reveal thatthroughout most of the Pleistocene the estimated δ18O of precipitationranged from−10 to−5‰, suggesting that the dilutewater at depthwasprimarily sourced from paleoprecipitation with minor contributionsfromglacialmeltwater fromeither subglacial rechargeorproglacial lakes.

The timing of groundwater recharge was calculated using noblegases. For the majority of samples the atmosphere derived 4Hecontribution is negligible, and is dominated by a crustal radiogenicsource, with near complete transfer of dissolved noble gases to the gasphase. In addition, mantle derived helium is negligible for all samples(b1%). 4He ages of the groundwater range from 0.082 to 1.2 Ma, withthermogenic methane associated with older fluids (average 1.0 Ma), ascompared to microbial methane (average 0.33 Ma). These results areconsistent with chloride and δ18O data from the New Albany Shale,which showelevated values toward the basin center in the thermogenicmethane area indicating limited flushing by freshwater, as compared tothemicrobial methane area nearer to the basin margins. All locations inthe study area have been influenced by Pleistocene recharge; however,deep salinebrineshavenot been completelyflushed, suggesting that thebasin fluids are fairly stagnant. Using 4He ages, themagnitude of in-situmicrobial methane production is estimated to range from 10 to1000 TCF/Ma,which is ~104 to106 times slower thanaverage laboratorymethane production rates from coals, implying that laboratoryexperiments may be fundamentally different than in-situ conditions.The accumulation of economic quantities of natural gas in associationwith Pleistocene fresh water recharge has implications for targetingundeveloped microbial gas accumulations, improving natural gasreservoir estimates world-wide, the potential of in-situ stimulation ofmethanogenesis, and understanding carbon cycling in organic-richsubsurface reservoirs.

Supplementarymaterials related to this article can be found onlineat doi:10.1016/j.chemgeo.2011.04.019.

Acknowledgments

Funding was provided by NSF (EAR-0635685; McIntosh), theUSGS, and RPSEA through the “Ultra-Deepwater and UnconventionalNatural Gas and Other Petroleum Resources” program. Co-operationfrom several energies companies in Indiana, Kentucky, and Illinois forwell access is very much appreciated. Also, tremendous field help wasprovided by JoeWade, Matthew Kirk, Samuel Miller, Julian Damashek,and Daniel Kekacs; and excellent laboratory support was given by TimCorley, Brittney Bates, and Stephen Osborn.

Appendix A. Discussion of the parameter values for 4He ages

Parameters for calculating 4He ages obtained from the literaturegenerally had more than one value. These ranges in values (minimumand maximum) and the most representative value are discussedbelow.

The density for the New Albany Shale is taken to be 2.7 g/cm3

(representative value) (Stevenson and Dickerson, 1969), which iswithin typical shale densities (2.6–2.8 g/cm3) (minimum and maxi-

mum values) (Patwardhan, 2010). Transfer efficiency is generallyassumed to be 1 for temperatures N50 °C. Currently bottom holetemperatures are as low as 27 °C (Curtis, 2002), however in the pasttemperatures probably exceeded 100 °C in areas sampled (Cluff andByrnes, 1990; Rowan et al., 2002; Schlegel et al., 2011). For thesecalculations we assume total transfer of 4He from the minerals to thewater (λ=1),whichmay result in an overestimate of groundwater age.Rock porosity for the New Albany Shale ranges from 0.1 to 0.14, with anaverage porosity for theDevonian systemof 0.14 (Howard, 1990; Curtis,2002). As the NewAlbany Shale is part of the Silurian–Devonian aquifersystem, a porosity of 0.14 is used as the representative value, 0.1 is usedas the minimum value, and 0.14 is used as the maximum value. Notethat because the porosity is in the denominator for 4He age calculations,the maximum value was used in the minimum calculations, and viceversa.

Uranium and Th concentrations in Devonian organic-rich shales inthe Illinois and Appalachian Basins were obtained from the literature(Ostrom et al., 1955; Leventhal and Hosterman, 1982; Tuttle et al.,2003; Algeo, 2004; Clauer et al., 2006; Perkins et al., 2008).Concentrations of U range from 1 to 140 ppm (used as minimumand maximum values), with published averages ranging from 15 to44 ppm. Concentrations of Th range from 3 to 21 ppm (used asminimum and maximum values), with published averages rangingfrom 4.2 to 13.9 ppm. Of the studies considered, only two (Tuttle et al.,2003; Clauer et al., 2006) had samples from the New Albany Shale inthe Illinois Basin with results for both U and Th concentrations. Claueret al. (2006) reported anomalously low Th concentrations (about6 ppm lower) as compared to other studies. For this reason, theaverage concentrations from Tuttle et al. (2003) were used as therepresentative concentrations for U (36.41 ppm) and Th (10.04 ppm).Thesemeasurementswere taken from a from a NewAlbany Shale corein east-central Kentucky (21 samples from 8 to 53 m depth; Tuttleet al., 2003).

The density of the upper and lower crust is 2.5 to 2.8 g/cm3 and 3.0 to3.3 g/cm3, respectively (takenasminimumandmaximumvalues) (TaylorandMcLennan, 1985). Representative valueswere chosen tobe thevaluesused by Zhou and Ballentine (2006) of 2.6 g/cm3 for the upper crust, and3.3 g/cm3 for the lower crust. Thickness of the crust for the Illinois Basinwas obtained from Ruff et al. (1994) and Earthscope Automated ReceiverStudy (EARS) (2010) (http://www.seis.sc.edu/ears.new/earsHasMoved.html). Ruff et al. (1994) estimates the crust thickness for Indiana (uppercrust=9.3 km, lower crust=26.6 km), and compares the estimate tomid-continent crustal thickness (upper crust=15 km, lowercrust=26.5 km). EARShas total crustal thickness estimates for5 locationsin the Illinois Basin ranging from52 to 48 kmdepth (average 49.2 km). InTaylor andMcLennan (1985), theupper crust is estimated tobe about25%of the total crustal thickness with an average value of 10 km +/− 3.5.Twenty-five percent of the average total crustal thickness in the IllinoisBasin is 12.3 km, resulting in a lower crustal thickness of 36.9 km.Minimum crustal thicknesses were taken from Ruff et al. (1994) forIndiana. Maximum crustal thicknesses were taken from Ruff et al. (1994)mid-continent estimates for the upper crust (15 km), and the lower crustthickness was estimated using the maximum total crustal measurementin the Illinois Basin (52 km) minus the maximum upper crust thickness(52 km−15 km=37 km). Representative crustal thicknesses used theaverage total crustal thickness for the IllinoisBasin (49.2 km),with25% forthe upper crust (12.3 km), and 75% for the lower crust (36.9 km).Thickness of the aquifer was taken to be the height of the New AlbanyShale in the areas sampled, and was estimated from Cluff and Byrnes(1990) with ranges from 2743 to 6096 cm.

Uranium and Th concentrations in the crust are from Taylor andMcLennan (1985). For the upper crust they report an average Uconcentration of 2.8 ppm (taken for the representative value), withmeasured values generallywithin+/−10%, giving aminimumestimateof 2.5 ppm and a maximum estimate of 3.1 ppm. The average Thconcentration is 10.7 ppm(take for the representative value),withmost

39M.E. Schlegel et al. / Chemical Geology 287 (2011) 27–40

measured values within +/−2%, giving a minimum estimate of10.5 ppm and a maximum estimate of 10.9 ppm. Taylor and McLennan(1985) calculate average lower crustal concentrations of U (0.28 ppm)and Th (1.06 ppm) (taken to be representative values) from upper crustestimates and total crust estimates of U and Th concentrations.Maximum concentrations for the lower crust are assumed to be thetotal crust concentrations (0.91 ppm U and 3.5 ppm Th). Minimumconcentrations are assumed to be very low, and are arbitrarily chosen as0.0001 ppm for both U and Th.

References

Algeo, T.J., 2004. Can marine anoxic events draw down the trace element inventory ofseawater? Geology 32 (12), 1057–1060.

Ballentine, C.J., Burnard, P.G., 2002. Production, release and transport of noble gases inthe continental crust. Rev. Mineral. Geochem. 47, 481–538.

Ballentine, C.J., Sherwood Lollar, B., 2002. Regional groundwater focusing of nitrogenand noble gases into the Hugoton-Panhandle giant gas field, USA. Geochim.Cosmochim. Acta. 66 (14), 2483–2497.

Ballentine, C.J., O'Nions, R.K., Oxburgh, E.R., Horvath, F., Deak, J., 1991. Rare gasconstraints on hydrocarbon accumulation, crustal degassing and groundwater flowin the Pannonian Basin. Earth Planet. Sci. Lett. 105, 229–246.

Ballentine, C.J., O'Noins, R.K., Coleman, M., 1996. AMagnus opus: He, Ne and Ar isotopesin a North Sea oil field. Geochim. Cosmochim. Acta. 60, 831–849.

Bekele, E., Rostron, B., Person, M., 2003. Fluid pressure implications of erosionalunloading, basin hydrodynamics and glaciations in the Alberta Basin, WesternCanada. J. Geochem. Explor. 78–79, 143–147.

Bethke, C.M., 1985. A numerical model of compaction-driven groundwater flow andheat transfer and its application to the paleohydrology of intracratonic sedimentarybasins. J. Geophys. Res. 90 (B8), 6817–6828.

Bethke, C.M., Reed, J.D., Oltz, D.F., 1991. Long-range petroleum migration in the IllinoisBasin. AAPG Bull. 75 (5), 925–945.

Bond, D.C., 1972. Hydrodynamics in Deep Aquifers of the Illinois Basin. Illinois StateGeological Survey, p. 41.

Boulton, G.S., Caban, P.E., 1995. Groundwater flow beneath ice sheets: part 2— its impacton glacier tectonic structures and moraine formation. Quat. Sci. Rev. 14, 563–587.

Boulton, G.S., Lunn, R., Vidstrand, P., Zatsepin, S., 2007. Subglacial drainage bygroundwater-channel coupling, and the origin of esker systems: part 1— glaciologicalobservations. Quat. Sci. Rev. 26, 1067–1090.

Breemer, C.W., Clark, P.U., Haggerty, R., 2002. Modeling the subglacial hydrology ofthe late Pleistocene Lake Michigan Lobe, Laurentide Ice Sheet. GSA Bull. 114 (6),665–674.

Buschbach, T.C., Kolata, D.R., 1990. Regional setting of the Illinois Basin. In: Leighton,M.W., Kolata, D.R., Oltz, D.F., Eidel, J.J. (Eds.), Interior Cratonic Basins. AmericanAssociation of Petroleum Geologists, Tulsa, pp. 29–58.

Christiansen, E.A., Sauer, E.K., 1998. Geotechnique of Saskatoon and surrounding area,Saskatchewan, Canada. Geol. Assoc. Can. Spec. Pap. 42, 117–145.

Christiansen, E.A., Genzwill, D.J., Meneley, W.A., 1982. Howe Lake; a hydrodynamicblowout structure. Can. J. Earth Sci. 19 (6), 1122–1139.

Clark, I., Fritz, P., 1997. Environmental Isotopes in Hydrogeology. Lewis Publishers, NewYork.

Clauer, N., Chaudhuri, S., Lewan,M.D., Toulkeridis, T., 2006. Effect of thermalmaturationon the K–Ar, Rb–Sr and REE systematic of an organic-rich New Albany Shale asdetermined by hydrous pyrolysis. Chem. Geol. 234 (1–2), 169–177.

Clayton, R.N., Friedman, I., Graf, D.L., Mayeda, T.K., Meents, W.F., Shimp, N.F., 1966. Theorigin of saline formation waters, 1. Isotopic composition. J. Geophys. Res. 71 (16),3869–3882.

Cluff, R.M., Byrnes, A.P., 1990. Lopatin analysis of maturation and petroleum generation inthe Illinois Basin. In: Leighton, M.W., Kolata, D.R., Oltz, D.F., Eidel, J.J. (Eds.), InteriorCratonic Basins. American Association of Petroleum Geologists, Tulsa, pp. 425–454.

Craig, H., Lupton, J.E., 1976. Primordial neon, helium, and hydrogen in oceanic basalts.Earth Planet. Sci. Lett. 31, 369–385.

Curtis, J.B., 2002. Fractured shale-gas systems. AAPG Bull. 86 (11), 1921–1938.Cutler, P.M., MacAyeal, D.R., Mickelson, D.M., Parizek, B.R., 2000. A numerical

investigation of ice-lobe permafrost interaction around the southern LaurentideIce Sheet. J. Glaciol. 46, 311–325.

Denniston, R.F., Gonzalez, L.A., Asmerom, Y., Polyak, V., Reagan,M.K., Saltzman,M.R., 2001.A high-resolution speleothem record of climatic variability at the Allerød-YoungerDryas transition in Missouri, central United States. Palaeogeol. Palaeoclimat.Palaeoecol. 176, 147–155.

Denniston, R.F., DuPree, M., Dorale, J.A., Asmerom, Y., Polyak, V.J., Carpenter, S.J., 2007.Episodes of late Holocene aridity recorded by stalagmites from Devil's Icebox Cave,central Missouri, USA. Quat. Res. 68, 45–52.

Dettman, D.L., Smith, A.J., Rea, D.K., Moore Jr., T.C., Lohmann, K.C., 1995. Glacialmeltwater in Lake Huron during early postglacial time as inferred from single-valveanalysis of oxygen isotopes in ostracodes. Quat. Res. 43, 297–310.

Dorale, J.A., González, L.A., Reagan, M.K., Pickett, D.A., Murrell, M.T., Baker, R.G., 1992. Ahigh-resolution record of Holocene climate change in speleothem calcite from ColdWater Cave, Northeast Iowa. Science 258, 1626–1630.

Dorale, J.A., Edwards, R.L., Ito, E., Gonzalez, L.A., 1998. Climate and vegetation history ofthe midcontinent from 75 to 25 ka: a speleothem record form Crevice Cave,Missouri, USA. Science 282, 1871–1874.

EARS (Earthscope Automated Receiver Study), 2010. accessed Nov. 2010 http://www.seis.sc.edu/ears.new/earsHasMoved.html.

Eberts, S.M., George, L.L., 2000. Regional ground-water flow and geochemistry in theMidwestern basins and arches aquifer system in parts of Indiana, Ohio, Michigan,and Illinois. Regional Aquifer-System Analysis — Midwestern Basins and Arches:U.S. Geol. Surv. Reston, VA., pp. C1–C103.

Edmunds, W.M., Milne, C.J., 2001. Palaeowater in Coastal Europe: evolution ofgroundwater since the late Pleistocene. Geol. Soc. London: Special Publication,p. 189.

Edwards, T.W.D., McAndrews, J.H., 1989. Paleohydrology of a Canadian Shield lakeinferred from O-19 in sediment cellulose. Can. J. Earth Sci. 26 (9), 1850–1859.

EIA (U.S. Energy Information Administration), 2010. Annual Energy Review 2009,Report No. DOE/EIA-0384 (2009) http://www.eia.doe.gov/emeu/aer. accessed Nov.2010.

Elliot, T., Ballentine, C.J., O'Nions, R.K., Ricchiuto, T., 1993. Carbon, helium, neon andargon isotopes in a Po Basin (northern Italy) natural gas field. Chem. Geol. 106,429–440.

Garven, G., Appold, M.S., Toptygina, V.I., Hazlett, T.J., 1999. Hydrogeologic modeling ofthe genesis of carbonate-hosted lead–zinc ores. Hydrogeol. J. 7 (1), 108–126.

Geyh, M.A., Kuenzl, R., 1981. Methane in groundwater and its effect on 14C groundwaterdating. J. Hydrol. 52, 355–358.

Gilfillan, S.M., Ballentine, C.J., Holland, G., Blagburn, D., Lollar, B.S., Stevens, S., Schoell,M., Cassidy, M., 2008. The noble gas geochemistry of natural CO2 gas reservoirsfrom the Colorado Plateau and Rocky Mountain provinces, USA. Geochim.Cosmochim. Acta. 72, 1174–1198.

Gooseff,M.N.,McKnight,D.M.,Runkel, R.L., Vaughn,B.H.,2003.Determining longtime-scalehyporheic zone flow paths in Antarctic streams. Hydrol. Processes. 17, 1691–1710.

Gooseff, M.N., Lyons, W.B., McKnight, D.M., Vaughn, B.H., Fountain, A.G., Dowling, C.,2006. A stable isotopic investigation of a polar desert hydrologic system, McMurdoDry Valleys, Antarctica. Arct. Antarct. Alp. Res. 38 (1), 60–71.

Grasby, S.E., Chen, Z., 2005. Subglacial recharge into the Western Canada SedimentaryBasin — impact of Pleistocene glaciations on basin hydrodynamics. GSA Bull.117 (3/4), 500–514.

Grasby, S., Osadetz, K., Betcher, R., Render, F., 2000. Reversal of the regional-scale flowsystem of the Williston Basin in response to Pleistocene glaciations. Geology 28,635–638.

Grootes, P.M., Stuiver, M., 1997. Oxygen 18/16 variability in Greenland snow and icewith 103 to 105-year time resolution. J. Geophys. Res. 102 (C12), 26,455–26,470.

Harmon, R.S., Schwarcz, H.P., 1981. Changes of 2H and 18O enrichment of meteoricwater and Pleistocene glaciations. Nature 290, 125–128.

Hoaglund III, J.R., Kolak, J.J., Long, D.T., Larson, G.J., 2004. Analysis of modern andPleistocene hydrologic exchange between Saginaw Bay (Lake Huron) and theSaginaw lowlands area. GSA Bull. 116, 3–15.

Howard, R.H., 1990. Hydrocarbon reservoir distribution in the Illinois basin. In:Leighton, M.W. (Ed.), Interior Cratonic Sag Basins. American Association ofPetroleum Geologists, Tulsa.

Johnson, W.H., 1990. Ice-wedge clasts and relict patterned-ground in central Illinoisand their environmental significance. Quat. Res. 33 (1), 51–72.

Jones, E.J.P., Voytek, M.A., Warwick, P.D., Corum, M.D., Cohn, A., Bunnell, J.E., Clark, A.C.,Orem, W.H., 2008. Bioassay for estimating the biogenic methane-generatingpotential of coal samples. Int. J. Coal Geol. 76, 138–150.

Jouzel, J., Souchez, R.A., 1982. Melting–refreezing at the glacier sole and the isotopiccomposition of the ice. J. Glaciol. 28 (98), 35–42.

Karri, S., Sierra-Alvarez, R., Field, J.A., 2006. Toxicity of copper to acetoclastic andhydrogenotrophic activities of methanogens and sulfate reducers in anaerobicsludge. Chemosphere 62, 121–127.

Killey, M.M., 1998. Illinois' Ice Age Legacy. Illinois State Geological Survey GeoScienceEducation Series, 14.

Kipfer, R., Aeschbach-Gertig, W., Peeteres, F., Stute, M., 2002. Noble gases in lakes andground waters. Rev. Mineral. Geochem. 47, 615–700.

Klump, S., Grundl, T., Purtschert, R., Kipfer, R., 2008. Groundwater and climate dynamicsderived from noble gas, 14C, and stable isotope data. Geology 36, 395–398.

Knight, P.G., 1987. Observations at the edge of the Greenland ice sheet: boundarycondition implications for modelers. Int. Assoc. Hydrol. Sci. Pub. 170, 359–366.

Knight, P.G., 1989. Stacking of basal debris layers without bulk freezing-on: isotopicevidence from West Greenland. J. Glaciol. 35, 214–216.

Kulongoski, J.T., Hilton, D.R., Izbicki, J., 2005. Source and movement of helium in theMorongo groundwater basin: the influence of regional tectonics on crustal andmantle helium fluxes. Geochim. Cosmochim. Acta. 69 (15), 3857–3872.

Larson, G.J., Ehlers, J., Gibbard, P.L., 2003. Large-scale glaciotectonic deformation in theGreat Lakes basin, USA–Canada. Boreas 32, 370–385.

Lemieux, J.-M., Sudicky, E.A., Peltier, W.R., Tarasov, L., 2008. Dynamics of groundwaterrecharge and seepage over the Canadian landscape during the Wisconsinianglaciations. J. Geophys. Res. 113, F01011. doi:10.1029/2007JF000838.

Lerche, I., Yu, Z., Tgorudbakken, B., Thomsent, R.O., 1997. Ice loading effects in sedimentarybasins with reference to the Barents Sea. Mar. Pet. Geol. 14 (3), 227–338.

Leventhal, J.S., Hosterman, J.W., 1982. Chemical and mineralogical analysis of Devonianblack-shale samples from Martin County, Kentucky; Carroll and WashingtonCounties, Ohio; Wise County Virginia; and Overton County, Tennessee, USA. Chem.Geol. 37 (3–4), 239–264.

Linick, T.W., Jull, A.J.T., Toolin, L.J., Donahue, D.J., 1986. Operation of the NSF-Arizonaaccelerator facility for radioisotope analysis and results from selected collaborativeresearch projects. Radio Carbon 28 (2), 522–533.

Lively, R.S., Jameson, R., Alexander, E.C., Morey, G.B., 1993. Radium in the Mt. Simon-Hinkley Aquifer, east-central and southeastern Minnesota. Inf. Circ. Minn. Geol.Surv. 36.

40 M.E. Schlegel et al. / Chemical Geology 287 (2011) 27–40

Lloyd Jr., O.B., Lyke, W.L., 1995. Groundwater Atlas of the United States, Illinois, Indiana,Kentucky. Ohio, Tennessee: Hydrologic Investigations Atlas 730-K. U.S. GeologicalSurvey, Reston.

Ma, L., Castro, M.C., Hall, C.M., 2004. A Late Pleistocene–Holocene noble gaspaleotemperature record in southern Michigan. Geophys. Res. Lett. 31, L23204.

Marksamer, A.J., Person, M.A., Day-Lewis, F.D., Lane Jr., J.W., Cohen, D., Dugam, B., Kooi,H., Willet, M., 2007. Integrating geophysical, hydrochemical, and hydrologic data tounderstand the freshwater resources on Nantucket Island, Massachusetts. In:Hyndman, D.W., Day-Lewis, F.D., Singha, K. (Eds.), Subsurface Hydrology: DataIntegration for Properties and Processes. Geophys. Monogr. Ser., AGU, WashingtonD.C.

Marshall, S.J., Clarke, G.K.C., 1999. Modeling North American freshwater runoff throughthe last glacial cycle. Quat. Res. 52, 300–315.

Martini, A.M., Walter, L.M., Budai, J.M., Ku, T.C.W., Kaiser, C.J., Schoell, M., 1998. Geneticand temporal relations between formation waters and biogenic methane: UpperDevonian Antrim Shale, Michigan Basin, USA. Geochim. Cosmochim. Acta. 62 (10),1699–1720.

McIntosh, J.C., Walter, L.M., 2006. Paleowaters in Silurian–Devonian carbonate aquifers:geochemical evolution of groundwater in the Great Lakes region since the LatePleistocene. Geochim. Cosmochim. Acta. 70, 2454–2479.

McIntosh, J.C., Walter, L.M., Martini, A.M., 2002. Pleistocene recharge to midcontinentbasins: effects on salinity structure and microbial gas generation. Geochim.Cosmochim. Acta. 66 (10), 1681–1700.

McIntosh, J.C., Garven, G., Hanor, J.S., 2011. Impacts of continental glaciation onvariable-density fluid flow and solute transport in a sedimentary basin. Geofluids11, 18–33. doi:10.1111/j.1468-8123.2010.00303.x.

Megonigal, J.P., Hines, M.E., Visscher, P.T., 2005. Anaerobic metabolism: linkages to tracegases and aerobic processes. In: Schlesinger, W.H., Holland, H.D., Turikian, K.K. (Eds.),Biogeochemistry, Treatise on Geochemistry, 8. Elsevier, Amsterdam, pp. 319–405.

OEE (Office of Energy Efficiency), 2008. Canada's Secondary Energy Use and GHGEmissions by Energy Source, Natural Resources Canada. http://oee.nrcan.gc.ca/english. accessed Nov. 2010.

Onstott, T.C., Lin, L.-H., Davidson, M., Mislowack, B., Borcsik, M., Hall, J., Slater, G., Ward,J., Sherwood Lollar, B., Lippmann-Pipke, J., Boice, E., Pratt, L.M., Pfiffner, S., Moser, D.,Gihring, T., Kieft, T.L., Phelps, T.J., Vanheerden, E., Litthaur, D., Deflaun, M., Rothmel,R., Wanger, G., Southam, G., 2006. The origin and age of biogeochemical trends indeep fracture water of the Witwatersrand Basin, South Africa. Geomicrobiol. J. 23,369–414.

Ostrom, M.E., Hopkins, M.E., White, W.A., Me-Vicker, L.D., 1955. Uranium in Illinoisblack shales. Ill. State Geol. Surv. Circ. 203, 15.

Patterson, C.J., Hansel, A.K., Mickelson, D.M., Quade, D.J., Bettis III, E.A., Colgan, P.M.,McKay, E.D., Stumpf, A.J., 2003. Contrasting glacial landscapes created by ice lobesof the southern Laurentide Ice Sheet. In: Easterbrook, D.J. (Ed.), Quaternary Geologyof the United States, INQUA 2003 Field Guide Volume. Desert Research Institute,Reno, pp. 135–153.

Patwardhan, A.M., 2010. The Dynamic Earth System, 2nd ed. PHI Learning PrivateLimited, New Delhi.

Pederson, Stig A.S., 2005. Structural analysis of the Rubjerg Knude Glaciotectonic complex,Vendsyssel, northern Denmark. Geol. Surv. Denmark Greenland Bull. 8, 192.

Perkins, R.B., Piper, D.Z., Mason, C.E., 2008. Trace-element budgets in the Ohio/Sunburyshales of Kentucky: constraints on ocean circulation and primary productivity inthe Devonian–Mississippian Appalachian Basin. Palaeogeog. Palaeoclimatol.Palaeoecol. 265 (1–2), 14–29.

Person, M., Dugan, B., Swenson, J.B., Urbano, L., Stott, C., Taylor, J., Willett, M., 2003.Pleistocene hydrogeology of the Atlantic continental shelf, New England. GSA Bull.115 (11), 1324–1343.

Person, M., McIntosh, J., Bense, V., Remenda, V.H., 2007. Pleistocene hydrology of NorthAmerica: the role of ice sheets in reorganizing groundwater flow systems. Rev.Geophys. 45, 28.

Rea, D.K., Moore Jr., T.C., Anderson, T.W., Lewis, C.F., Dobson, D.M., Dettman, D.L., Smith,A.J., Mayer, L.A., 1994. Great Lakes paleohydrology: complex interplay of glacialmeltwater, lake levels, and sill depths. Geology 22, 1059–1062.

Rice, D.D., 1992. Controls, habitat and resource potential of ancient bacterial gas. In:Vially, R. (Ed.), Bacterial Gas Editions. Technip, Paris, pp. 91–120.

Rowan, E.L., de Marsily, G., 2001. Infiltration of Late Paleozoic evaporative brines in theReelfoot rift: a possible source for Illinois basin formation waters and MVTmineralizing fluids. Pet. Geosci. 7, 269–279.

Rowan, E.L., Goldhaber, M.B., Hatch, J.R., 2002. Regional fluid flow as a factor in thethermal history of the Illinois Basin: constraints from fluid inclusions and thematurity of Pennsylvanian coals. AAPG Bull. 86 (2), 257–277.

Ruff, L., LaForge, R., Thorson, R.,Wagner, T., Goudaen, F., 1994. Geophysical Investigation ofthe Western Ohio-Indiana Region, Final Report 1986–1992. U.S. Nuclear RegulatoryCommission: NUREG/CR-3145, 10, p. 73.

Schlegel, M.E., McIntosh, J.C., Bates, B.L., Kirk, M.F., Martini, A.M., 2011. Comparison offluid geochemistry and microbiology of multiple organic-rich reservoirs in asedimentary basin: evidence for controls on methanogenesis and microbialtransport. Geochim. Cosmochim. Acta. 75, 1903–1919.

Schoell, M., 1980. The hydrogen and carbon isotopic composition of methane fromnatural gases of various origins. Geochim. Cosmochim. Acta. 44, 649–661.

Siegel, D.I., 1991. Evidence for dilution of deep, confined ground water by verticalrecharge of isotopically heavy Pleistocene water. Geology 19, 433–436.

Siegel, D.I., 1992. Discussion of “Geochemical evolution of the Cambrian–Ordovicianaquifer, eastern Wisconsin: 1. Major ion and radionuclide distribution” and“Geochemical evolution of the Cambrian–Ordovician sandstone aquifer, easternWisconsin: 2. Correlation between flow paths and groundwater chemistry”.Ground Water 30 (2), 273-273.

Siegel, D.I., Mandle, R.J., 1984. Isotopic evidence for glacial melt water recharge to theCambrian–Ordovician Aquifer, north-central United States. Quat. Res. 22, 328–335.

Souchez, R., Lemmens, M., Lorrain, R., Tison, J.-L., Jouzel, J., Sugden, D., 1990. Influence ofhydroxyl-bearingminerals on the isotopic composition of ice from the basal zone ofan ice sheet. Nature 345, 244–246.

Stevenson, D.L., Dickerson, D.R., 1969. Organic geochemistry of the New Albany Shale inIllinois. Illinois State Geological Survey, Illinois Petroleum, 90. Urbana.

Stichler, W., Schotterer, U., 2000. From accumulation to discharge: modification of stableisotopes during glacial and post-glacial processes. Hydrol. Process. 14, 1423–1438.

Stichler, W., Schotterer, U., Frohlich, K., Ginot, P., Kull, C., Gaggeler, H., Pouyaud, B., 2001.Influence of sublimation on stable isotope records recovered from high-altitudeglaciers in the tropical Andes. J. Geophys. Res. 106 (D19), 22,613–22,620.

Strąpoć, D., Mastalerz, M., Eble, C., Schimmelmann, A., 2007. Characterization of theorigin of coalbed gases in southeastern Illinois Basin by compound-specific carbonand hydrogen stable isotope ratios. Org. Geochem. 38, 267–287.

Stueber, A.M., Walter, L.M., 1991. Origin and chemical evolution of formation waterfrom Silurian–Devonian strata in the Illinois basin, USA. Geochim. Cosmochim.Acta. 55, 309–325.

Stueber, A.M., Walter, L.M., 1994. Glacial recharge and paleohydrologic flow systems inthe Illinois Basin — evidence from chemistry of Ordovician carbonate (Galena)formation waters. GSA Bull. 106 (11), 1430–1439.

Stueber, D.I., Walter, L.M., Huston, T.J., 1993. Formation waters from Mississippian–Pennsylvanian reservoirs, Illinois Basin, USA— chemical and isotopic constraints onevolution and migration. Geochim. Cosmochim. Acta. 57 (4), 763–784.

Swann, D.H., 1968. A summary geologic history of the Illinois Basin. Geology andPetroleum Production of the Illinois Basin. Schulze Printing Co, Evansville, pp. 3–21.

Tarasov, L., Peltier, W.R., 2007. The co-evolution of continental ice cover and permafrostextent over the last glacial–interglacial cycle in North America. J. Geophys. Res. 112,F02S08. doi:10.1029/2006JF000661.

Taylor, S.R., McLennan, S.M., 1985. The Continental Crust: Its Composition andEvolution. Blackwell Scientific, Oxford.

Taylor, S., Feng, X., Kirchner, J.W., Osterhuber, R., Klaue, B., Renshaw, C.E., 2001.Isotopic evolution of a seasonal snowpack and its melt. Water Resour. Res. 37 (3),759–769.

Teller, J.T., 1990. Volume and routing of late-glacial runoff from the southern LaurentideIce Sheet. Quat. Res. 34, 12–23.

Torgersen, T., 1980. Controls on pore-fluid concentration of 4He and 222Rn and thecalculation of 4He/222Rn ages. J. Geochem. Explor. 13, 57–75.

Torgersen, T., 2010. Continental degassing flux of He-4 and its variability. Geochem.Geophys. Geosys. 11, 6.

Torgersen, T., Clarke, W.B., 1985. Helium accumulation in groundwater: I. An evaluationof sources and the continental flux of crustal 4He in the Great Artesian Basin,Australia. Geochim. Cosmochim. Acta. 49, 1211–1218.

Torgersen, T., Kennedy, B.M., 1999. Air-Xe enrichments in Elk Hills oil field gases: role ofwater in migration and storage. Earth Planet. Sci. Lett. 167 (3–4), 239–253.

Tuttle, M.L.W., Breit, G.N., Goldhaber, M.B., 2003. Geochemical data from New AlbanyShale, Kentucky: a study in metal mobility during weathering of black shales. U.S.Geological Survey Open-File Report 03-207, p. 60.

Wood, J.R., Forman, S.L., Pierson, J., Gomez, J., 2010. New insights on Illinoian deglaciationfrom deposits of Glacial Lake Quincy, central Indiana. Quat. Res. 73, 374–384.

Yde, J.C., Knudsen, N.T., 2004. The importance of oxygen isotope provenance in relationto solute content of bulk meltwaters at Imersuaq Glacier, West Greenland. Hydrol.Process. 18, 125–139.

Zhou, Z., Ballentine, C.J., 2006. 4He dating of groundwater associated with hydrocarbonreservoirs. Chem. Geol. 226, 309–327.

Zhou, Z., Ballentine, C.J., Kipfer, R., Schoell, M., Thibodeaux, S., 2005. Noble gas tracing ofgroundwater/coalbed methane interaction in the San Juan Basin, USA. Geochim.Cosmochim. Acta. 69 (23), 5413–5428.

Zuppann, C.W., Keith, B.D., Keller, S.J., 1988. Geology and Petroleum Production of theIllinois Basin vol 2. Illinois and Indiana. Kentucky Geol. Soc. 272.