. . I. 4.?- [h ‘ "1‘. 2: m... "3"22-2 :c' w"; v; . :7 ~ . I‘ “.3; as}? hi‘l‘f c“ 3% mack “1 a. . H: “:5 . ‘ = 1.“; at"... 't~ ‘: LU. _ . fl ‘ .. 554-. ‘W-‘Z’w‘rz ' hr . w a»; 1U *4 ‘N ‘. E4” u 5' , ' 9:3». lg.‘ L‘ “.331. ' ‘v‘ ‘3. ~. _ ‘ ‘ u. v’. 'af-S'J _ v ‘ ‘ “‘43:.“ 1’ ’1 1 ~ ’1‘. X. 91“ w $.51 i . “Marya: , [Egg-«2‘» . '34?!» 1.: y; m 4 A J . «Qiflq hing, ’3‘ 4 E: ‘ r1. .m. ‘ I 4 ‘.§1 .1 r it: “I (h ‘1: ‘ ' . a A' Z in , , : 14"}; 33%}! ”:53: ”it. ‘5' £13"; .2 V“ ‘9 . {1' v .W j‘t“ :' "fi‘k - 35.“ w . h. ‘nv: m ~v~ TE NWERS mmlllllllllulmluIllllllll‘ll 3 1293 01021 9735 This is to certify that the thesis entitled Geochemical Mass Balance Models of Sandstone Weathering in the Pennsylvanian Recharge Beds of South-Central Michigan presented by Jason Rodney Price has been accepted towards fulfillment of the requirements for M. S. degree in Geological Sciences fl/Léé/ZMQ M/é/ Major professor Date 7/f/V7é/ 0—7639 MS U is an Affirmative Action/Equal Opportunity Institution LEERARV Michigan State University PLACE II RETURN BOXto manuals mum ywrrocord. TOAVOID FlNESMumonorbdondatodm. MSU lsAnNflrmntlvo Action/Emu! Oppomnlty Indium: W1 GEOCHEMICAL MASS BALANCE MODELS OF SANDSTONE WEATHERING IN THE PENNSYLVANIAN RECHARGE BEDS OF SOUTH-CENTRAL MICHIGAN BY Jason Rodney Price A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Geological Sciences 1994 ABSTRACT GEOCHEMICAL MASS BALANCE MODELS OF SANDSTONE WEATHERING IN THE PENNSYLVANIAN RECHARGE BEDS OF SOUTH-CENTRAL MICHIGAN BY Jason Rodney Price Outcrop Pennsylvanian sandstones located near Grand Ledge, Michigan.were studied.in order to evaluate the chemical and textural effects of weathering. The exposure of these sandstones is the result of post-Pleistocene river down- cutting, and reflect weathering since that time. A comparison is made to subsurface Pennsylvanian sandstones which serve as pre-weathering analogs. Mass balance calculations suggest that iron is being conserved in the sandstone with the joint block interiors serving as net exporters of iron to the 1 cm thick case- hardened joint faces. All other major cations, including A1“, are being mobilized from the outcrop through dissolution. The mass balance calculations for each ionic species are compared with shallow aquifer water chemistry data. This comparison suggests that quartz, carbonates, and pyrite are currently altering in the subsurface. Outcrop K-feldspar appears stable, while muscovite and kaolinite are weathering and the aluminum is being mobilized into the aquifer. DEDICATION This thesis is dedicated to a dear friend and mentor, Mr. Richard C. Noreen. Dick’s sudden hospitalization in November 1993, and death in December 1993, resulted in the loss of a valued member of the NIU and DeKalb communities. His genuine interest in people, unending willingness to offer a helping hand, and sincere concern for both colleagues and students at Northern Illinois University will long be remembered. Thanks for all the perverted jokes, meals, gifts, and. general unselfishness over the years, buddy. You'll never be replaced. iii ACKNOWLEDGMENTS The individual who could not possibly be thanked enough for his help, is my thesis advisor, Dr. Michael Velbel. His humble disposition, cynical sense of humor, and tactful, patient criticisms are tremendously appreciated Also deserving of recognition are my committee members, Dr. Sibley and Dr. Westjohn. Dr. Sibley always returned no- nonsense answers to my all too often spontaneous questions. Dru Westjohn.provided voluminous material on the subsurface of the Michigan basin, including beautiful thin sections, clay mineral data, and diagenesis research. A sincere thank you to my parents who empathetically and patiently listened to my stories of frustration, as well as indirectly funding a large portion of my education. Noteworthy, too, is Tina Beals who was kind enough to perform the silica analyses of my Grand Ledge spring water. I also cannot forget all of my fellow graduate students who served as both friends and colleagues during my tenure at Michigan State. Kris Huysken, Ditters, Wei Huang, Bill Sitarz, Steve Reigel, Jolene & Bruce Meissner, Cheol Woon Kim, Jaeman. Lee, Jon, Kolak, and.'many' others too :numerous to mention. Thanks folks! Partial funding for this project was provided by the Michigan Basin Geological Society. iv TABLE OF CONTENTS LIST OF TABLES .......................................... vii LIST OF FIGURES ......................................... ix CHAPTER 1: INTRODUCTION Purpose and Scope ............................... 3 Previous Work ................................... 5 Surficial Geology ....................... 5 Subsurface Geology ...................... 12 Hypothesis ...................................... 16 CHAPTER 2: PETROGRAPHY Sample Collection ............................... 18 Methods ......................................... 18 Petrographic Description and Interpretation ..... 19 Quartz .................................. 25 K—Feldspar .............................. 3O Muscovite ............................... 30 Clay Minerals ........................... 31 Pyrite .................................. 35 Ferruginous Oxy-Hydroxides .............. 37 Comparison With Subsurface Petrography .......... 44 Carbonate ............................... 45 Quartz .................................. 48 K-Feldspar .............................. 48 Clay Minerals ........................... 48 Pyrite .................................. 53 Ferruginous Oxy-Hydroxides .............. 53 CHAPTER 3: CLAY MINERALOGY Methods ......................................... 54 Results ......................................... 55 CHAPTER 4: MASS BALANCE MODELING Method of Calculation ........................... 57 Mineral Compositions ............................ 60 Pyrite .................................. 60 Carbonates .............................. 61 Kaolinite .............. 1 ................. 61 Goethite, Muscovite, and Quartz ......... 62 Balanced Chemical Reactions ..................... 63 Carbonate Equilibria .................... 63 Silica Equilibria ....................... 65 Aluminosilicate Solubility .............. 65 Ion Mass Balance ................................ 66 Modal Mineral Changes ................... 67 Precipitation Inputs .................... 67 Joint Face vs. Joint Block Interior ..... 69 Comparison With Subsurface Water Chemistry and Mineralogy ................................. 72 Mineral Stability ....................... 72 Long Term vs. Instantaneous Water Chemistry .......................... 74 CHAPTER 5: UNCONFORMITIES AND HYDROCARBON RESERVOIRS....82 CHAPTER 6: HONEYCOMB WEATHERING ......................... 84 CHAPTER 7: SUMMARY AND CONCLUSIONS Summary ......................................... 88 Conclusions ..................................... 9O APPENDICES Appendix A ...................................... 92 Appendix B ...................................... 94 Appendix C ...................................... 95 Appendix D ...................................... 99 BIBLIOGRAPHY ............................................ 101 vi LIST OF TABLES Table 1. Comparison of subsurface and outcrop mineralogy of Carboniferous sandstones in the Michigan basin .......................... 6 Table 2. Outcrop Pennsylvanian sandstone modal mineralogies ................................... 24 Table 3. Subsurface Pennsylvanian sandstone modal mineralogies ................................... 44 Table 4. Report on carbonate abundances in subsurface thin sections ....................... 45 Table 5. Summary of XRD data from the outcrop Pennsylvanian Eaton sandstone .................. 56 Table 6. Summary of U.S.G.S. subsurface Pennsylvanian sandstone XRD data ............... 56 Table 7. Mineral data for phases used in the mass balance model ......................... 62 Table 8. Chemical reactions used for mass balance modeling ............................... 63 Table 9. Mass balance calculations for joint face and joint block interior .................. 66 Table 10. Precipitation data from Lansing and East Lansing, Michigan ........................ 68 Table 11. Mineral stability comparison for outcrop, glacial till, and Pennsylvanian aquifers ........................ 73 Table 12. Water chemistry data from the Grand River aquifer near Grand Ledge, Michigan ............................... 77 vii Table 13. Table 14. Table 15. Water chemistry data from the Saginaw aquifer near Grand Ledge, Michigan ............................... 78 Long term vs. present day water chemistry in Pennsylvanian sandstones ......... 78 Sample collection data ........................ 94 viii LIST OF FIGURES Figure 1. Partial stratigraphic column of the Michigan basin ................................ 2 Figure 2. Geology of south-central Michigan and index map of the Grand Ledge area .......................................... 4 Figure 3. Ternary diagram showing the composition of the Eaton sandstone ............ 9 Figure 4. (a) Photomicrograph of the quartzose Eaton sandstone ............................... 20 Figure 4. (b) Photomicrograph of a subsurface Upper Pennsylvanian sandstone ..................................... 21 Figure 5. (a-b) Photomicrographs of the typical Eaton sandstone ............................... 22 Figure 6. Scanning electron micrograph of weathered quartz showing corrosion of quartz over-growths ........................ 26 Figure 7. Photomicrograph of a diagenetically altered K-feldspar with etch pits ............. 29 Figure 8. (a) Photomicrograph of altering muscovite showing compaction features and alteration to vermiculite ................................... 32 Figure 8. (b) Scanning electron micrograph of parental muscovite with exfoliating vermiculite layers ............................ 33 Figure 8. (c) Photomicrograph showing evidence of muscovite dissolution ............. 34 Figure 9. Photomicrograph of a typical diagenetic clay patch ......................... 36 Figure 10. (a-b) Photomicrographs displaying evidence of kaolinite dissolution ..................... 38 ix Figure 10. (c-d) Scanning electron micrographs displaying evidence of kaolinite dissolution .................................. 40 Figure 11. Scanning electron micrograph of a vermicular kaolinite in the 1 cm case—hardened joint face ..................... 42 Figure 12. Photomicrograph of what is believed to be pyrite crystals ........................ 43 Figure 13. (a-b) Photomicrographs of subsurface "pre-weathering" carbonates displaying both poikilitic carbonate and rhombs ................................... 46 Figure 14. (a-b) Photomicrographs showing iron- bearing carbonates altering to iron oxy- hydroxides in the subsurface ................. 49 Figure 15. (a) Photomicrograph of a subsurface quartz grain altering to chlorite ............ 51 Figure 15. (b) Photomicrograph of the Eaton sandstone showing result of weathering the quartz-hosted chlorite above ............................... 52 Figure 16. Percent goethite vs. distance from joint face ................................... 70 Figure 17. (a-b) Selected diffractograms ................ 92 Figure 18. (a—d) Photos of outcrop sampling ............. 95 CHAPTER 1 : INTRODUCTION This investigation examines the weathering of the Upper Pennsylvanian (Conemaugh?) Eaton sandstone (Figure 1) which crops out near Grand Ledge, Michigan (SW 1/4, section 2, T 4 N, R 4 W, Oneida Township, Eaton County) (Figure 2). These outcrops are the result.ofjpost-Pleistocene river down—cutting and, therefore, exhibit weathering, much of which is likely post—glacial» These and other Pennsylvanian sandstones in the subsurface of the Michigan basin presently produce fresh- potable waters, with the outcropping and near surface sandstones providing a recharge zone for meteoric water. Paragenetic sequences and dissolution textures of authigenic minerals in Pennsylvanian sandstone aquifers suggest that basin-wide evolution of a brine produced the authigenic mineral suite prior to post-Pleistocene time (Westjohn et al., 1991; Westjohn & Sibley, 1991). Therefore, the pre-Pleistocene authigenic mineral suite of the Pennsylvanian aquifers in the Michigan basin may be viewed as the pre-weathering mineralogy of the subaerially exposed Eaton sandstone. a 8 5 If, ‘g 8 Glacialion Stratigraphic Unit Hydrogeologic Unit (1 m 0) C d) U 2 ’2 2 :6 g E Glacial dritt o a: . g 1;; 2 aQurfers o 6 8 Wisconsin 3 lllinoian .9 Pre-Illinoian 2 a. 2 .2 3 3’. o Unnamed Glacial till-red beds § 3 3; red beds confining unit 2 9: c a ------ ““‘f “““““ .2! E Grand my“ loma. Eaton. and 5 2 Formation Woodville Sandstone _ 2 5 Members Grand River- 3 __ ““‘ ‘ u ‘ ““““““““““““““““““ Saginaw E I}? Saginaw Parma Sandstone agualer 8 (5 Formation Member Bayport 3 8:13:31 ”“9510“ Bayport-Michigan cu . . confining unit _. Group Michigan 5 AAAAAAAAAAAA - AAAAA formation AAAAAAAAAAAAAAAAAAAA 3 ' .9 Marshal Sandstone Napoleon Sandstone Marshall aquiler 9 g Member .9; 3 Goldwater Shale £ 2 >. Sunbury 3-," 5“” >Eastern 0.! Bare. Michigan Sandst n . o e Coldwater-Antnm ‘ confining unit 2 Ellsworth Western l c Shale Michigan 6 'a 5 Bedtord Eastern 7 90?: Shale Michigan ”co £Q> 0 o g 0 Armin Shale Figure 1. Partial stratigraphic column of the Michigan basin. (From Dannemiller & Baltusis, 1990.) 3 PURPOSE AND SCOPE The purpose of this study is to compare the authigenic mineralogy of subaerially exposed Pennsylvanian sandstones with the age-equivalent subsurface rocks found deeper in the Michigan basin (Table 1). Although it has been established that the Pennsylvanian sequence near Grand Ledge is the most extensive natural outcrop of rocks of such age in Michigan (Kelly, 1933), little work has been performed to identify the effects of weathering, and no attempt has been made to relate the authigenic mineralogy of the subsurface age-equivalent rocks to the alteration of the subaerially exposed sandstones. Outcrop characteristics of sandstones are important in the evaluation of certain categories of potential reservoirs fom' oil and. gas. If correct conclusions on reservoir potentials are to be made, the effects of weathering on the sandstone must be established. Weathered zones exhibit significant secondary porosity which, when trapped below unconformities, may provide potential hydrocarbon reservoirs (Heald et al., 1979; Shanmugam & Higgins, 1988; Shanmugam, 1988, 1990). By comparing the effects of weathering with the pre-weathering mineralogy established from subsurface drill cuttings, this investigation will provide insight into the nature of porosity evolution at unconformities. Clinton County + Eaton County N \ l \ \\ \‘ 43' m °" Fitzgerald Donia Park a... M (3 4 83' 86' 35° 34° 0 40 80 m Sec. 10 0 80 KM Figure 2. Geology of south-central Michigan and index map of the Grand Ledge area. Dots with numbers represent outcrop sample locations (Appendix B). (Hashed area is Pennsylvanian Saginaw formation, stippled area is Pennsylvanian Grand River formation, and the inverted wave area is Upper Jurassic rocks; modified from Dannemiller and Baltusis, 1990, Figures 1 & 2; and Martin, 1982, Figure 1.) 5 PREVIOUS WORK Surficial Geology Eaton Sandstone The Eaton sandstone of the Grand River Formation forms the ledges of the Grand River and its tributary, Sandstone Creek, in the northern part of Eaton and southern part of Clinton counties, Michigan (Figure 2). Being the most extensive natural exposure of Pennsylvanian strata in the state of Michigan, the Eaton sandstone provides an important resource for investigations into many aspects of Pennsylvanian geology, including the diagenesis and weathering of Pennsylvanian sandstones. It is a porous, thickly bedded, medium grained, buff—colored quartz arenite to subarkose (Figure 3), having a maximum thickness in outcrop of approximately 18 m (Kelly, 1933; Hudson, 1957; Martin, 1982) . Diagenetic alterations include quartz cement and feldspar alteration (Martin, 1982) (Table l) . Honeycomb weathering is well developed in the Eaton sandstone, and is apparently a function of the presence of salts on the outcrop surface, the aspect of the outcrop, and the massiveness (homogeneity of texture and fabric) of the unit (Wallis & Velbel, 1985). The same researchers also noted the presence of limonite cements forming a well—indurated 1 cm thick zone along joint surfaces. The limonite cement must have formed following sufficient induration of the sandstone to allow brittle deformation. Velbel & Genuise (1988) have studied the non-bedded mudrocks Table 1. Comparison of subsurface and outcrop mineralogy of Carboniferous sandstones in the Michigan basin. A. PETROGRAPHIC MINERALOGY Detrital Outcrop Subsurface Win—WW Quartz x2 x6 xu K-Feldspar X2 x6 X” Rock Fragments X2 x6 Muscovite x2 X6 xu Chlorite X12 Authigenic Outcrop Subsurface Melon—WWW Anhydrite X” xm Ankerite x5.8.10 X8,1o Barite X8 Xe Calcite XS-lo X8-12 Chlorite x6,7,8,10 x3-11 Dolomite x5.7.8.10 xe.1o,12 Feldspar xLlO qun Glauconite X9 x3 Gypsum x7,a,1o xs,1o,12 Illite x13 X6,7,10 X9,1o,11 Iron Oxy-Hydroxides X2(Goethite)13 x7!“o x8o1°.12 KaOlinite x13 x6,7,8,10 x8-12 Mixed Clays X8 X8 Rhodocrosite Xa1° X10 Siderite X&7J° X10 Quartz X1.2. 13 x8, 10, 13 x3-12 Witherite Xi1° X10 Vermiculite X13 Table 1. (cont’d). B. HEAVY MINERALOGY Detrital Outcrop Subsurface Mineralogy Pennsylvanian Pennsylvanian Mississippian Actinolite X” Apatite X2 Biotite X” Cassiterite X1 Chlorite X” Epidote X” Garnet X1 X” Hornblende X” Ilmenite X” Kyanite X1 Leucoxene X” Magnetite X” Monazite X1 Pyroxene“ X2 X” Rutile X” Staurolite X1 Tourmaline Xl'z'4 X6 X” Zircon X1'2'4'” X6 X” “Pyroxene in outcrop is pigeonite, and in the subsurface is enstatite and hvnersthene. Authigenic Outcrop Subsurface Mineralogy Pennsylvanian Pennsylvanian Mississippian Celestite X” Chlorite XELBJ° X141 Pyrite .Xfi7“° XE?” Hudson, 1957 Martin, 1982 Davis & Bredwell, 1978 Kelly, 1936 Kramer & Westjohn, 1991 Westjohn, written communication Long et al., 1988 Westjohn et al., 1990 Westjohn & Sibley, 1991 Westjohn et al., 1991 Zacharias, 1992 Stearns, 1933 This study \DQQGUIDWNH HHHH UNHO 8 which occur as lenses in the Eaton sandstone and determined them to consist of kaolinite, illite, lepidocrocite, minor amounts of chlorite, and interstratified illite-vermiculite. The lepidocrocite is believed to be related to local groundwater flow in the Eaton sandstone (Velbel & Brandt, 1989). Studies of Other Basins General" Numerous authors have described subaerial diagenesis. Notable among these is Fairbridge (1967) who suggested the term "epidiagenesis" for surficial weathering and the resulting development of new textures and minerals. Al-Gailani (1981) has described the diagenesis of unconformities with emphasis on authigenic mineral formation at paleo-surfaces and the resulting adverse effects on reservoir characteristics. Emery et al. (1990) have used potassium feldspar leaching by meteoric water and kaolinite abundances to demonstrate the presence of burial unconformities. 'Tardy (1971) and. Bjorlykke (1984) have discussed the importance of secondary porosity by describing potassium feldspar dissolution in freshwater and the associated kaolinite precipitation. Typically, lateral and vertical variations in lithology complicate the study of sandstone weathering by making it difficult to compare weathered and unweathered rock. Furthermore, it is not always clear which secondary alterations are of deep diagenetic origin and which are truly the products of subaerial alteration. In a study of the Quartz Quartz Arenite 0.00 5 o o o 5 O. O 0 '1 O o O Subarkose Sublitharenite O K (a Q O é“ 6.; 4‘” " a K )4 Figure 3. Ternary' diagrant showing the composition. of the Eaton sandstone. Solid circles are outcrop samples and open circles represent subsurface samples. See text for discussion. (Classification of Pettijohn et al., 1987). 10 spheroidal weathering of the Pennsylvanian Kanawha Formation of central and southern.West Virginia, Heald et al. (1979) had great success when comparing the weathered rock to fresh rock cores a short distance away, treating the core mineralogy as pristine and pre-weathering. These researchers believe that an absence of oxidized minerals in outcrop suggests a lack of weathering, and supported this conclusion with subsurface data. In addition, they attributed their outcrop K-feldspar voids to subsurface diagenesis. Weathering patterns similar to those found in the Eaton sandstone at Grand Ledge, Michigan have been described for the Virgelle Member of the Cretaceous-aged Milk River Formation in Alberta, Canada (Campbell, 1991). This investigator found.the sandstone to exhibit alveolar weathering, and thin case- hardened resistant zones of iron-rich varnish, and attributed much of the weathering to ice and salt crystal growth. Geomorphology. Thiry et al. (1988) and Thiry and Milnes (1991) took a geomorphologic approach to sandstone weathering. These researchers found pedogenic and groundwater silcretes, and attributed their formation to the lowering of the water table during river down-cutting. This work implies that quartz cement does not behave as a seal, and as long as groundwater rises (in the case of early diagenesis), or falls (in the case of subaerial weathering) sufficiently slowly, continuous quartz cementation.may proceed as an uninterrupted formation of groundwater silcretes. Furthermore, Thiry et al. (1988) mention that quartz dissolution occurs above the water 11 table and quartz cementation occurs below the water table. Meissner (1993) found that the waters of the Pennsylvanian aquifers in the Michigan basin are saturated with respect to quartz. As Thiry et al. (1988) have demonstrated for the Paris basin, the Michigan basin exhibits quartz dissolution in outcrop, and quartz saturation below the water table. Bromley (1992) suggested that the early formation of quartz in the Navajo sandstone of the Colorado Plateau.was due to the presence of the 'unconformably' underlying' Kayenta mudstone which maintained an elevated groundwater level in the early' Navajo sandstone, with. evaporation. inducing' quartz precipitation. However, initial cement accumulation may behave as a seal, preventing further evaporation, and thus cementation (Goudie, 1973; Summerfield, 1983). Pedogenesis. From a pedogenic perspective glauconitic quartzites have been shown to lateritically weather into "red beds" by glauconitic grains weathering to ferruginous ooids and pisolites in ferricretes (Nahon et al., 1980; Parron & Nahon, 1980). vermiculite. Illite may weather to vermiculite (e.g. Adams & Kassim, 1983), but in a study by White (1962), it was demonstrated that the weathering of muscovite produced a 14 A XRD peak, while the weathered illite peak was simply less intense than the unweathered illite peak. Chittleborough (1989) invokes the opposite weathering reaction, whereby illite weathers from vermiculite. 12 Ferruginization. ‘Young (1987) reports concentrations of goethite along joint faces of sandstones in the East Kimberley region of Australia, while Nott et al. (1991) found two stages of ferruginization of the Long Beach formation of Australia. Both pyrite weathering and CO2 dissolution have been shown to occur in sandstones in a humid region of Japan (Chigira and Sone, 1991), who demonstrated that the formation of iron oxy-hydroxide cement in the oxidation front of the outcrop strengthens the rocks, while beneath this front dissolution of cements weakens the sandstone. Weathering of the pyritic Mahoning sandstone of West Virginia has completely dissolved all pyrite and carbonate, creating sufficiently acidic minesoils to inhibit revegetation (Singh et al., 1982). Weed & Ackert (1986) report ferruginous oxy-hydroxide precipitation occurring early in their weathering sequence for antarctic sandstones. Subsurface Geology Carboniferous Strata of.M1chigan Recent work has been conducted on the mineral-water interactions, paragenesis, and diagenesis of the subsurface Pennsylvanian strata.of the Michigan.basin, as the Grand.River Formation is one of the principal bedrock aquifers in the Michigan.basin (Westjohn et al., 1990) (Table 1). ‘Westjohn et al. (1990) demonstrated that the cements in the Pennsylvanian sandstones are mineralogically diverse; cements of the poorly to well-cemented sandstones include silica, calcite, ankerite, dolomite, kaolinite, and iron oxide, and lesser amounts of 13 chlorite, glauconite, barite, mixed-clays, and gypsum cements. Westjohn et al. (1991) observed paragenetic sequences which are the same for both Mississippian and Pennsylvanian sandstones from the Michigan basin, and suggested that the identical authigenic minerals found in these sandstone aquifers are the product of pre-Pleistocene basin-wide chemical evolution of groundwater in Carboniferous strata. A paragenetic sequence for’ the Marshall sandstone (Mississippian) as determined by Stearns (1933) has quartz precipitation, with calcite filling the interstices between sand grains, followed by pyrite, magnetite (possibly marcasite), and small amounts of celestiteu Westjohn.& Sibley (1991) state that there is no evidence that flushing with meteoric water during or since the Pleistocene has altered Mississippian clastic sediment authigenic minerals. Studies by Zacharias (1992) and Zacharias et al. (1992) on Mississippian strata in the Michigan basin interpret isotopic data and mineral paragenesis and suggest that cements (chlorite, carbonate, and kaolinite) did not form in equilibrium with present-day pore-fluids. Furthermore, the illite distribution throughout the unit suggests that it may have formed prior to the differentiation of modern—day interstitial fluids (Zacharias, 1992; Zacharias et al., 1992) . Since kaolinite is the final phase (Westjohn et al., 1990; Zacharias, 1992) of the paragenetic sequence, and.it is not in isotopic equilibrium with modern-day interstitial fluids, it follows that phases precipitated prior to kaolinite are also 14 not in equilibrium with present-day pore fluids. Long et. al. (1990) use major element geochemistry and isotopic chemistry of water from deep formations, near surface bedrock, and glacial drift to suggest that any water-rock interaction presently occurs only in the glacial drift. The interacting water is characterized by high salinities, which are believed to be the result of long residence times of the groundwater, thus allowing time for upward diffusion/advection of formation brine into glacial-lacustrine clay. Meissner et al. (1992) inferred.that dissolved.solids in thelMississippian Marshall sandstone subcrop originated from meteoric water-rock interactions in overlying glacial drift. These same researchers suggest that the evolution of ground water in Mississippian. aquifers by' dilution. of marine brine with meteoric water occurred following geochemical processes such as clay interaction and sulfate reduction, and that this ground. water evolution is ‘very similar to that of the underlying Devonian formations. Wahrer et al. (1992) studied ground water in glacial-drift and near-surface-bedrock aquifers and found that the ground water in the latter is at or near equilibrium with respect to calcite, and possibly dolomite. Studies of Other Basins General. Bjorkum & Gjilsvik (1988) have described the disagreement regarding open vs. closed systems for authigenic 'mineral growth and propose an isochemical model for authigenic kaolinite, potassium feldspar, and illite formation. These 15 authors point out that in.an isochemical system, kaolinite may forntby the degradation of mica or potassium feldspar, and the ap/am.ratio of the pore water will increase and the reaction will reach equilibrium unless either a K+ sink or, a H” source exists. For the Carboniferous sandstones in the Illinois basin, Nesbitt (1980) has suggested that subsurface sodium ions are incorporated in clay minerals rather than bonding to weak neutral complexes in shales. Ferruginizationm McBride (1987) provides a case history of the diagenesis of a subarkose sandstone which, like the Eaton sandstone, contains limonite cement, displays upward- fining of grain size, is a fluvio-deltaic deposit, is well sorted and rounded, and bioturbated. In addition, four plausible diagenetic pathways are presented, with final uplift and weathering resulting in the oxidation of siderite to limonite, goethite, and hematite, as well as the dissolution of calcite producing porosities of up to 20%. Arditto (1983) conducted a study on the mineral-water interactions of the intake beds of the Great Australian (Artesian) basin, and determined that weathering of subsurface siderite cement explains the development of secondary limonite and goethite over the outcrop exposure. Furthermore, the same researcher noted lateral anisotropy with respect to authigenic kaolinite in both outcrop and in the subsurface, and found very porous, water—saturated zones which are generally coated with red- brown iron-oxides. Arditto (1983) believes the more porous zones represent either primary porosity or secondary porosity 16 by the leaching of carbonate-cemented zones. Quartz dissolution. A study by Morris and Fletcher (1987) found that redox reactions involving iron result in a much more rapid dissolution of quartz than would be predicted from the known solubility of quartz in water. These researchers conclude that in a ferrous iron solution a single- layer ferrous iron/silica complex forms on the quartz grain surface; this layer breaks down under oxidizing conditions, resulting in a rapid release of silica to solution. In essence, an absence of oxidized minerals suggests a lack of weathering. HYPOTHESIS The hypothesis tested here is that subaerial exposure of the Pennsylvanian Eaton sandstone has produced an authigenic mineralogy that can be explained in terms of an open chemical system (allochemical) with a high freshwater flux. This is in contrast to the differing authigenic mineralogy of the subsurface age-equivalent rocks found deeper in the Michigan basin, which may be explained in terms of a closed chemical system (isochemical) with a low water flux. Since the surficial sandstone outcrops are exposed to precipitation, it is expected that ions produced by the degradation of minerals will be transported out of the system. If appreciable weathering of the Eaton sandstone has occurred, the alteration should be evident in thin section. 17 Weathering textures may include quartz grains and quartz overgrowths displaying corrosion textures, dissolved calcite, and leached feldspar with.concomitant kaoliniteu Since illite has been found in the subsurface (Table 1) it may be found weathering to kaolinite or ferruginous oxy—hydroxides, or may be leached away completely. In general, weathering tends to decrease the amount of Si, K, and Mg present, while increasing the quantities of Al and Fe, and. producing oxides and hydroxides. CHAPTER 2: PETROGRAPHY SAMPLE COLLECTION Sampling was performed at 4 different locations in the vicinity of Grand Ledge, Michigan (Figure 2, Appendix C). Additional thin sections from a previous worker were also used. At each of the 4 locations 4 to 6 samples were taken starting at the joint face and working toward the interior of the joint block. The orientation of the joint plane, aspect of the outcrop, trend of sampling line, and distance from the joint face that each sample was taken was recorded (Appendix B). The subsurface thin sections were provided by the U.S. Geological Survey, Lansing, Michigan. All samples with a sample name starting with "B" are from wells drilled in the Bunkerhill area, Ingham County; all those samples starting with an "S" are from the Standish area, Arenac County. METHODS Thin sections observed in this study (43 total) were impregnated with blue epoxy. The sandstones were first impregnated, then cut, with the cut side of the billet being 18 l9 adhered to the slide, so that any plucking of framework grains would create an isotropic hole in the epoxy and be readily distinguishable by the petrographer. Petrographic evaluation and point counting of the sandstones was performed on a Nikon Labophot-Pol microscope. A.mechanical stage was used to advance the thin section and to record the coordinates of noteworthy features. A grid reticule was inserted into the right ocular for the purpose of making petrographic measurements and point counting. PETROGRAPHIC DESCRIPTION AND INTERPRETATION The suite of thin sections exhibits extremely little variability. The sandstones are typically' very porous (averaging 22.9%, Table 2), with mild alteration of detrital phases, patches of clay, quartz cement, sparse detrital zircon, and limonite coatings on all of the above. The regions of highest porosity are often devoid of limonite coatings, and in these areas detrital grains are commonly rounded (Figure 4a), as opposed to less porous areas of grain angularity (Figure 5a-b). Quartz dissolution features are manifested in thin section and under the SEM as corroded quartz overgrowths (Figures 5 & 6) . Point count data indicate an average of 1.8% of quartz dissolution has occurred. All phases, including clay, exhibit dissolution features at some location, although limonite may be found coating dissolution features in detrital and authigenic phases. Dissolution 20 Figure 4 . (a) Photomicrograph of the quartzose Eaton sandstone. Notice region devoid of goethite (limonite) , and associated grain roundness. Sample JP-95—21. Crossed-polars; field of view is 1.5 mm across . Figure 4. 21 (b) Photomicrograph of a subsurface Upper Pennsylvanian sandstone. Notice similar grain rounding as in (a), as well as carbonate and quartz cements. U.S.G.S. sample J3—81. Crossed—polars; field of view is 2.9 mm across. (Photo courtesy of U.S. Geological Survey, Lansing, Michigan) Figure 5. 22 (a) Photomicrograph of the typical Eaton sandstone. Notice goethite (dark brown material), high porosity, microcline, and grain angularity. Sample JP—95-1. Plane-polarized light; field of view is 1.5 mm across. 23 Figure 5. (b) Same view as in (a) but under crossed—polars. Goethite is birefringent orange material. 24 Table 2. Outcrop Pennsylvanian sandstone modal mineralogies. Inches from rite Lithica Goethite ‘Other JP-95-1A 0.00 61.5 6.6 0 2.9 1.2 0 0 2.2 19.2 1.0 JP-95-1B 0 75 61 5 25 8 0 2.9 1.2 0 0 2.2 5.4 1.0 JP-95-2 2 75 67 6 21.5 0 3.4 1.7 0.3 0 0.8 4.2 0.6 JP-95-3 5 25 64 8 20.3 0 2.6 1.0 0 0 1.3 8.9 1.1 JP-95-4 7 50 68 8 23.6 0 1.8 1.0 0 0 1.5 2.0 1.5 JP-95-5 10 00 60 0 31.0 0 3.6 1.0 0 0 1.9 2.4 0.2 JP-95-6 13 25 64 2 25.8 0 3.1 1.4 0 0 1.1 2.5 1.9 JP-95-7A 0 00 56 7 5.3 0 3.7 0.0 0 0 1.0 21.3 0.7 JP-95-7B 0 75 56 7 5.3 0 3.7 0.0 0 0 1.0 11.3 0.7 JP-95-8 3 00 67 0 16.2 0 1.7 0.3 0 0 0.8 13.7 0.3 JP-95-9 5 25 58 S 25 0 0 1.2 0.2 0.2 0 0.5 14.4 0.0 JP-95-10 7.50 64.1 18 4 0 1.1 0 0 0 2.5 13.3 0.6 JP-95-11 9.75 61.9 24.9 0 3.5 0 0 0 0.8 7.5 1.3 JP-95-12A O 00 59 7 12.1 0 1.3 0.5 0 0 1.1 20.0 0.8 JP-95-12B 0 75 59 7 32.1 0 1.3 0.5 0 0 1.1 4.5 0.8 JP-95-13 3 13 66 4 24 4 0 1.5 0.6 0 0 0.9 5.9 0.3 JP-95-14 5 25 65 4 20 4 0 1.4 1.4 0 0 0.3 10.6 0.6 JP-95-15 7 38 69 1 19.7 0 1.4 1.4 0 0 1.4 6.6 0.3 JP—95-16A 0 00 66 8 4.7 0 1.4 0.6 0 0 0.9 18.4 0.3 JP-95-16B 1 00 66 8 23.1 0 1.4 0.6 0 0 0.9 6.9 0.3 JP-9S-17 3 75 64 7 19 4 0 1.2 2.6 0.6 0 2.6 8.1 0.9 JP-95-18 6 00 64 4 26.2 0 2.7 0.8 0.5 0 1.6 3.5 0.3 JP-95-19 8 25 64.5 24.9 0 2.6 0.3 0.3 0 0.6 6.3 0.6 JP-9S-20 10.75 73.6 17 4 0 1.0 0.7 0 0 0 7.0 0.3 JP-95-21 13.00 66.4 29 2 0 2.1 0.6 0 3 0 0 1.2 0.3 JP-95-22 -- 73.2 20 0 O 0.6 1.3 0 0 2.3 2.6 0.0 JW-1-3 -- 69.1 20.1 0 1.5 1.2 0 0 1.2 5.8 1.2 JW-3-4 -- 68.7 18 9 0 0.8 0.8 0.8 0 2.2 7.5 0.3 JW-l-l -- 74.6 20 5 0 1.6 0.6 O 0 1.6 0 1.0 JW-13-2 -- 74.8 21 5 0 1.9 0.3 0 0 0.9 0 0.6 JW-24-6 -- 69.1 22 2 0 2.1 0.3 0 0 0.6 5.7 0.0 JW6-3(6) -- 61.1 26 1 0 1.6 1.1 0 0 2.6 3.5 4.0 JW-S-8 -- 68.7 18.1 0 1.6 0.8 0.8 0 2.1 7.6 0.3 JW-E-1-3 -- 64.3 26.5 0 1.7 0.6 0 0 1.1 5.0 0.8 JW-23 -- 67.5 25 S 0 2.0 0.8 0.8 0 0.3 2.8 0.3 JW-25 -- 65.1 24.4 0 1.4 0.3 0 0 0 8.6 0.3 JW-8(10) -- 72.1 18.7 0 2.1 1.8 0.3 0 1.2 1.8 2.1 JW-0(11) -- 73.9 15.4 0 1.6 0 0 0 1.0 8.0 0.0 JW-l7-8 -- 61.2 25.0 0 3.2 0.6 0.2 0 1.4 8.0 0.4 JW-8(l3) -- 72.5 20.1 0 2.2 0.6 0.3 0 1.0 2.9 0.3 JW-19(14) -- 68.5 14.1 0 0.6 0 0 0 0 16.2 0.6 JW12-17-3 -- 66.7 18.8 0 1.1 0.3 0 0 0.3 12.8 0.0 JW12-17-4 -- 65.6 26.2 0 1.3 1.2 0.2 0 2.5 3.0 0.0 JW12-17-5 -- 64.5 26.1 0 2.7 0.8 0.5 0 1.7 3.4 0.3 JW-ll-S -- 58.9 30.1 0 1.3 0.6 0 0 0.9 7.9 0.3 JW-14-2 -- 61.1 25.9 0 1.5 1.1 0 0 2.4 4.3 3.7 JW-11-2 -- 56.7 26.0 0 3.1 0.3 0 0 1.8 11.0 1.1 Mean 65.9 22.9 0 2.0 0.7 0.1 0 1.2 6.4 Standard Dav. 4.8 4.2 0 0.8 0.6 0.2 0 0.8 4.0 Minimum 56.7 14.1 0 0.6 0 0 0 0 0 Maximum 74.8 32.1 0 3.7 2.6 0.8 0 2 21.3 a"0ther" refers to plucked grainsI zirconsI or unidentifiable material. 25 features are manifested as jagged edges bounding a pore. Those samples collected fromioutcrop #3 (Figure 2) contain far more limonite than any other outcrop. Quartz Description Quartz comprises 56-75% of the framework grains (Table 2). Intergrown boundaries with quartz cement are common, but dust rims are extremely scarce, suggesting very early quartz cementation. The persistence of well-rounded argillaceous rock fragments suggests a lack of compaction which may be the result of early quartz cementation at a shallow depth. The reader is referred to page 10 for a discussion of shallow quartz cementation. Some quartz grains exhibit a maroon oxy-hydroxide fracture filling which petrographically appears to be inherited from the source region since it does not extend beyond the limits of the host quartz grain” .Also observed.was what appears to be quartz replacement by clay. This replacement manifests itself as small YBIlOW'ClaY flakes being inserted into the outer edge of the quartz grain. In the subsurface it was verified that illite was replacing quartz using EDS on Upper Pennsylvanian sandstones (Westjohn, written communication). XRD data on the overlying Eaton sandstone of this study confirms the presence of illite (Table 5). Isolated vermicular molds are also present, and are sometimes limonite filled in a few detrital quartz grains (Figure 15). This iron-bearing vermicular clay was leither' inherited from 26 7—7 r i, i , i 7 fl #7 '7 7 1‘_,__\u 7— Figure 6. Scanning electron micrograph of weathered quartz showing corrosion of quartz over—growths. Sample JP-95-17. Bold bars are 10 mm apart. 27 the source region or formed diagenetically; it will be seen later in this thesis that the mineral is chlorite. The Role of Ferruginization in Quartz Solubility As stated above, quartz grains are angular in areas where porosity has been partially occluded by limonite, while sub- rounded to rounded.in areas of greatest porosity and limonite is absent. The lack of limonite is likely a result of the higher iron mobility due to the higher rate of the meteoric pore—water flow. Nedkvitne & Bjorlykke (1992) use the same theory, with aluminum however, to explain an absence of kaolinite in facies of high secondary porosity. This may suggest that the limonite is responsible for quartz dissolution. llzis questionable*whether quartz dissolution is associated with the precipitation or dissolution of neoformed limonibe. The observation that limonite fills dissolution cavities and fractures implies that dissolution of quartz occurred prior to the precipitation of limonite. However, it is generally believed that quartz is only soluble in alkaline waters (e.g Krauskopf, 1956; Siever, 1962), and the precipitation of limonite releases protons, which would produce acidic conditions. It is possible, therefore, that the dissolution of limonite (an hydrogen consuming reaction) may provide an alkaline microenvironment suitable for quartz dissolution” ILimonite dissolution. is exhibited. in thin section by the presence of limonite in dissolution cavities, while being absent on the outermost edges of the quartz grain. For alkaline meteoric pore-water to reach the quartz surface, 28 it must infiltrate the limonite surface layer. If this had occurred, then a dissolution void would exist between the quartz and limonite; a feature not observed. This discussion leads the author to Ibelieve that quartz dissolution is associated with limonite precipitation and not dissolution. The reader is referred to page 16 for a discussion of quartz dissolution. Quartz Dissolution and pH Several studies have demonstrated. that quartz grain etching may occur in acidic to near neutral-pH environments, in the presence of high dissolved organic carbon concentrations (i.e. Young, 1987, 1988; Bennett & Siegel, 1987; Bennett et al., 1988; 1991). These studies suggest that organic-acid-silica complexes increase quartz solubility and dissolution rates at near neutral-pH environments. However, even though. lush 'vegetation. covers ‘most horizontal rock surfaces at Grand Ledge, only thin (<1 m) soils cover the sandstones. Furthermore, a pH measurement of spring water draining the ledges in the Spring of 1994 indicates a pH of 8.8. The high pH is attributed to the dissolution of limonite, but that hypothesis has not been confirmed. The silica concentrations in the spring water is 5.6 mg/l (Beals, personal communication) which indicates silicate mineral dissolution (dominated.by~quartz) is occurring at present, but is not exceeding equilibriunlconcentrations, regardless of the high pH (Krauskopf, 1956; Siever, 1962). In addition, at the onset of weathering’ quartz dissolution. was the result of 29 Figure 7. Photomicrograph of a diagenetically altered K- feldspar with etch pits. K—feldspar appears stable in outcrop. Sample JP-95—5. Crossed—polars; field of view is 0.6 mm across. 30 goethite precipitation, but today quartz dissolution is simply the result of dilute meteoric water reaching equilibrium.with respect to silica. K-Feldspar Description K-feldspar composes <1-2% of the framework grains (Table 2), is often fractured, altered along cleavage, and/or exhibiting etch pits (Figure 7). Limonite may or may not be found filling these features. Microcline may be found very pristine. muscovite Description .Although.muscovite accounts for less than.1% of the rock, it is still a common, and important, constituent. It may be found conforming around more competent framework grains, as well as exhibiting no indication of compaction (Figure 8a,c). It may also be observed showing signs of dissolution (Figure 8c) . Tables 2 & 3 show an average muscovite loss due to dissolution of about 0.3%. SEM photomicrographs (Figure 8b) show muscovite undergoing exfoliation as a result of weathering. Energy Dispersive Spectroscopy (EDS) indicates a potassium depletion in the exfoliating layers relative to the unweathered muscovite. Vermiculite was found to be present in the clay-sized fraction of the sandstone (Table 5), and was not identified in the subsurface (Table 6), and is, therefore, a true weathering product. 31 Interpretation The author believes that muscovite is parental to vermiculite based on several lines of evidence (for a complete discussion.on vermiculite, the reader is referred to page 11). Chlorite appears in such small quantities in the subsurface (Table 6; words such as "hint," "trace," and "minor" are used to describe chlorite abundances), that it is unlikely that it could result in a vermiculite peak like that observed on the XRD jpattern. for* the outcrop (Appendix .A). The above discussion (as well as the discussion found on page 11) implies that muscovite is a more likely vermiculite parent than illite. The alteration of muscovite to dioctahedral vermiculite is well established (e.g. Rich, 1958), who states that the widespread occurrence of dioctahedral vermiculite suggests that the formation of this mineral from muscovite is common. Furthermore, both Rich (1958) and Lin and Clemency (1981) describe the removal of K+ from the interlayer sites of muscovite with the destruction of the tetrahedral sheets (Si-O bonds) being the rate-controlling mechanism of muscovite dissolution. The removal of the K? allows the structure to expand to 14 A. As stated earlier, potassium depletion was observed in the exfoliating layers of an Eaton sandstone muscovite. Clay Minerals Clays occur as patches of mixed yellow-orange, and gray vermicular clays (Figure 9), and often as products of in situ weathering' of detrital muscovite (Figure 8). The lack of Figure 8. 32 (a) Photomicrograph of altering muscovite showing compaction features and alteration to vermiculite (arrow). Sample JP—95-1. Plane—polarized light; field of view is 1.5 mm across. 33 Figure 8. (b) Scanning electron micrograph of parental muscovite with exfoliating vermiculite layers. Sample JP—95-1. Bold bars are 100 mm apart. Figure 8. 34 (c) Photomicrograph showing evidence of muscovite dissolution. Sample JP—95-12. Crossed—polars; field of view is 1.5 mm across. 35 compaction, and.presence and persistence of spherical pelitic rock fragments, suggests that any clay present formed post- depositionally, and is not squashed argillaceous rock fragments. Clay loss due to weathering is observed in the Eaton sandstone (Figure 10), and is very important in weakening the sandstone, since the increased porosity may make the rock more sensitive to frost action (Vicente, 1983) or salt wedging. XRD data indicates the presence of abundant, well crystallized kaolinite, and lesser amounts of illite and vermiculite (Table 5; Appendix A). The delicate vermicular crystals of kaolinite suggest a diagenetic origin (Wilson and Pittman, 1977). Reports have been.made of vermicular molds in the limonite-rich. case-hardened. joint faces (Wallis, unpublished data), however, a vermicular kaolinite was found in the joint face using the SEM (Figure 11). The mixed clay patches appear to be preserved where largely enclosed by surrounding framework grains. Clay bounded by relatively large pores typically exhibit dissolution features, and may be coated with limonite. Clay is never found optically continuous with the detrital grains it surrounds, and kaolinite is rarely found associated with microcline. Pyrite Three 0.01 mm cubic euhedral opaque grains were also observed (Figure 12). Based on subsurface petrography (Westjohn, written communication) it is believed that these grains are pyrite. Since the pyrite occurs amidst a patch of brown limonitic matrix, no red. haloes ‘were observable 36 Figure 9. Photomicrograph of a typical diagenetic clay patch. Composition is predominantly kaolinite with lesser illite. Sample JP-95-2. Crossed—polars; field of view is 0.6 mm across. 37 surrounding them. The euhedral shape of these limonite enclosed pyrite grains suggests that they have not experienced any subaerial weathering, likely'duerto the protective coating of the limonite. Ferruginous Oxy-Hydroxides Description The presence of limonite (goethite) is characteristic of the outcrop samples. Although most abundant in the 1 cm thick case-hardened.joint faces (Figure 16), the goethite also forms the 1-5 mm weathering skins found on the outcrop surface. These skins may penetrate as deep as 3-4 cm in the most heavily weathered.outcrop #3 (Figure 2), and.are the result of iron.being mobilized and reprecipitated at the outcrop face in response to evaporation (Williams & Robinson, 1989). Qualitative XRD results show the presence of goethite, and an absence of detectable hematite in the sandstones. The neoformed goethite typically coats all phases in the rock, including dissolution cavity fillings, fracture fillings, dissolution features in detrital and authigenic minerals, as well as penetrating into the clay patches. Wallis & Velbel (1985) noted the presence of vermicular molds in the limonite at the 1 cm case-hardened joint face, however this study found a goethite covered vermicular kaolinite in.tflua joint face (Figure 11). Interpretation Extensive iron staining at springs suggests water draining the outcrop is saturated.‘with respect to the iron 38 Photomicrograph displaying evidence of ' JP—95—1. Figure 10. (a) kaolinite dissolution. Sample Plane—polarized light; field of view is 0.6 mm across . 39 Figure 10. (b) Same view as in (a) but under crossed— polars. 4O Figure 10. (c) Scanning electron micrograph displaying evidence of kaolinite dissolution. Sample JP— 95-21. Bold bars are 100 mm apart. 41 Figure 10. (d) Scanning electron micrograph displaying evidence of kaolinite dissolution. Sample JP— 95—10. Bold bars are 10 mm apart. Figure 11. 42 Scanning electron micrograph of a vermicular kaolinite in the 1 cm case—hardened joint face. Wallis (unpublished data) reported finding "vermicular molds" in goethite in the joint face. Sample JP—95-7. Bold bars are 10 um apart. Figure 12. 43 Photomicrograph of what is believed to be pyrite crystals. The crystals appear to be protected from meteoric pore-water by a thick layer of goethite. Sample JP-95-7. Crossed— polars; field of view is 0.24 mm across. 44 mineral that makes up the stain. This iron is the result of goethite dissolution. Hollingsworth (1977) determined that localized iron oxidation alone was not responsible for changes in outcrop coloration of the Pennsylvanian Kanawha Group in West Virginia. The same author believes that additional iron was introduced via groundwater and deposited as an iron hydroxide. COMPARISON‘WITH SUBSURFACE PETROGRAPHY Subsurface petrography is extracted from photomicrographs (Figures 13, 14, & 15) and other researchers' notes (Westjohn, written communication) , as well as from the author’s own point counts and petrographic observations. 18 qualitative comparison of subsurface/outcrop mineralogy is provided in Table 1. Table 3. Subsurface Pennsylvanian sandstone modal mineralogies. 1. art: Porosit Carbo For + Carbon. R-s er Kaolin. Wu c rite Lithics ‘Other 82-76 61.9 17.0 4.0 21.0 3.8 3.5 0.3 0 8.3 1.3 82-80 61.2 12.0 5.1 17.1 4.0 1.1 1.7 0 10.0 4.3 82-85 66.8 14.3 2.3 16.6 4.3 1.4 1.1 0 8.3 1.4 B2-92 59.9 14.0 8.0 22.0 2.8 2.1 0 0 10.1 2.1 $20-45 69.2 18.7 3.8 22.5 1.3 0.8 0 0 5.7 0.6 82-47 69.7 18.4 0.3 18.7 0.6 3.1 0 0 2.0 0.3 $13-51 67.9 18.7 8.4 27.1 0.6 0.6 0 0 3.8 0.0 S3-41 66.0 18.4 8.6 27.0 1.2 0.3 0.3 0 4.9 0.3 83-118 66.2 15.8 9.9 25.7 1.6 0.6 0.6 0 3.4 1.2 S3-65.5 78.1 10.4 4.7 15.1 1.3 0.7 0 0 2.7 1.0 83-109.5 74.2 17.5 3.1 20.6 0.6 0.6 0.3 0 3.4 0.0 83-134 71.1 19.0 2.3 21.3 1.4 0 0.3 0 5.1 0.9 “can 67.7 16.2 5.0 21.1 2.0 1.2 0.4 0 5.6 Std. Dov. 7.0 5.3 3.0 3.96 1.4 1.1 0.5 0 2.9 Kinimnn 59.9 10.4 0.3 15.1 0.6 0.0 0.0 0 2.0 Maximum 78.1 19.0 9.9 27.1 4.3 3.5 1.7 O 10.1 A"Other" refers to plucked grainsI zirconsI or unidentifiable material. 45 Carbonate Carbonate phases compose the most abundant cement in the subsurface (Figure 13). Point count data from these sandstones shows modal carbonate ranges of 0-10%, but in several samples where carbonate is recorded as occurring in trace abundances it is noted that the carbonate "looks as though it was extensive, mostly dissolved," (Westjohn, written communication). If carbonate and porosity values are combined, values of 15-27% are obtained, averaging 21% (Table 3). Subsurface carbonate dissolution is reported to be producing iron oxide (Table 4), carbonate being a source for the goethite found in outcrop (Figure 14). Numerous researchers have reported a diagenetic siderite source of epidiagenetic iron oxy-hydroxides (e.g. Heald et al., 1979; Arnold, 1978; Hollingsworth, 1977; Arditto, 1983; Young & Young, 1988). Table 4. Report on carbonate abundances in subsurface thin sections. (Westjohn, written communication.) Carbonate Subsurface Report Siderite Forms abundant "concretions." Reported altering to iron oxide, and filling fractures. Ankerite Patchy carbonate cement in "S" and "B" suites is ankerite. Initially "...constitutes a majority of observed carbonates." Dolomite Ankerite zones to dolomite on edges. "Dolomite is rare..." Calcite Samples with all carbonates present only have " . . .rare calcite. . . " which is " . . .almost exclusively poikiloblastic..." Figure 13 . 46 (a) Photomicrograph of subsurface "pre- weathering" carbonates displaying both poikilitic carbonate and rhombs. U.S.G.S. sample 32-92. Crossed-polars; field of view is 2.9 mm across. (Photo courtesy of U.S. Geological Survey, Lansing, Michigan). Figure 13. 47 (b) Photomicrograph of subsurface "pre- weathering" carbonates displaying both poikilitic carbonate and rhombs. U.S.G.S. sample 813—51. Crossed-polars; field of view is 2.9 mm across. (Photo courtesy of U.S. Geological Survey, Lansing, Michigan). 48 Quartz Subsurface quartz grains may be found rounded, as well as euhedral . Several photomicrographs show vermicular kaolinite, illite, and.chlorite "growing out of quartz grains" (Westjohn, written communication) (Figure 15). Outcrop samples may be found containing vermicular molds, sometimes goethite filled. This observation suggests that quartz either contained chlorite in the source region, or diagenetically altered to chlorite, which in turn is being weathered to goethite, or being completely dissolved (Figure 15b). K-Feldspar Feldspars in the subsurface are characterized by being altered by carbonate. Microcline may be found very altered (often with etch pits), sometimes to vermicular clays. In general, feldspar is very similar between outcrop and subsurface (appears stable), which is reflected in the point count data (Tables 2 & 3) (Figure 7). This stability is very intriguing in that muscovite shows evidence of dissolution and alteration (Figure 8), yet has a dissolution rate constant in the laboratory nearly one order of magnitude lower than that for K-feldspar (Busenberg & Clemency, 1976; Lin & Clemency, 1981). Clay Minerals Clays frequently occur as pore fillings of kaolinite and illite in the subsurface, exactly like that found in outcrop, except dissolution of the clay is apparent in outcrop (Figure 10). Absent in outcrop is abundant illite cement with a Figure 14. 49 (a) Photomicrograph showing iron—bearing carbonates altering to iron oxy—hydroxides in the subsurface. U.S.G.S. sample S3—41. Plane-polarized light; field of view is 1.00 mm across. (Photo courtesy of U.S. Geological Survey, Lansing, Michigan). Figure 14. . .. ...> . «z- m'P" -‘ ..~.'fl‘ f 01% ~ . y...‘ ‘ 50 ( . - q ‘ w . r a ‘0 3’} "‘ - . _ .. . s a .. f ¢-» . (b) Photomicrograph showing iron-bearing carbonates altering to iron oxy-hydroxides in the subsurface. U.S.G.S. sample B2-76. Crossed—polars; field of view is 0.5 mm across. (Photo courtesy of U.S. Geological Survey, Lansing, Michigan). Figure 15. 51 (a) Photomicrograph of a subsurface quartz grain altering to chlorite. U. S. G. S. sample J14- 199. Plane- -polarized light; field of view is 0.50 mm across. (Photo courtesy of U.S. Geological Survey, Lansing, Michigan). 52 Figure 15. (b) Photomicrograph of the Eaton sandstone showing result of weathering the quartz-hosted chlorite in (a). Sample JP-95-13. Crossed- polars; field of view is 0.6 mm across. 53 fibrous habit growing perpendicular to grain surfaces. Kaolinite may be found inside pores, completely surrounded by carbonate. This kaolinite nanifests itself :hi outcrop as vermicular books "floating" in the pores spaces. Pyrite Authigenic pyrite may be framboidal or occur as individual crystals in the subsurface. As stated earlier, only three euhedral grains were observed in the outcrop, compared to 2.7% of the total rock in the subsurface (Westjohn, written communication). It is likely that the pyrite is also a contributor of iron for goethite formation. In addition, pyrite oxidation is likely responsible, at least in. part, for' producing the acidic conditions which. are dissolving the carbonate. This idea will be investigated further in the mass balance section of this thesis. The reader is referred to page 12 for a discussion of pyrite weathering. Ferruginous Oxy-Hydroxides Subsurface thin sections generally contain relatively little iron oxy-hydroxides. However, one thin section observation records heavy iron oxide in pores, some nearly spherulites, and related to carbonate cement (Westjohn, written communication) (Figure 14). Hematite is observed in the subsurface, but is scarce, and is not unequivocally identified. Otherwise, all subsurface ferruginous oxy- hydroxides are associated with carbonate dissolution. CHAPTER 3: CLAY MINERALOGY METHODS Five samples were chosen for XRD identification of clays and.oxy-hydroxides. Two of these were taken from.the freshest outcrop (JP-95-21 and JP-95-17), and the remaining three were sampled from each of the other outcrop locations. Four oriented mounts of each sample were prepared, one saturated with potassium, one saturated with magnesium, one saturated with magnesium and glycolated at room temperature, and one with only the naturally—occurring exchange ions. After the initial XRD analyses, the potassium-saturated samples were heated to 575°CL and rescannedu Scans were made on a Rigaku "Geigerflex" x-ray diffractometer system from 2° to 35° 28, at 35 kV, 25 mA, with a scan rate of 1° 28/minute, and divergence, receiving, and scatter slits were 1/6°, 0.3 mm, and 2°, respectively, using Ni-filtered CuKa radiation. Complete clay-mineral preparation techniques may be found in Appendix D. 54 55 RESULTS The results of the XRD analyses of the samples are summarized in'Table 5, and.selected.diffractograms are located in Appendix A. No attempt has been made to quantify the clay phases, but the diffractograms ‘may’ suggest the relative abundance or degree of crystallinity; that is, the sharper and taller the peak, the more abundant and/or the better crystallized the clay. The freshest samples (JP-95-21 and JP-95-17) show the presence of vermiculite, illite, kaolinite, and goethite. No chlorite or carbonate was found in any sample as reported by Martin (1982). Since no vermiculite has been identified in the subsurface (Table 1; Westjohn, written communication; Table 6), it may be concluded that it is a weathering product. Furthermore, since vermiculite is absent from the more intensely weathered samples, increased weathering appears to be removing this phase“ The vermiculite parent is believed to be muscovite, as stated previously. Sample JP—95-10 is the most intensely weathered sample, and petrographically contains the most abundant limonite. Not surprisingly, the XRD analyses for this sample exhibits the tallest and sharpest goethite peak of all the samples analyzed, and the kaolinite and illite peaks are not nearly as intense as the other samples (Appendix A). It may be concluded. that weathering results in. the removal and/or decreases the crystallinity of kaolinite and illite, and 56 increases the abundance and/or degree of crystallinity of goethite. Table 5. Summary of XRD data from the outcrop Pennsylvanian Eaton sandstone. Sample Mineralogy JP-95-2 Kaolinite, illite, goethite, vermiculite JP-95-10 Kaolinite, illite, goethite JP-95-13 Kaolinite, illite, goethite JP-95-17 Kaolinite, illite, goethite, vermiculite JP-95:21g Kaolinite, illite1,qoethite. vermiculite Table 6. Summary of U.S.G.S. subsurface Pennsylvanian sandstone XRD data. (From Westjohn, written communication). Sample Mineralogy 72 Kaolinite >>> illite 80 Kaolinite >> illite, trace chlorite 85 Kaolinite >> illite, trace chlorite 93 Kaolinite >> illite, trace chlorite 106 Kaolinite >>> illite 110 Kaolinite >>> illite, trace chlorite 121 Kaolinite >>> illite, hint chlorite 125 Kaolinite >>> illite, hint chlorite 127 Kaolinite > illite, trace chlorite 133 Kaolinite >> illite, trace chlorite 134 Kaolinite >>> illite, hint chlorite PSS-l Kaolinite >> illite, hint chlorite PSS-3 Kaolinite >> illite, trace chlorite PSS-2 Illite > kaolinite, minor chlorite JS-l Kaolinite > illite, minor chlorite JS-4 Kaolinite = illite, minor chlorite JS-5 Kaolinite = illite, minor chlorite JS-6 Illite = kaolinite, minor chlorite JS-8 Kaolinite > illite, minor chlorite JS-9 Kaolinite > illite, minor chlorite JS-10 Kaolinite > illite, minor chlorite JS-ll Kaolinite > illite, minor chlorite JS-13 Kaolinite > illite, minor chlorite JS-14-1 Kaolinite > illite, trace chlorite USGS 110 Kaolinite, minor illite, trace chlorite USGS 106 Kaolinite, minor illite, trace chlorite USGS 52 Kaolinite. trace illite CHAPTER 4: MASS BALANCE MODELINQ METHOD OF CALCULATION The mass balance calculations performed in this study follows the method described by Merino (1975a, 1975b), and utilized by Land & Milliken (1981). It requires that molar volumes (v—) of all mineral phases used in the balance be known. The units of molar volume are cmP/mol, the inverse of which (mol/cwfi) is desired in order to generate the number of moles released per given volume of rock (in this investigation the reference volume of rock for mass balance calculations will be cubic meters). The inverse molar volume, therefore, may be calculated using the following expression: 1/9 = G * (1/mw) * 106cm3/m3 the inverse molecular volume in mol/m’, where l/v G the specific gravity of the mineral in g/cm3 (all specific gravity values are taken from Klein & Hurlbut, 1985), mw = the molecular weight of the mineral in g/mol. The inverse molecular volumes used in the calculations of this communication may be found in Table 7. The volume of a given mineral lost or gained during 57 58 weathering was determined by point counting, with the subsurface thin sections representing the pre—weathering mineral abundances. Approximately 80 Upper Pennsylvanian subsurface thin sections were examined in order to select those which texturally and compositionally match the outcrop. Only 12 of these thin sections were qualitatively and quantitatively similar enough to the outcrop to justify their use as pre-weathering representatives of the outcrop. The outcrop and subsurface sandstone detrital abundances are plotted in Figure 3. It should be noted that the subsurface samples are more lithic rich, departing from the quartz arenite classification and into the sublitharenite classification. The lithic fragments in the subsurface thin sections are typically schistose fragments; small polycrystalline and monocrystalline quartz grains bound by platy muscovite. In outcrop, these lithic fragments are manifested as clustered silt size quartz grains, separated by dissolution voids where muscovite once was, but has since been weathered away. It is possible that the outcrop rocks are simply a different lithofacies than the selected subsurface rocks (the most lithic rich subsurface rocks are from the "B" suite of Westjohn (written communication) , while the remainder are from their "S" suite). However, the persistence of clustered silt size quartz grains observed in the subaerial samples suggests that the outcrop rocks may have been more lithic-rich prior to weathering. With the quartz and feldspar modal ranges being extremely similar between outcrop and 59 subsurface samples (Tables 2 & 3), the possibility that differences in primary depositional facies are responsible for the compositional variation observed in Figure 3 is very small, although not negligible. The author point-counted both the subsurface and the outcrop, the data and statistics of which may be found in Tables 2 & 3. At least 300 points were counted for each slide, with the final totals almost always equalling 400 points or more. The net loss or gain of each mineral may be found in Table 7. The only mineral that was not actually point counted by the author is pyrite. All of the 12 subsurface thin sections are within 135 feet of the Earth's surface and are beginning to alter. All of the slides exhibit high porosity and some carbonate dissolution” Since pyrite is one of the most highly soluble minerals it has already been removed. However, Westjohn (written communication) reports a thin section containing 2.7% pyrite and.6% porosity; If it is assumed that the sandstone was completely occluded with cement prior to uplift, then.the 6% porosity combined with the ease with.which pyrite weathers suggests that even 2.7% is a minimum value. Furthermore, it is likely that pyrite weathering is responsible for producing the acid.which, at least in.part, is dissolving the carbonate. 'Fherefore, the absence of pyrite in the 12 subsurface thin sections used in this study does not imply that no pyrite was precipitated, but, rather, that pyrite dissolution has already occurred. With the molar volumes of each mineral calculated and the 60 percent change (created or destroyed) determined from point count data, the number of moles of a given mineral formed or removed from a cubic meter of rock may be determined. Before doing the elemental mass balance, however, accurate mineral compositions must be known, and balanced chemical reactions must be identified. MINERAL COMPOSITIONS Table 7 lists the mineral compositions used in the material balance calculations, and the sources of the formulas. Formulas are either idealized textbook formulas or actual formulas determined by quantitative energy dispersive spectroscopy (Westjohn, written communication). Pyrite Elemental stoichiometries of Fe2+ and S‘ were determined by EDS to be FeSz, the idealized textbook formula. Carbonates Westjohn (written communication) performed. extensive analyses of the carbonate phases. Quantitative EDS data was conducted on carbonate patches, nearly all rendering very good major element oxide totals close 100%. Those analyses which did not sum to Mathin 11.5% of 100% were not used in this study. The four carbonates found to be present in significant quantities are calcite (typically poikilotopic), ankerite, siderite, and dolomite. The exact modal distribution of each 61 has not been determined. However, when performing mass balance calculations the modal percent of each carbonate is needed. Based on the report by Westjohn (written communication), as well as petrographic observations made by the author, a geologically reasonable carbonate distribution may be established. Admittedly, this distribution makes a number of inferences, but nonetheless, the distribution is geologically reasonable, and is constrained by all available information. The constraints are summarized in Table 4. The information in Table 4 implies the following relationship between carbonates: Ankerite > Siderite > Calcite > Dolomite The siderite composition is actually an average since the siderite concretions are reported to be zoned from a Mg—poor core to a Mg-rich edge. The average is between the core and the Edge (Feo . ascao . onno . 03C03 and Femeccao. oaMgo. 14MI10.03CO3 I respectively). The total carbonate percentage is the subsurface modal carbonate + modal porosity, which assumes that all porosity in the subsurface samples resulted from the removal of carbonate cement (Table 3). Kaolinite The reported oxide abundances for kaolinite have extremely poor totals. The error is likely attributed to the inherent difficulty associated with analyzing any mineral composed of clay-size crystals, combined with the fact that EDS cannot determine the structural water content of hydrous 62 minerals. Since there is essentially no compositional variation in kaolinite, using an idealized formula is justified. Goethite, Muscovite, and Quartz No data is available on the true compositions of any of these phases. However, none of these minerals exhibits large stoichiometric variability in nature, and, therefore, traditional chemical formulas will be used for the material balance calculations. Table 7. Mineral data for phases used in the mass balance model. Net loss or gain (%): 1/% Joint block Joint Phase (molelm’) Chemical formula interior tag; Pyrite 4.17*104 FeSi“° -2.7 -2.7 Calcite 2.70*104 Cao.97Feo.o3CO3° -4.4 -4.4 Ankerite 1.42*1o4 CaLlMgOMFeO.“ (C03) 2" -7 . 3 -7 . 3 Siderite 3 . 50*10“ Feo.87Cao.o3Mgo.o7Mno.03CO3° -6 . l —6 . 1 Dolomite 1.5331104 Cao.98Mgo_95Feo.06 (C03) 2" -3 .3 -3 .3 Goethite 4.92*104 a-FeO(OH)’ +6.4 +19.7 Muscovite 7. 08*103‘” KA12(AlSi3010) (OH) 2’ -o .3 -o .3 Kaolinite 1 . 01*104 15.1231205 (OH) ,’ -o . 5 -o .5 Quartz 4.41*10‘ SiOJ -1.8 -1.8 C} The specific value of muscovite varies over a small range. In this case the geometric mean was used for the inverse molecular volume calculation. T Standard formula from Klein & Hurlbut, 1985. ¢ Microprobe data from Westjohn, written communication. 63 BALANCED CHEMICAL REACTIONS As stated previously, balanced chemical reactions are needed to perform mass balance modeling. All of the chemical reactions used in this study are listed in Table 8. Table 8. Chemical reactions used for mass balance modeling. Reaction Title Balanced reaction Pyrite dis. Fes2 + 3.5 02 + H20 « Fe2+ + 2 H‘ + 2 so,” Calcite dis. H‘ + 0.88 Caomiveomco3 .. 0.86 Ca“ + 0.03 Fe2+ + 0.12 112cc3 + 0.76 HCO,‘ Ankerite dis. H‘ + 0.44 Ca1.1Mgo_“Feo.“(CO3)2 .. 0.49 Ca2+ + 0.20 Fe" + 0.19 Mg" + 0.12 H,C03 + 0.764 Hco; Siderite dis. H’ + 0.88 Fee.8.0-1,,,nlvxgomlvmomco3 .. 0.77 Fe“ + 0.03 Ca" + 0.06 Mg2+ + 0.03 Mn” + 0.12 H2C03 + 0.76 HCO,’ Dolomite dis. H‘ + 0.44 Ca,“Mg,flap-em;(00,)2 .. 0.43 Ca” + 0.03 Fe" + 0.42 Mg" 4» 0.12 Hzco, + 0.76 sec; Goethite reci . Fe2+ + 0.25 O -+21JSI{O e aFeO(OH) + 2 H+ P P 2 2 Muscovite dis. KA12 (A1813010) (0H)2 + 12 H20 «- K‘ + 3 A1(0H),' + 3 14,810, + 2 H+ Kaolinite dis. A128i205(OH). + 7 11,0 .. 2 Anon),- + 2 H,Sio, + 2 11* Quart; dis. SiO. + 2 H.O e H.Si0. Carbonate Equilibria Since carbonate dissolution is both temperature and pH dependent, it deserves special attention (Drever, 1988). In finding the appropriate pH for carbonate dissolution due to pyrite weathering, subsurface water chemistry was found which had high dissolved Fe“, high dissolved sulfate, and high alkalinity (Dannemiller & Baltusis, 1990). In doing so, the 64 idea is to find those wells where pyrite oxidation appears to be occurring, and average those pH and temperature values in order to determine the CO3 and bicarbonate (HCO3’) concentration in the weathering waters. The average values finally used were based on 12 wells, which average a pH of 7.25, and a temperature of 12.42° C. Using the following general reaction: K1 2 H2C03 .. 2 H‘ + 2 HCO3' which may be written as follows: 4 H+ + 2 CaCO3 a 2 H‘ + 2 HCO3' + 2 Ca“. The dissociation.reaction of bicarbonate is not important here since the pH is 7.25. Since, the equilibrium constant K1 varies linearly with temperature at temperatures below approximately 25° C, then K1 may be extrapolated from Drever’s (1988) data. At 12.42° C, the equilibrium constant K1 equals 3.63*10”. If K1 [H‘] [HCO3‘]/[H2CO3], then Kl/[H*] = [HCO3']/[H2C03] 3.63*10”/5.62*104 6.46/1 Therefore, at pH = 7.25 and T = 12.42°CL inorganic carbon is 65 distributed between the following species: 0.87 Hco; 0.13 H2C03 which provides: 2 H2CO3 .. 0.27 H2C03 + 1.73 H‘ + 1.73 HCO3' Calcite dissolution involves a net 2.27 H7, and the overall reaction becomes: H‘ + 0.88 Caco3 a 0.12 142003 + 0.88 Ca2+ + 0.76 HCO3'. It must be kept in mind that the above solution is the general case for the dissolution of any carbonate. Silica Equilibria Silica solubility, too, involves the production of the weak monosilicic acid (H4SiO,) . However, in this case, H,,SiO4 is not particularly problematic because measured pH values of spring water draining the Eaton sandstone show a pH of 8.8, and at pH below about 9 only H4SiO, contributes significantly to the 281 (Richardson & McSween, 1989). Therefore, the simple congruent reaction of silica and water in Table 8 is sufficient here. Aluminosilicate Solubility The aluminosilicate minerals involved in the mass balance calculations are muscovite and kaolinite. What is important about these two phases is that the aqueous aluminum species resulting from their dissolution is pH dependent (Drever, 66 1988). At an outcrop pH of 8.8, Al(OH); will be the dominant aqueous aluminum.species (Drever, 1988), which.is reflected in the chemical reactions in Table 8. ICE! EHUSS EHELAmflIE With balanced precipitation and dissolution reactions identified from thin sections (Table 8), and modes of new and destroyed minerals determined (Table 7), quantities of each element imported and exported for a given volume of rock may be established (Table 9). Table 9. Mass balance calculations for joint face and joint block interior. (All abundances in moles/u? of sandstone.) A. JOINT FACE % of Moles needed Mblee released Ph... rock g4 32+ §+ p.24 814+ c220 EEO 92+ L130 x40 so 3- Pyrite -2.7 2300 1100 2300 Calcite -4.4 1300 35 1200 Ankerite -7.3 2300 470 1100 450 Siderite -6.1 2400 1900 63 150 63 Dolomite -3.3 1100 29 490 480 Goethite +19.7 10300 20700 Muscovite -0.3 42 64 64 21 Kaolinite -0.5 100 100 100 Quartz —1.8 790 Total; 7100 10300 23142 3534 954 2853 1080 63 164 21 2300 B. JOINT BLOCK INTERIOR % of Moles needed Moles releeeed Phase rock I? Fej‘ 13‘ Fe” 81“ ct“ m 3" 41” r so," Pyrite -2.7 2300 1100 2300 Calcite -4.4 1300 35 1200 Ankerite -7.3 2300 470 1100 450 Siderite -6.1 2400 1900 63 150 63 Dolomite -3.3 1100 29 490 480 Goethite +6.4 3100 6300 Muscovite -0.3 42 64 64 21 Kaolinite -0.5 100 100 100 Quartz -1.8 790 Total; 7100 3100 8742 3534 954 2853 1080 63 164 21 2300 67 Model Mineral Changes The point count data in Tables 2 & 3 includes maximum, minimum, and average values. In performing material balance only two numbers for a given mineral are needed; a pre— weathering percentage and a present day percentage. In this study, the author has elected to use average values from the point count data for several reasons. First, choosing the minimum values often provides 0% of a phase which is known to be present. For example, the minimum goethite value is zero, which is certainly not geologically reasonable for the Eaton sandstone. If one chooses the maximum values, then suddenly almost 2/3 of the original pore space is now occluded by goethite, which, again, is neither consistent with petrographic observations, nor is geologically reasonable. Therefore, this study utilizes mean modal mineralogies for the material balance calculations, which are the most reasonable based on petrographic and field observations. The "Net Loss or Gain" column in Table 5 is based on the mean values from Tables 2 & 3. Precipitation Inputs When performing any mass balance calculation.both inputs and outputs must be identified. Since precipitation provides an input into this systeMIOf study, its chemistry becomes very important. Bulk precipitation chemical analyses were performed in Lansing, Michigan (11 miles east of Grand Ledge) (Peters & Bonelli, 1982), and.East Lansing (Wood, 1969) (Table 10). Compared to the contributions of ions by the minerals, 68 the precipitation inputs are negligible. Admittedly hydrogen ion.is present.in.the precipitation in significant quantities, but this study does not attempt to mass balance protons. Sulfate is significantly concentrated in the precipitation of south central Michigan. In fact, the groundwater chemistry for one of the wells near Grand Ledge (Table 13) has a sulfate concentration exactly the same as the Lansing precipitation measurement (2.3 mg/l). At least for that particular well, precipitation is tflua only source of sulfate at the present time. Table 10. Precipitation data from Lansing and East Lansing, Michigan. (All concentrations in mg/l.) Specific Conductance pH Ca“ Mg” Na’ K’ 80." Cl' Si" Fe” Mn’+ 1§igmfi 5.34 0.39 <0.1 <0.2 <0.1 2.3 0.61 <0.01 0.02 0.004 ‘ <50 5.8 10 0 -- -- 3.5 0.8 -- -- -- ‘ <50 5.3 10 0 -- -- 2.0 0.9 -- -- -- O Peters & Bonelli, 1982 ‘ Wood. 1969 An attempt was made to quantify the ionic inputs from precipitation through time into the Eaton sandstone, but such calculations are very complicated and elusive. In order to perform such calculations one needs to know the time of exposure of the outcrop, the change in porosity as a function of time, the changes in precipitation volume and chemistry as a function of time, the actual volume of precipitation to enter the rock, and whether evapotranspiration effects are causing increase in elemental concentrations. Obtaining 69 accurate numbers on the above is extremely difficult, and is beyond the scope of this thesis. However, it does suffice to conclude that the contributions of elements from precipitation is extremely small compared to the contributions from the weathering minerals. Wood (1969) came to the same conclusion for the Grand Ledge area. Furthermore, Wood (1969) states that chloride is the only major ion for which precipitation may be a significant source. Since chloride is not one of the elements involved in the material balance calculation, quantifying precipitation inputs is not necessary. Joint Face vs. Joint Block Interior The mass balance calculations were performed.for both the joint face (1 cm thick), and the joint block interior (Table 9). It is apparent in the field that the 1 cm joint faces are much better indurated than the much more friable joint block interiors. Figure 16 shows a plot of modal percent goethite vs. distance from joint face, and what is readily apparent is the abundances of goethite in the 1 cm thick joint face. Furthermore, Table 9 shows that the joint faces behave as net importers of iron, while the joint block interior behaves as a net exporter. For the diffusion of iron to occur to the joint face the sandstone must have been completely, or nearly completely saturated with water. This suggests that the formation of goethite occurred in the shallow subsurface, likely in the saturated zone. In addition, jointing must have occurred prior to pyrite oxidation and carbonate dissolution, since 7O -.— Outcrop 2 _+_. Outcrop 3 —e— Outcrop 4 —B— Outcrop 6 Percent goethite I l 4 6 8 10 12 14 Distance from joint (inches) a 0 <__ Joint face \5 2 Figure 16. Percent goethite vs. distance‘ from joint face. 71 these phases are the iron source. The idea that perhaps these outcrop sandstones may not have been buried sufficiently deep to achieve the subsurface diagenetitrmineralogy should also be dismissed since significant burial, and resulting pressure, must have occurred to permit fracturing during uplift. This uplift and epidiagenetic sequence is supported petrographically. Subsurface thin sections exhibit initial pyrite oxidation, followed by carbonate dissolution. Table 9, however, shows that insufficient H‘ is released by pyrite weathering to dissolve all of the carbonate. Carbonic acid resulting from CO2 dissolution in rainwater is the likely other acid responsible for carbonate dissolution. It is reasonable to conclude that groundwater flow along subsurface joint.p1anes provided.the oxidant responsible for the goethite formation at the joint faces (groundwater ferricretes). The iron mass balance suggests that the Eaton sandstone is behaving as a closed system with respect to iron. The joint faces require approximately 6766 moles of Fe+2 per m3cflf sandstone more than they themselves can supply, while the joint block interiors release approximately 434 moles of Fe+2 per HP of sandstone than their secondary minerals need. These numbers indicate that each unit volume of case-hardened joint face material requires the importation of Fe+2 equivalent to that derived from over 15.5 unit volumes of joint block interior material. 72 COMPARISON'WITH SUBSURFACE WATER CHEMISTRY AND MINERALOGY Mass balance tabulations represent net import and export of ions through time. Any water chemistry data collected at the present represents the concentration of elements in solution at an instant in time. Therefore, a comparison of water chemistry over time (the mass balance) and instantaneous water should provide insight into the epidiagenetic sequence. Mineral Stability Table 11 provides a comparison of mineral stability for outcrop, glacial till aquifers (Wahrer, 1993), and Pennsylvanian aquifers (Meissner, 1993) in the Michigan basin. The stability of several of these phases deserves special attention. Quartz Outcrop» petrography’ reveals quartz grains ‘which. are embayed, and.particularly angular when bounding pores (Figure 6). An analysis of spring water draining the outcrop reveals SiO2 concentrations of 5.62 mg/l (Beals, personal communication). Such a concentration reflects quartz saturation, and. is typical for IEarth’s surface conditions (Krauskopf, 1956; Siever, 1962). As (discussed. earlier, initial quartz dissolution occurred in response to goethite precipitation. Today, however, it is simply the rain water attempting to reach.saturation.with.respect to quartz, and.may or may not be a result of the percolating waters having a pH of 8.8. 73 Table 11. Mineral stability comparison for outcrop, glacial till, and Pennsylvanian aquifers. *Glacial Till **Pennsylvanian Mineral Outcrop Aggifer Aggifer Calcite Unstable Stable Stable Dolomite unstable Slightly Unstable Slightly Unstable Gypsum Unstable Unstable Unstable Anhydrite Unstable Unstable ? Quartz Unstable Stable Stable Illite Unstable Stable Unstable Kaolinite Unstable Stable Stable Muscovite Unstable Unstable Stable K-Feldspar Stable (?) Unstable Unstable * Wahrer, 1993 ** Meisner1,1993 K-Feldspar K-feldspar petrographically appears stable. Qualitatively the subsurface feldspar is as altered.as that of the outcrop. Point count data reveals no loss of feldspar due to weathering (Tables 2 & 3). However, glacial till aquifer waters (Wahrer, 1993), and Pennsylvanian aquifer waters (Meissner, 1993) are undersaturated with respect to K- feldspar. Numerous physical and chemical conditions may explain the outcrop stability (e.g. water contact times, high pH, volume of water contacting the sandstone, etc.), but regardless of these conditions both kaolinite and muscovite show evidence of leaching. 74 Sulfate Both Meissner (1993) and Wahrer (1993) believe that the sulfate in Pennsylvanian and glacial till aquifers, respectively, is from gypsum (CaSO,-2HZO) and anhydrite (CaSO,) weathering in overlying Jurassic red beds (Cohee, 1965). Any sulfate found in the water of the Upper Pennsylvanian aquifers of south central Michigan is not believed to be from red beds, but rather from pyrite oxidation. The maximum value of sulfate found in the waters of either the Saginaw or Grand River formations near Grand Ledge is 160.00 mg/l (Table 13), which may be entirely accounted for by oxidizing 2.7% pyrite. Therefore, when comparing outcrop mass balance calculations to shallow water chemistry, it may be assumed that the only elemental inputs into the aquifer are from precipitation and mineral-water interactions occurring within the Pennsylvanian recharge beds. Long Term.vs. Instantaneous water Chemistry .Method The mass balance calculations for the outcrop sandstone represent elemental imports and exports since the onset of weathering, and, therefore, may provide a way to estimate an average long term water chemistry. Since SiC5 concentrations for a spring draining the outcrop are known, a normalization factor may be calculated and used to convert all of the other major ions into concentrations in mg/l. The method begins by taking the known, contemporary $105 concentration of spring water and converting this concentration from mg/l to mol/l. 75 To do so, simply multiply the known SiO2 concentration (5.62 mg/l; Beals, personal communication) by the molecular weight of quartz in milligrams (1 mol/60080 mg): 5.62 mg/l * 1 mOl/60080 = 9.35””10'5 mOl/l. Now that the concentration of SiO2 is in mol/l, it may be divided by the number of moles of SiO2 per m3 of sandstone from the mass balance (Table 9) in order to clear the moles and leave units of m3 of sandstone per liter (m3 ss/l): 9.35410-5 mol/l + 954 mol/m3 88 = 9.81*10'° m3 ss/l. 9.81*10*’ufi ss/l is the conversion factor used to convert the mass balance values in moles per HP of sandstone (Table 9) into concentrations in mg/l. The conversion factor is the inverse of the volume of water (with the measured dissolved SiO2 concentration) required to flush a cubic meter of sandstone to remove the amount of silica calculated by the mass balance (Table 9). If it is assumed that all solid phases in the sandstone have interacted with this same volume of water, then by multiplying the quantity of each element/specie in Table 9 (in.amd/m? ss) by the conversion factor (in m3 ss/l) a value in mol/l may be obtained. A concentration in mol/l is easily converted into mg/l by using the molecular weight of the particular element/specie and converting from grams to milligrams. For example, Ca2+ is as 76 follows: 2853 mol/nfi ss * 9.81*10’8 09 88/1 = 2.8*10'4 mol/l (From Table 9) (conversion factor) 2.8*10'4 mol/l * 40.08 g/mol * 1000 mg/g = 11.2 mg/l (molecular weight) The last value (11.2 mg/l) is placed in the "Outcrop (average over time)" column of Table 14. Iron is neglected from these calculations because of the difficulty involved in calculating the volume of the joint face, the volume of the joint block interior, the volume of the weathered skin, and the amount leached away, all of which are needed in order to sum the total number of moles of iron. The "Subsurface (Present)" column of Table 14 is the major element water chemistry data from Upper Pennsylvanian aquifers near Grand Ledge, Michigan. Ratios are used for comparisons since the groundwater has undergone a longer flow path and has accumulated more solutes. When possible, data from the Grand.River formation.is used, however, data does not exist for all elements, in.which case, Saginaw aquifer data is used. When comparing the Grand River data and Saginaw data (Tables 12 & 13), significant differences exist. For example, Grand River aquifer wells are shallower and more dilute. However, concentrations do not vary by orders of magnitude and, as will soon be seen, return very useful results. 77 Table 12. Water chemistry data from the Grand River aquifer near Grand Ledge, Michigan. (All concentrations 1n mg/l.) —m ufi'fl'. Reference Depth Conductance pH Ca" Mg” Ne’ r’ (c.co,) 80.” c1' ms Fe" aco, Identifier (feet) (Helen) CLINSS 137 -- -- -- -- -- -- 102 12.0 3.0 -- 0.36 230 °CLIN56 120 -- -- -- -- -- -- 342 13.0 1.0 -- 2.40 410 °KENT20 107 -- -- -- -- -- -- 276 40.4 3.0 -- 1.15 312 ‘CLINGRl 180 -- 7.5 60 26.0 14.0 1.9 330 11.0 1.2 288 0.73 351 ‘CLINGR2 120 625 7.6 -- -- -- -- -- 13.0 1.0 -- 2.40 410 ‘CLINGR3 200 -- 7.3 70 32.0 13.0 2.4 306 11.0 1.2 325 0.25 397 ‘CLINGR4 200 -- 7.5 77 36.0 13.0 2.2 340 9.0 1.0 377 1.40 480 ‘CLINGRS 255 -- 7.3 56 26.0 16.0 2.0 288 7.0 1.0 288 2.40 361 ‘CLINGRG 180 -- 7.4 82 34.0 16.0 1.8 345 19.0 2.1 363 3.50 471 ‘CLINGR7 230 -- 7.6 62 18.0 30.0 2.6 238 9.0 1.5 280 0.20 316 ‘CLINGRB 250 -- 7.6 79 24.0 6.5 1.8 296 7.0 1.0 313 1.80 390 ‘CLINGR9 13s -- 7.8 84 31.0 13.0 3.4 337 36.0 3.5 374 1.20 407 ‘CLINGRlO 230 -- 7.5 82 29.0 15.0 1.4 324 23.0 6.3 358 2.80 402 ‘CLINGRll 137 381 7.9 -- -- -- -- 102 12.0 3.0 -- 3.60 230 ‘CLINGR12 200 653 7.3 79 32.0 14.0 1.4 329 24.0 10.0 359 1.00 398 ‘CLINGR13 71 487 7.3 67 22.0 8.6 1.1 258 11.0 0.0 285 0.38 325 ‘CLINGR14 209 344 7.7 35 8.8 28.0 3.1 124 5.2 0.0 208 0.34 220 ‘EATOGRl 111 -- 7.8 75 30.0 8.0 2.8 325 24.0 4.4 323 2.40 354 Mean 171 498 7.5 70 27.0 15.0 2.2 274 15.9 2.5 319 1.57 359 Stand. Dev. 55 139 0.2 14 7.4 6.9 0.7 84 10.0 2.5 49 1.13 76 Minimum 71 344 7.3 35 8.8 6.5 1.1 102 5.2 0.0 208 0.20 220 lexgggg 255 653 7.9 84 36.0 30.0 3.4 345 40.4 10.0 377 3.60 480 s tate of Michigan, 1966 ‘ Wood, 1969 CLIN a Clinton County EATO . Eaton County KENT = Kent County 510,, Ca”, Mg”, .50," One way to check the data is to calculate the ratio of the elemental concentrations against SiO2 for both outcrop and subsurface. This is the column entitled "Species/$102" in Table 14. Notice how S102, Ca”, Mg“, and SO,” all show very similar numbers for outcrop and subsurface (within the same order of magnitude). This similarity suggests that the minerals which contain those species (quartz, a.r1]< e r 1.1: e / s i.tn i t e , ankerite/siderite/dolomite/calcite, and pyrite, respectively) are weathering at present in the shallow Upper Pennsylvanian aquifers. Notice, too, for these same species the similarity between the ratio of the long term average to the pmesent (Table 14, column 6), as well as them. all having a negative 78 Table 13. Water chemistry data from the Saginaw aquifer near Grand Ledge, Michigan. (Data from Dannemiller & Baltusis, 1990; All concentrations in mg/l.) —mfic um. Reference Depth Conduct. pa Ca" Mg" Na’ R‘ (CaCO,) 80," Cl‘ Si" TDS Al" Fe" Mn" Identifier (Feet) (Ellen) CLINl 200 660 7.1 85 33 6.8 1.8 520 2.3 1.1 13 360 <0.01 0.66 0.02 CLINZ 305 557 7.4 69 32 13.0 2.2 316 8.4 1.2 18 315 <0.01 0.53 0.01 CLIN3 155 505 7.5 62 24 20.0 1.6 283 4.8 1.3 17 281 <0.01 0.23 0.19 CLIN4 462 556 7.2 71 26 11.0 2.6 297 17.0 2.6 9 324 <0.01 0.43 0.02 CLIN6 355 642 7.6 78 32 12.0 1.8 362 10.0 1.5 15 344 <0.01 0.56 0.02 EATOl4 360 634 7.3 78 27 20.0 2.3 320 32.0 9.6 13 372 <0.01 0.44 0.01 EAT016 440 821 7.1 110 35 17.0 1.2 352 63.0 33.0 17 511 <0.01 1.10 0.04 EAT017 441 870 7.1 110 38 16.0 1.3 398 80.0 29.0 18 500 <0.01 1.10 0.05 INGH20 160 589 7.5 77 26 3.7 1.6 312 5.9 1.7 13 306 <0.01 0.71 0.02 INGHZG 490 652 7.3 86 26 13.0 2.2 300 26.0 20.0 12 356 <0.01 0.51 0.02 IONI4 245 601 7.5 73 29 9.4 1.0 314 11.0 1.1 18 320 <0.01 0.90 0.03 IONIB 485 551 7.4 76 29 16.0 1.9 333 5.9 9.3 17 338 0.01 0.78 0.03 IONI9 450 774 7.5 120 37 40.0 3.4 367 160.0 21.0 12 590 <0.01 3.20 0.07 Mean 350 647 7.3 84 30 15.2 1.9 344 32.8 10.2 15 378 <0.01 0.86 0.04 Std. DOV. 125 111 0.2 18 S 8.9 0.6 62 45.1 11.6 3 94 0.00 0.75 0.05 Minimum 155 505 7.1 62 24 3.7 1.0 283 2.3 1.1 9 281 <0.01 0.23 0.01 “.818“! 490 870 7.6 120 38 40.0 3.4 520 160.0 33.0 18 590 0.01 3.20 0.19 CLIN a Clinton County EATO - Eaton County INGH a Ingham County IONI = Ionia County The number following the four letter county abbreviation refers to the well number in Dannemiller & Baltusis (1990). Table 14. Long term vs. present day water chemistry in Pennsylvanian sandstones. (All concentrations in mg/l.) mEcrop Sfieurface SPOCIOI7 SIC, 8102 5.6 15‘ 1 1 0.38 -9.2 Ca” 11.2 70 2.0 4.7 0 16 -58 8 Mg” 2.58 27 0.46 1.8 0.10 -24.4 Mn” 0.34 0.04: 0.06 0.003 8.5 +0.3 Al” 0.43 <0.01: 0.08 0.01 <43 +0.42 K‘ 0.08 2.2 0.01 0.15 0.04 -2.12 803' 1.55 15 9 0.28 1.1 0.10 -14.35 ‘Data taken from the Saginaw aggifer (Table 13) rather than the Grand River aggifer. difference when subtracting the present from the long term average (Table 14, column 7). It may seem.strange that quartz dissolution appears to be occurring in the Pennsylvanian aquifer when Meissner (1993) reported that quartz is stable in these aquifers. It must be remembered that the water chemistry data acquired for this comparison are from the shallowest, most dilute aquifers where weathering is likely occurring. Meissner’s (1993) mineral stability data is for the entire Michigan basin which includes 79 a large number of relatively deep, solute-rich wells. In summary, the wells utilized in this study are sufficiently shallow to reside above the weathering front, permitting minerals like quartz and carbonate to experience weathering, while Meissner (1993) incorporates numerous, very deep, solute-rich wells into his study, which is a generalization of the entire Michigan basin. K’ Potassium, too, has a negative value in column 7 of Table 14, but columns 4 and 5 (ratio of chemical species to SiOfl differ by 15 times. The fact that the [K*]/[Si02] ratio is smaller in outcrop than in the subsurface suggests that less dissolution of K*—bearing minerals is occurring in outcrop. The possible K*-bearing phases are K-feldspar and muscovite. According to Meissner (1993) muscovite is stable in the subsurface, while K—feldspar is unstable. Muscovite occurs in very small abundances in both outcrop and subsurface (mean of 0.1% and 0.4%, respectively) and is therefore, not a major contributor of potassium to either outcrop or aquifer. K- feldspar, in contrast, has a mean modal percent of 2% in the subsurface and outcrop (Tables 2 & 3) . This relatively large K-feldspar abundance coupled with instability provides a substantial source for the potassium ion in the groundwater. Though in similar abundance in outcrop, the K-feldspar's apparent stability at the Earth's surface prevents significant K“ from being leached from the weathering environment. 80 Mn2+ and 1113* Mn” and AP” also deserve explanation. For these two cations, column 7 in Table 14 is positive, indicating that they are being mobilized in outcrop more so than in the aquifers. It also implies that the phases in the subsurface which.contain.manganese and.aluminunlhave either not commenced weathering (are stable), or have been completely removed by meteoric water. The only source of Mn2+ is siderite. Westjohn (written communication) reports that for the most porous subsurface sandstones, the remnant carbonate is ankerite and/or calcite. The petrographic observations of this study confirm that statement. This implies that siderite is one of the earliest phases to dissolve. If so, then the very low subsurface Mn2+ concentrations likely reflect the complete absence of siderite in the present day, or, at least, a lack of Mn mobility. Since the water chemistry data used for this study is from shallow wells, the above discussion is altogether geologically reasonable. Aluminum is particularly interesting. The primary sources of aluminum are muscovite and kaolinite. From Table 11 it is observed that both. phases are stable in the subsurface, thus explaining the minimal.Alh’concentrations in the subsurface. In other words, subsurface Al3*-bearing phases have not yet begun to weather. However, both muscovite and kaolinite are unstable in outcrop and are undergoing dissolution” 'What is of particular importance to note here is 81 that aluminum (and possibly manganese) is mobilizing in the outcrop sandstone. CHAPTER 5: UNCONPORMITIES AND HYDROCARBON RESERVOIRS The Eaton sandstone represents porosity enhancement as a result of acidic meteoric water leaching below a modern-day unconformity surface. Such porous weathered zones as this may provide potential hydrocarbon reservoirs if trapped beneath unconformities (Heald et al., 1979; Shanmugam, 1988, 1990; Shanmugam and Higgins, 1988). Furthermore, this study allows one to see first hand the secondary changes associated with weathering; such alterations are sometimes difficult to distinguish from deep diagenetic changes when studying subsurface unconformity surfaces (Heald et al., 1979). This study found that other than the alteration of muscovite to vermiculite, the major chemical operator acting on the Eaton sandstone at the present time is dissolution. What is noteworthy, however, is that K-feldspar appears stable. Emery et al. (1990), for example, found that the abundance of potassium feldspar below an unconformity increased with depth, while the kaolinite abundance decreased with depth, suggesting potassium feldspar leaching by meteoric water resulted in concomitant kaolinite precipitation as a. result of subaerial exposure. Chittleborough (1989), too, reports microcline altering to kaolinite in a soil developed on a feldspathic sandstone. ‘Young (1986) had similar findings 82 83 for the Bungle Bungle Massif of AustralLa. An absence of kaolinite below the Cimmerian unconformity in the northern North Sea was explained by Bjorkum et al. (1990) to be the result of an erosion rate exceeding the propagation rate of the dissolution/weathering front. Such an explanation is inappropriate for explaining the absence of epidiagenetic kaolinite at Grand Ledge since the weathering zone is well preserved and erosion is at an absolute minimum. In addition, Arditto (1983) found.kaolinite to be the major stable phase in the intake beds of the Great Australian basin, with K-feldspar and muscovite dissolution providing the ions for the kaolinite formation. The reasons why K-feldspar may be stable, or metastable, in outcrop have already been discussed. Let it simply be stated here that the Eaton sandstone is a modern day exposure surface in which neither potassium feldspar is altering to kaolinite, nor is the erosion rate exceeding the propagation rate of the dissolution front. This study agrees with Bjorkum et al. (1990) who state that kaolinization may not be as important below unconformity surfaces as once was believed. CHAPTER 6: HONEYCOMB WEATHERING As mentioned earlier, honeycomb weathering is ubiquitous at Grand Ledge, and is apparently a function of the presence of salts on the outcrop surface, the aspect of the outcrop, and the massiveness of the sandstone (Wallis & Velbel, 1985). Mustoe (1982) attributed. honeycomb ‘weathering in coastal exposures of arkosic sandstones to the evaporation of salt water deposited by wave splash, an the resulting salt precipitation physically disaggregating the sand grains. This same researcher 'mentions that the cavity’ walls are not reinforced by the weathering skin (but rather green algae), which has been suggested by other researchers. Wave splash action has also been proposed for forming miniature pits in the quartz arenites of India (Ganesh & Sathyanarayan, 1991). Smith (1982) pays great tribute to Mustoe for taking such a scientific approach to honeycomb weathering, but, in as much as Smith approves of Mustoe’s approach, he is quick to point out that honeycomb weathering occurs in other geographic regions besides the shoreline. lk1fact, Smith (1983) presents a NASA photograph of a Martian boulder on which resides honeycomb weathering! The purpose of this section is not to provide an exhaustive review of the literature on honeycomb weathering, 84 85 but rather to make a comparison of the Grand Ledge honeycombing with other researchers work. Robinson & Williams (1992) discuss the difficulty in finding an explanation for honeycombing. These researchers make the following observations about honeycomb weathering: concentrated along bedding planes, favored on massive joint blocks, and usually occur on the south and west sides of the joint blocks. These researchers propose that frost action, salt wedging, water seepage along bedding planes, and patchy protective crusts are all factors in the generation. of honeycombs. These observation correspond with those made by Wallis & Velbel (1985). Mustoe (1983) evaluated the honeycomb weathering at an inland locality, Capitol Reef Desert, Utah. He utilized chemical analyses, XRDldata, and.field observations to suggest that salt weathering is the most important cause of disintegration, possibly aided by calcite dissolution in the calcareous sandstones. Kelletat (1980) records the same observations for calcareous sandstones in western Scotland and southern Greece, for which honeycombs are full of sand and mixed with fine salt crystals. Such sandstones are, however, coastal, with saltwater spray reaching the honeycomb zone. In contrast, Gill et al. (1981) examined a honeycomb weathered greywacke with a scanning electron microscope (SEM) , and found that salt was not associated with honeycombs, even though the honeycombs occurred in the supratidal zone where the rocks are frequently wet by sea spray. 86 Several researchers have suggested that mineral weathering is responsible for initiating honeycombing. Butler and Mount (1986) propose that the chemical dissolution of selected minerals (quartz, feldspar, and.phyllosilicates), as well as salt-weathering and/or heat-moisture expansion processes resulting from splash-zone wetting and drying that produce the rock corrosion (honeycombing). Certainly quartz corrosion and phyllosilicate dissolution are occurring at Grand Ledge. Sancho and Benito (1990) suggest that for the Ebro basin.of Spain, the initiation of honeycombing is related to the presence of easily weatherable minerals. It has also been reported.that the shape of honeycombs depends on textural factors which control water circulation in the sandstone, which has a direct effect on feldspar hydrolysis, wetting- drying, and salt weathering (Sancho &(Gutierrez, 1990). Young & Young (1992) also discuss the importance of sandstone permeability to honeycombing. They suggest that the reason honeycombs are ubiquitous in.the Aztec sandstone and.absent in the Navajo sandstone is because the Aztec sandstone contains significant porosity, while the Navajo sandstone contains large joints through which rainwater is funneled, preventing meteoric water from entering the interior of the joint blocks. Admittedly, the Eaton sandstone contains many joint surfaces. However, the large porosity and mineral dissolution associated with the Eaton sandstone suggests substantial infiltration of rainwater. Finally, McGreevy (1985) reports the presence of gypsum 87 salts on the honeycombs of a Carboniferous sandstone in northern Ireland, as well as quartz etching, but is not sure if they influence honeycombing: He even states that the cause of honeycomb formation is not known. This study has little to add to the already large body of observations made regarding honeycomb weathering. Grand Ledge honeycomb characteristics that have not been dismissed by other researchers are that they occur on massive very porous joint block interiors, and that quartz and phyllosilicate dissolution are occurring. The only study which did not find salts associated with honeycombing was Gill et al. (1981). Regardless, it is readily apparent that the cause of honeycomb weathering still eludes geomorphologists. CHAPTER 7: SUMMARY AND CONCLUSIONS SUMMARY This study compares the weathering mineralogy of the subaerially exposed Pennsylvanian Eaton sandstone with the age-equivalent subsurface rocks found deeper in the Michigan basin. By comparing the effects of weathering with the pre- weathering mineralogy established from subsurface drill cuttings, this investigation has provided insight into the nature of porosity evolution below unconformities. The hypothesis that the authigenic mineralogy of the Eaton sandstone may be explained in terms of an open chemical system is not entirely supported” It holds true to nearly all elements, except for iron, and. possibly' manganese. Admittedly, in the present day goethite dissolution is occurring. However, portions of the dissolved iron load are being reprecipitated as the weathering crust, as well as at springs. The sandstone is typically'very'porous, with.corrosion.of quartz grains, and.dissolution of kaolinite and muscovite. In the subsurface, the sandstone may be classified as a quartz arenite to sublitharenite, while epidiagenetic alterations appear to have modified the composition into»a quartz arenite. 88 89 Goethite is ubiquitous in the outcrop samples, with joint faces containing anomalously high goethite abundances, and forming fracture-related groundwater ferricretes. The goethite has its iron source in pyrite and iron-bearing carbonate phases. Mass balance calculations indicate that dissolution is the most important chemical mechanism.in the outcrop. Iron is being conserved in the sandstone with the joint block interiors serving as net exporters of iron to the 1 cm thick case-hardened joint faces. .All other species (Si“, Ca”) Mg”) Mn”, Kf,£xh”y apg.Al“) are being removed in solution from the outcrop. The ion imports/exports calculated in the mass balance model may be converted into‘water chemistry concentrations and compared with the modern day water chemistry of shallow Pennsylvanian.aquifers. 'The results indicate that quartz, all carbonate phases, and pyrite are currently weathering in the shallow aquifers. The fact that K-feldspar is stable in outcrop explains why the K” concentrations of the aquifers are higher than those calculated from the mass balance model. A higher concentration of Mn2+ in the outcrop calculations of water chemistry indicate that the only manganese-bearing phase (siderite) is being dissolved early in the aquifers, and that the aquifers may be either presently flushed of Mn", or possibly the manganese has been exported from the joint block interiors to the joint face, much.as iron.is doing; .Aluminum, too, has a higher concentration in reconstructed losses from 9O outcrop relative to the aquiferu This may be explained.by the observation that all Alhebearing phases in the subsurface are stable, while unstable in outcrop. In terms of being a modern day unconformity surface, the Eaton sandstone exhibits the typical porosity enhancement. What is different, however, is that kaolinite is not forming as a‘weathering product, and.diagenetic kaolinite presently is undergoing dissolution. Unfortunately this study is unable to shed new light on the formation of honeycomb weathering. CONCLUSIONS (1) Outcrop mineral-water interactions are responsible for the silica saturation observed in the Pennsylvanian aquifers of the Michigan basin. (2) Quartz dissolution. was the result of goethite precipitation at the onset of weathering, but today is simply the result of quartz dissolving in undersaturated meteoric water. (3) Muscovite is undergoing weathering to vermiculite, as well as being dissolved. (4) Mass balance calculations indicate that the Eaton sandstone is behaving as a closed system with respect to iron, yet aluminum, as well as silica, calcium, magnesium, manganese, potassium, and sulfate, are all mobile. (5) (6) (7) (8) 91 Pyrite oxidation alone is not responsible for creating the acidic conditions 'which. dissolved. the carbonate cement. All kaolinite in the outcrop is a diagenetic clay and is not the result of feldspar' weathering; K-feldspar appears stable in outcrop. Kaolinite in outcrop is undergoing dissolution, thus implying that kaolinization of aluminosilicates below unconformity surfaces is not as important a factor in reservoir quality as was once believed. Epidiagenetic alterations appear to have modified the composition of the sandstone from a sublitharenite/quartz arenite in the subsurface to a predominantly quartz arenite in outcrop. APPENDICES APPENDIX A 92 APPENDIX A JP-95-10 Kaofinhe Kaolinite ”Me "Me 9 Goethite “lite MgGw {E Rmv WW I I I l I I ’1 30 20 10 3 We Figure 17. (a) Selected diffractograms. Sample JP-95-10 93 APPENDIX A (Cont’d) JP-95-17 Vermiculite Kaolinite Kaolinite 4U W K-575°C we Illite Goethite Raw _ .— __..J _ (A) O M O ~A O (l)— Figure 17. (b) Selected diffractograms. Sample JP-95—17 APPENDIX B 94 JAPPTHUDIXIIB Table 15. Sample collection data. Joint Sample Trend of Height Distance from Outcrop Samples Plane Surface Sample Line From Base Joint Face 2 JP-95-1 NSOE, West N8W 6’6" JOINT-1%" JP-95-2 888E 1%-4" JP-95-3 4-6%" JP-95-4 GX-BX" JP-95-5 8%-11%" JP-95-6 11%-15" 3 JP-95-7 N53W, Southeast N65W 5’6" JOINT-2" JP-95-8 84NE 2-4" JP-95-9 4-6%" JP-95-10 GX-BX" JP-95-ll BX-ll" 4 JP-95-12 N40W, North NBOW 5’0" JOINT-2" JP-95-13 87SE 2"-4%" JP-95-14 4%-6%" JP-95-15 SK-BX" 6 JP-95-16 N55E, Southwest N55W 7’0" JOINT-2% JP-95-17 908E ZX-S" JP-95-18 5-7" JP-95-19 7-9%" JP-95-20 9%-12" JP-95-21 12-14" APPENDIX C 95 APPENDIX C Figure 18. (a) Photo of outcrop sampling. Outcrop #2 96 APPENDIX C (cont’d) u 44- 7'1?" "'1 ’9‘”, '7 ._ .3 ‘giu Figure 18. (b) Photo of outcrop sampling. Outcrop #3 97 APPENDIX C (cont’d) Figure 18. (c) Photo of outcrop sampling. Outcrop #4 98 APPENDIX C (cont’d) Figure 18. (d) Photo of outcrop sampling. Outcrop #6. APPENDIX D 99 APPENDIX D Clay Mineral Preparation and Identification Preparation Techniques Clay mineral mounts were prepared for x-ray diffraction (XRD) by a method described by Keller et al. (1986). Samples were mechanically disaggregated with mortar and pestle, and then dispersed by ultrasonification in distilled water. Separation into the clay size fraction was performed by gravity settling; The <2um. particle size fraction. was separated with a pipette, and the aliquots were then filtered onto a 0.45/1m Millipore filter and rinsed with distilled water, the suspensions being drawn through the filter with a vacuum. The filter cakes were transferred to standard petrographic glass slides, placed sample-side down, and "rolled" onto the glass using a glass stirring rod. The filter paper was then pealed off, leaving the ion-saturated clay cake adhering to the glass slide. Four oriented mounts of each sample were prepared; one saturated with potassium, one saturated with magnesium, one saturated with magnesium and glycolated at room temperature, and one with only the naturally—occurring exchange ions ("raw"). Clay Mineral Identification Identification of clays follows the procedure outlined in Eslinger and Pevear (1988). Illite was identified by sharp peaks near 10 A, 5 A, and 3.33 A. Identification of kaolinite 100 was based on two sharp peaks near 7.2 A and 3.6 A. Goethite was identified by a single, weak broad peak near 4.18 A. 2:1 clay identification required a comparison of all four preparations of each individual sample, as well as heating to 57S°CL Identification.of vermiculite relied on the expansion from 10 A to 14 A when magnesium was added to the sample preparation. Its failure to expand to 17 A when glycolated distinguished it from smectite. The absence of any 14 A or 7 A peaks after heating to 575° C, confirmed the lack of any chlorite in the sample. BIBLIOGRAPHY BIBLIOGRAPHY Adams, W.A., and J.K. Kassim, 1983. The Origin of Vermiculite in Soils Developed from Lower Palaeozoic Sedimentary Rocks in Mid-Wales: Soil Science Society of America Journal, v. 47, pp. 316-320. Al-Gailani, M.B., 1981. Authigenic Mineralizations at Unconformities: Implications for Reservoir Characteristics: Sedimentary Geology, v. 29, pp. 89-115. Arditto, P.A. , 1983 . Mineral-Groundwater Interactions and the formation of Authigenic Kaolinite Within the Southeastern Intake Beds of the Great Australian (Artesian) Basin, New South Wales, Australia: Sedimentary Geology, v. 35, pp. 249-261. Arnold, G.E., 1978. A Petrographic Study’ of Sandstone Weathering: Unpublished M.S. Thesis, West Virginia University, 54 pp. Bennett, P.C., and.D.I. Siegel, 19873 Increased Solubility of Quartz in Water due to Complexing by Organic Compounds: Nature, v. 326, no. 6114, pp. 684-686. Bennett, P.C., Melcer, M.B., Siegel, D.I., and J.P. Hassett, 1988. The Dissolution of Quartz in Dilute Aqueous Solutions of Organic Acids at 25° C: Geochimica et Cosmochimica Acta, v. 52, pp. 1521-1530. Bennett, P.C., Siegel, D.I., Hill, B.M., and P.H. Glaser, 1991. Fate of Silicate Minerals in a Peat Boga Geology, v. 19, pp. 328-331. Bjorkum, P.A., and N. Gjelsvik, 1988. An Isochemical Model for Formation of Authigenic Kaolinite, K-Feldspar, and Illite in Sediments: Journal of Sedimentary Petrology, v. 58, no. 3, pp. 506-511. Bjorkum, P.A., Mjos, R., Walderhaug, O., and A. Hurst, 1990. The Role of the Late Cimmerian Unconformity for the Distribution of Kaolinite in the Gullfaks Field, North Sea: Sedimentology, v. 37, pp. 395-406. 101 102 Bjorlykke, K., 1984. Formation of Secondary Porosity. How Important is it?: LI; McDonald, D.A. and R.C. Surdam (eds.), Clastic Diagenesis, American .Association of Petroleum Geologists Memoir 37, pp. 277-286. Bromley, M., 1992. Topographic Inversion of Early Interdune Deposits, Navajo Sandstone (Lower Jurassic), Colorado Plateau, U.S.A.: Sedimentary Geology, v. 80, pp. 1-25. Busenberg, E., and D.V. Clemency, 1976. The Dissolution Kinetics of Feldspars at 25° C and 1 atm CO2 Partial Pressure: Geochimica et Cosmochimica Acta, v. 40, pp. 41- 50. Butler, P.R., and J.F. Mount, 1986. Corroded Cobbles in Southern Death Valley: Their Relationship to Honeycomb Weathering and Lake Shorelines: Earth Surface Processes and Landforms, v. 11, pp. 377-387. Campbell, I.A., 1991. Classification of Rock Weathering at Writing-on-Stone Provincial Park, Alberta, Canada: 11 Study in Applied Geomorphology: Earth Surface Processes and Landforms, v. 16, pp. 701-711. Chigira, M., and K. Sone, 1991. Chemical Weathering Mechanisms and Their Effects on Engineering Properties of Soft Sandstone and Conglomerate Cemented by Zeolite in a Mountainous Area: Engineering Geology, v. 30, pp. 195- 219. Chittleborough, D.J., 1989. Genesis of a Xeralf on Feldspathic Sandstones, South Australia: Journal of Soil Science, v. 40, pp. 235-250. Cohee, G.V., 1965. Geologic History of the Michigan Basin: Journal of the Washington Academy of Sciences, v. 55, pp. 211-224. Dannemiller, G.T., and MLA. Baltusis, 1990. Physical and Chemical Data for Ground Water in the Michigan Basin, 1986-1989: U.S. Geological Survey Open-File Report 90- 368, 155 pp. Davis, M.W., and H.D. Bredwell, 1978. Geology; it The Nature of Grand Ledge; Grand Ledge Area American Revolution Bicentennial Commission, pp. 7-37. Drever, J.I., 1988. The Geochemistry of Natural Waters: Prentice Hall, Inc., Englewood Cliffs, 437 pp. Emery, D., Myers, K.J., and R. Young, 1990. .Ancient Subaerial Exposure and Freshwater Leaching in Sandstones: Geology, v. 18, pp. 1178-1181. 103 Eslinger, E., and D. Pevear, 1988. Clay' Minerals for Petroleum Geologists and Engineers: SEPM Short Course Notes no. 22, 413 pp. Fairbridge, R.W., 1967. Phases of Diagenesis and Authigenesis: lg G. Larsen and G.V. Chilinger (eds.), Diagenesis in Sediments, Elsevier, Amsterdam, pp. 19-89. Ganesh, A., and S. Sathyanarayan, 1991. Origin of Miniature Pits in the Quartz-Arenites of the Badami Group (Younger Proterozoic), Umatar, Belgaum District, Karnataka: Journal if the Geological Society of India, v. 38, pp. 621-624. Gill, E.D., Segnitt, E.R., and N.H. McNeill, 1981. Rate of Formation of Honeycomb Weathering Features (Small Scale Tafoni) on the Otway Coast, S.E. Australia: Proceedings of the Royal Society of Victoria, v. 92, pp. 149-154. Goudie, A., 1973. Duricrusts in Tropical and Subtropical Landscapes: Clarendon Press, Oxford, 174 pp. Heald, M.T. , Hollingsworth, T.J. , and R.M. Smith, 1979. Alteration of Sandstone as Revealed by Spheroidal Weathering: Journal of Sedimentary Petrology, v. 49, pp. 901-910. Hollingsworth, T.J., 1977. Spheroidal Weathering of Sandstones in Nicholas County, West Virginia: Unpublished M.S. Thesis, West Virginia University, 84 pp. Hudson, R.J., 1957. Genesis and Depositional History of the Eaton Sandstone, Grand Ledge, Michigan: Unpublished M.S. Thesis, Michigan State University, 49 pp. Keller, W.D., Reynolds, R.C., and A. Inoue, 1986. Morphology of Clay Minerals in the Smectite-to-Illite Conversion Series by Scanning Electron Microscopy: Clays and Clay Minerals, v. 34, no. 2, pp. 187-197. Kelletat, D., 1980. Studies on the Age of Honeycombs and Tafoni Features: Catena, v. 7, pp. 317-325. Kelly, W.A., 1933. Pennsylvanian Stratigraphy Near Grand Ledge, Michigan: Journal of Geology, v. 41, pp. 77-88. ----------- , 1936. The Pennsylvanian System in Michigan: Michigan Geological Survey Publication 40, Geological Series 34, part 2, pp. 149-226. Klein, C., and C.S. Hurlbut, 1985. Manual of Mineralogy: John Wiley & Sons, Toronto, 596 pp. 104 Kramer, R.S., and D.B. Westjohn, 1991” Textures, and.Chemical and Isotopic Compositions of Authigenic Carbonates in Pennsylvanian Sandstones in the Michigan Basin: Unpublished ‘United States Geological Survey' Report, Lansing, Michigan. Krauskopf, K.B., 1956. Dissolution and Precipitation of Silica at Low Temperatures: Geochimica et Cosmochimica Acta, v. 10, pp. 1-26. Land, L.S., and K.L. Milliken, 1981. Feldspar Diagenesis in the Frio Formation, Brazoria County, Texas Gulf Coast: Geology, v. 9, pp. 314-318. Lin, F., and C.V. Clemency, 1981. The Kinetics of Dissolution of Muscovite at 25° C and 1 atm (Kt Partial Pressure: Geochimica et Cosmochimica Acta, v. 45, pp. 571-576. Long, D.T., Wilson, T.P., Takacs, M.J., and D.H. Rezabek, 1988. Stable-Isotope Geochemistry of the Saline Near- Surface Ground Water: East Central Basin: Geological Society of America Bulletin, v. 100, pp. 1568-1577. Long, D.T., Badulamenti, L., and T.P. Wilson, 1990. The Role of Glaciation in Controlling the Geochemical Interaction of Near Surface Groundwater and Formation Brine: Geological Society of America Abstracts with Programs, v. 22, p 31. Martin, J.R., 1982. Pennsylvanian Deltaic Sedimentation in Grand Ledge, Michigan: ‘Unpublished M.S. Thesis, Western Michigan University, 131 pp. McBride, E.F. , 1987. Diagenesis of the Maxon Sandstone (Early Cretaceous), Martin Region, Texas: A.Diagenetic Quartz- Arenite: Journal of Sedimentary Petrology, v. 57, pp. 98-107. McGreevy, J.P., 1985. A Preliminary Scanning Electron Microscope Study of Honeycomb Weathering of Sandstone in a Coastal Environment: Earth Surface Processes and Landforms, v. 10, pp. 508-518. Meissnery B.D., 1993. The Geochemistry and Source for Solutes in Ground Water from the Pennsylvanian Bedrock Sequence in the Michigan Basin: Unpublished M.S. Thesis, Michigan State University, 115 pp. theissner, B.D., Long D.T., Wahrer, M.A., Bauer, P.N., Lee, R.W., and T.P. Wilson, 1992. Geochemistry and Source of Solutes in Ground Water from the Marshall Sandstone Regional Aquifer, Michigan Basin: Geological Society of America Abstracts with Programs, v. 24, no. 7, p. A240. 105 Merino, E., 1975a. Diagenesis in Tertiary Sandstones from Kettleman North Dome, California--I. Diagenetic Mineralogy: Journal of Sedimentary Petrology, v. 45, no. 1, pp. 320-336. Merino, E., 1975b. Diagenesis in Tertiary Sandstones from Kettleman North Dome, California-~II. Interstitial Solutions: Distribution.of.Aqueous Species at 100°C3and Chemical Relation to the Diagenetic Mineralogy: Geochimica et Cosmochimica Acta, v. 39, pp. 1629-1645. Morris, R.C., and A.B. Fletcher, 1987. Increased Solubility of Quartz Following Ferrous-Ferric Iron Reactions: Nature, v. 330, pp. 558-561. Mustoe, G.E., 1982. The Origin of Honeycomb Weathering: Geological Society of America Bulletin, v. 93, pp. 108- 115. Mustoe, G.E., 1983. Cavernous Weathering in the Capitol Reef Desert, Utah: Earth Surface Processes and Landforms, v. 8, pp. 517—526. Nedkvitne, T., and K. Bjorlykke, 1992. Secondary Porosity in the Brent Group (Middle Jurassic), Huldra Field, North Sea: Implications for Predicting Lateral Continuity of Sandstones?: Journal of Sedimentary Petrology, v. 62, no. 1, pp. 23-34. Nahon, D., Carozzi, A.V., and C. Parron, 1980. Lateritic Weathering as a Mechanism for the Generation of Ferruginous Ooids: Journal of Sedimentary Petrology, v. 50, pp. 1287-1298. Nesbitt, W.H., 1980. Characterization of Mineral-Formation Water Interactions in Carboniferous Sandstones and Shales of the Illinois Sedimentary Basin: American Journal of Science, v. 280, pp. 607-630. Nott, J.F., Idnurn., M., and R.W. Young, 1991. Sedimentology, Weathering, Age and Geomorphological Significance of Tertiary Sediments on the Far South Coast of New South Wales: Australian Journal of Earth Sciences, v. 38, pp. 357-373. Parron, C., and D. Nahon, 1980. Red Bed Genesis by Lateritic Weathering of Glauconitic Sediments: Journal of the Geological Society (London), v. 137, pp. 689-693. Peters, N.E., and.J.E, Bonelli, 1982. Chemical Composition of Bulk Precipitation in the North-Central and North-eastern United States, December 1980 Through February 1981: Geological Survey Circular 874, 63 pp. 106 Pettijohn, F.J., Potter, P.E., and R. Siever, 1987. Sand & Sandstone: Springer-Verlag, New York, 553 pp. Rich, C.I., 1958. Muscovite Weathering in.a Soil Developed in the Virginia Piedmont» .Ig Clays and Clay'Minerals, Fifth National Conference on Clays and Clay Minerals, National Academy of Sciences-National Research Council, publication 566, Ada Swineford (ed.). pp. 203-212. Richardson, S.M. , and H.Y. McSween, 1989. Geochemistry: Pathways and Processes: Prentice Hall, Englewood Cliffs, 488 pp. Robinson, D.A., and R.B.G. Williams, 1992. Sandstone Weathering in the High Atlas, Morocco: Z. Geomorph. N.F., v. 36, no. 4, pp. 413-429. Sancho, C., and G. Benito, 1990. Factors Controlling Tafoni Weathering in the Ebro Basin (N.E. Spain): Z. Geomorph. N.F., v. 34, no. 2, pp. 165-177. Sancho, C., and M. Gutierrez, 1990. Analysis Morfometrico De La Tafoniza De La Arenisca De Peraltilla (Anticlinal De Barbastro, Depresion. Del Ebro): Influencia De Los Factores Mineralogic-Texturales: Cuaternario y Geomorfologia, v. 4, pp. 131-145. Shanmugam, G., 1988. Origin, Recognition and Importance of Erosional Unconformities in Sedimentary Basins: lg New Perspectives in Basin Analysis, Kleinspehn, K.L. and C. Paola (eds.), Springer-Verlag, New York, pp. 83-108. Shanmugam, G., and J.B. Higgins, 1988. Porosity Enhancement from Chert Dissolution Beneath Neocomian Unconformity, Ivishuk Formation, North Slope, Alaska: American Association of PetroleuntGeologists Bulletin, v. 72, pp. 523-535. Shanmugam, G., 1990. Porosity Prediction in Sandstones Using Erosional Unconformities: Lg Prediction of Reservoir Quality Through Chemical Modeling, Meshri, I.D. and P.J. Ortoleva (eds.), American Association Petroleum Geologists Memoir 49, pp. 1-23. Siever, R., 1962. Silica Solubility, 0-200° C! and. the Diagenesis of Siliceous Sediments: Journal of Geology, v. 70, pp. 127-150. 107 Singh, R.N., Grube, W.E., Jr., Smith, R.M., and R.F. Keefer, 1982. Relation of Pyritic Sandstone Weathering to Soil and Minesoil Properties: lg Acid Sulfate Weathering, Kittrick, J.A., Fanning, D.S., Hossner, L.R., Kral, D.M. and S. Hawkins (eds.), Soil Science Society of America Special Publication Number 10, pp. 193-207. Smith, P.J., 1982. Why Honeycomb Weathering?: Nature, v. 298, pp. 121-122. Smith, P.J., 1983. Can Honeycomb Weathering be ET?: Nature, v. 301, p.291. State of Michigan, 1966. Laws Relating to Water: A Comprehensive Planning Study of the Grand River Basin, Michigan: Lansing, Michigan, 415 pp. Stearns, M.D., 1933. The Petrology of the Marshall Formation of Michigan; .Journal of Sedimentary Petrology: vu 3, pp. 99-112. Summerfield, M.A., 1983. Silcrete: lg Chemical Sediments and Geomorphology, Goudie, A.S., and K. Pye (eds.),.Academic Press, London, pp. 59-91. Tardy; Y., 1971” Characterization of the Principal Weathering Types by the Geochemistry of Waters from some European and African Crystalline Massifs: Chemical Geology, v. 7, pp. 253-271. Thiry, M., Ayrault, M.B., and J.C. Grisoni, 1988. Ground- Water Silicification and Leaching in Sands: Example of the Fontainebleau Sand (Oligocene) in the Paris Basin: Geological Society'of.America.Bulletin” v. 100, pp. 1283- 1290. Thiry, M., and A.R. Milnes, 1991. Pedogenic and Groundwater Silcretes at Stuart Opal Field, South Australia: Journal of Sedimentary Petrology, v. 61, pp. 111-127. Velbel, M.A., and D.S. Brandt, 1989. Sedimentology and Paleography of the Pennsylvanian Strata of Grand Ledge, Michigan: Michigan Basin Geological Society Field Trip Guidebook, 33 pp. Velbel, M.A., and J. Genuise, 1988. Clay' Minerals in Underclays and Other Kaolinitic Mudrocks of the Saginaw Formation and Eaton Sandstone (Pennsylvanian Grand.River Group), Grand. Ledge, Michigan; .Abst., 1988 .Annual Meeting, Clay Minerals Society, p. 35. 108 Vicente, M.A., 1983. Clay Mineralogy as the Key Factor in Weathering of "Arenisca Dorada" (Golden Sandstone) of Salamanca, Spain: Clay Minerals, v. 18, pp. 215-217. Wahrer, M.A., 1993. The Geochemistry and.Source of Solutes in Ground Water from the Glacial Drift Regional Aquifer, Michigan Basin: ‘Unpublished.M.S. Thesis, Michigan State University, 111 pp. Wahrer, M.A., Long, D.T., Meissner, B.D., and D.B. Westjohn, 1992 . Geochemistry and Source of Solutes in Ground Water from.‘Near-Surface-Bedrock and. Glacial-Drift Regional Aquifers, Michigan Basin: Geological Society of America Abstracts with Programs, v. 24, no. 7, p. A240. Wallis, J.L., and M.A. Velbel, 1985. Honeycomb Weathering of Pennsylvanian Sandstones, Grand Ledge, Michigan: Geological Society of America Abstracts with Programs, v. 17, p. 330. Weed, R., and R.P. Ackert Jr., 1986. Chemical Weathering of Beacon Supergroup Sandstones and Implications for Antarctic Glacial Chronology: South African Journal of Science, v. 82, pp. 513-516. Westjohn, D.B., H.W. Olsen, and A.T. Willden, 1990. Matrix- Controlled Hydraulic Properties of Mississippian and Pennsylvanian Sandstones from the Michigan Basin: U.S. Geological Survey Open-File Report 90-104, 18 pp. Westjohn, D.B. , and D.F. Sibley, 1991. Geophysical and Hydraulic Properties of Mississippian Sandstones from the Michigan Basin, and Their Relations to Mineralogy and Stratigraphy: Geological Society of America Abstracts with Programs, v. 23, p 26. Westjohn, D.B., Sibley, D.F., and J.A. Eluskie, 1991. Authigenic Mineral Paragenesis in. Mississippian and Pennsylvanian Sandstone Aquifers in the Michigan Basin: Geological Society of America Abstracts with Programs, v. 23, p. 327. White, J.L., 1962. X-Ray Diffraction Studies on the Weathering of Muscovite: Soil Science, v. 93, no. 1, pp. 16-21. Williams, R., and D. Robinson, 1989. Origin and Distribution of Polygonal Cracking of Rock Surfaces: Geografiska Annaler, v. 71A, pp. 145-159. 109 Wilson, M.D., and E.D. Pittman, 1977. Authigenic Clays in Sandstones: Recognition and Influence on Reservoir Properties and Palaeo-Environmental Analysis: Journal of Sedimentary Petrology, v. 47, pp. 3-31. Wood, W.W., 1969. Geochemistry of Ground Water of the Saginaw Formation in the Upper Grand River Basin, Michigan: Unpublished Ph.D. Thesis, Michigan State University, 104 PP; Young, R.W., 1986. Tower Karst in Sandstone: Bungle Bungle Massif, Northwestern Australia: Z. Geomorph. N.F., v. 30, no. 2, pp. 189-202. Young, R.W., 1987. Sandstone Landforms of the Tropical East Kimberley Region, Northwestern Australia: Journal of Geology, v. 95, pp. 205-218. Young, R.W. , 1988. Quartz Etching and Sandstone Karst: Example from the East Kimberleys, Northwestern Australia: Z. Geomorph N.F., v. 32, no. 4, pp. 409-423. Young R.W., and R.M. Young, 1988. "Altogether Barren, Peculiarly Romantic": The Sandstone Lands Around Sydney: Australian Geographer, v. 19, no. 1, pp. 9-25. Young R.W., and A. Young, 1992. Sandstone Landforms: Springer-Verlag, New York, 163 pp. Zacharias, K.F., 1992. Clay' Mineral Diagenesis in the Marshall Sandstone, Michigan Basin: Unpublished M.S. Thesis, Michigan State University, 95 pp. Zacharias, K.F., Sibley, D.F., and D.T. Long, 1992. Detrital Diagenesis, Hydraulic Conductivity and Pore Water Chemistry in the Marshall Sandstone Regional Aquifer, Michigan Basin: Geological Society of America Abstracts with Programs, v. 24, no. 7, p. A240. "11111111711111111113