lllllllllllllllllllllHIlHlllllllllllllllllllllllllllll 1111:1313 3 1293 10681 735Q ' u jmresuj :‘xka w ~ 1‘11“ 5‘ 13 c i- 3 l Mum; "1 warm-tar: A-“'— ’5 This is to certify that the thesis entitled FIELD EVIDENCE FOR SHALE MEMBRANE FILTRATION OF GROUNDWATER, SOUTH-CENTRAL MICHIGAN presented by David F. Slayton has been accepted towards fulfillment of the requirements for Masters degree in Geo] 091 Mal-A/ 44m [ Major professor Date May 21, 1982 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution MSU LIBRARIES n. RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. 1’ ma? 6 r? m“ £31 at "mags ”I #31326 502304 FIELD EVIDENCE FOR SHALE MEMBRANE FILTRATION OF GROUNDWATER, SOUTH-CENTRAL MICHIGAN by David F. Slayton A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Geology 1982 ABSTRACT FIELD EVIDENCE FOR SHALE MEMBRANE FILTRATION OF GROUNDWATER, SOUTH-CENTRAL MICHIGAN By David F. Slayton Shale membrane filtration is the process that controls the groundwater geochemistry of part of the Saginaw Formation aquifer in south-central Michigan. This process purports that certain rock types such as shales act as semi-permeable membranes which can selectively retard the movement of dissolved ions in a water migrating through the membrane. Based on the recharge information and surficial bedrock lithology, two study areas were chosen in south-central Michigan to test the occurrence of shale membrane filtration in the Saginaw Formation. Tritium analysis of selected wells, in one area, supports the presence of local "point recharge" through sands and gravels to sandstone at the surface of the Saginaw Formation. The other area, based on tritium analysis, does not show the presence of local point recharge. Ion ratios from bedrock waters in the two study areas were found to increase or decrease relative to each other as would be predicted by shale membrane filtration theory, laboratory experiments, and field studies. Water! rock interactions and mixing with other chemically different waters do not correctly explain the observed ion ratios, or can not explain all the ratios as does shale membrane filtration. Shale membrane filtration best explains the observed geochemistry of Saginaw Formation's waters from the study area where conditions are favorable for the occurrence of membrane filtration. ACKNOWLEDGEMENTS I wish to thank my master's thesis committee Chairmen, Dr. Grahame Larson and Dr. David Long, who gave freely of their time throughout the course of this investigation and were a constant source of encouragement. I also wish to thank Dr. James ‘l'row for serving on my thesis guidance committee and critically reviewing this manuscript. Thanks is offered also to Jim Vanderkloot who helped with sampling and Spent many hours assisting me with computer programs. I would also like to thank Mr. Lloyd Ness for collecting rock samples of the Saginaw Formation from Lansing Board of Water and Light water wells. ii Lo."-' - . .. TABLE OF CONTENTS LIST OF FIGURES . . . . . . . . ....... . ...... . iv LISTOFTABLES........................ v INTRODUCTION . . . . . . . . . . . . . . . . . . ...... l PurposeandScope..................... PreviousInvestigations................... I 2 GEOLOGY AND HYDROGEOLOGY OF THE STUDY AREA . . . . . . 6 6 GeologyoftheStudyArea . . . . . . . . . . . . . . . . . Aquifer Characteristics ................... 9 Recharge to the Saginaw Formation . ........ . . . . 10 Source of Dissolved Solids in Groundwater ........... 12 Tritium in the Groundwater . . ............... 14 Tritium Data ...................... 16 Interpretation of Tritium Data ............... . l6 SHALE MEMBRANE FILTRATION ................ 20 Previous Works - Theoretical ....... . ........ 20 Previous Works — Experimental ..... . . . ........ 23 Previous Works - Field Study Example ............ 25 Test of Shale Membrane Filtration Hypothesis in the Tri'County Area 0 o o ooooo o o o ooooooooo 27 METHOD OF ANALYSIS ................... . . 29 RESULTS AND DISCUSSION . . . . . . . . . . . . . ..... . 32 Results of Chemical Predictions . . . . . . . . . . . . . . . 36 SO /HC03 O O O O O O O O I O O O O O O O O O O O O O O 0 40 Cl OCO3. O O O O O O O O O O O O O O 0 O O O O O O 0 O 42 Cl/SO 0000000000000000000 o o o o o 43 CdMé. O O 00000 O O O ......... O O O O O O a“ NdK O O O O O O O O 0 O O O O O O O O O O O O O O O O O O 45 CdNa O O O O O O O O O O O O O O ....... O O O O O #6 Cl/Ca O O O O O O O O O O O ......... O O O O O O (‘7 TBS O O O O O ........ O O O O O O O O O O I O O O #7 summary 0 ..... O O O O O O ...... O O O O O O O #7 Conclusion ........................ 49 BIBLIOGRAPHY ........................ 52 APPENDIX ........................... 56 iii Figure 1: Figure 2: Figure 3: Figure 4: Figure 5: Figure 6: LIST OF FIGURES LocationofStudyArea . . . . . . . . . . . . . . . Generalized Surficial Bedrock Lithology ........ LocationofStudyAreasAandB . . . . . . . . . . . Wells Analyzed for Tritium Content ..... . . . . . Wells Analyzed for Water Chemistry. . . . ...... Generalized Piper Diagram .............. iv 11 13 15 30 33 Table 1: Table 2: Table 3: Table 1:: Table 5: Table 6: Table 7: Table 8: LIST OF TABLES Tritium Content (T.U.) of Water Wells ..... . . . . . Water Analysis of Drift and Rock Wells ......... Average Ion Ratios for Drift and Saginaw Fm. . . . . . . Average Ion Ratios for Study Areas A and B . . . . . . . Comparison of Study Area A and B Ratios . . . . . . . . Comparison of Another Field Study to This Study . . . . . Ion Ratios of All Drift and Rock Wells . . . . . . . . . . Wells from Study Areas A and B Used for Ion Ratio Comparisons . . . . . . . . . ........ . 17 66 37 38 39 41 73 80 INTRODUCTION Purpose and Scope This research tested the hypothesis that shale membrane filtration controls the groundwater geochemistry of the Saginaw Formation in south-central Michigan. Shale membrane filtration (reverse osmosis) is the process whereby certain rock types can selectively remove or retard the movement of ions in water migrating through the formation (White, 1965; Berry, 1969; Hanshaw and Coplin, 1973; Kharaka and Berry, 1973). The theory of shale membrane filtration states that some geologic formations such as clay and shale can act as semi-permeable membranes. Specifically, if water is forced through the membrane, selective retardation of dissolved ions can occur. However, due to various factors discussed later, certain ions are filtered out or have their movement through the membrane retarded more with respect to other ions. As a result, water on the "input" side would eventually become more concentrated than water on the "output" side in ions that are retarded or could not pass through the membrane. Hence, waters on Opposite sides would be chemically different. The working hypothesis of this study is that these differences in chemistry are predictable from experiments, field data, and the general theory of shale membrane filtration. These experimental and field data will be compared with groundwater chemistry data from the Tri-County study area to see if shale membrane filtration can account for the observed groundwater quality. Previous Investigations The geology of the Saginaw Formation in the Tri-County area has been described in detail by Kelly (1936), and by Mecenberg (1963). The hydrology of the Saginaw Formation has been studied by Stuart (1915), Firouzian (1963), Wheller (1967), Vanlier and Wheeler (1968), Radfar (1979), and Ritter (1981). In addition, a report by Vanlier (1964) contains a general discussion of the geology, hydrology, and water quality of the Tri-County area. A similar but more detailed report was also done by Vanlier, et al., 1973. Previous work on the geochemistry of the groundwater in the Saginaw Formation has suggested shale membrane filtration (reverse osmosis) as the best explanation for the observed difference in water chemistry between the bedrock Saginaw Formation and the overlying glacial drift (Wood, 1969, 1976; Vanlier, et al., 1973). Another study by Radfar (1979) studied potential recharge areas using water quality data and concluded that shale membrane filtration was not the controlling process on groundwater chemistry in the Saginaw Formation. Wood (1976) preposed ion filtration as the explanation for the observed distribution of dissolved solids in the Saginaw Formation aquifer. The 1976 paper hypothesized that shale beds in the Saginaw Formation aCt as ion filters that allow water of lower dissolved-solids content to pass through, into sandstone of the aquifer. Although the Kharaka and Berry (1973) paper was cited, the retardation sequences presented are not used, and Wood assumed all ions are filtered nearly the same. As well, when comparing chemical data, Wood does not use ion ratios, only individual ions total dissolved amount. The comparisons were only between drift water and Saginaw Formation water (in which only it of 9 parameters measured to test the hypothesis were accepted at 10% level of significance), and between wells cased in sandstone versus wells cased in shale (in which comparison only 3 of 9 hypothesis were accepted at 1096 level of significance). Additionally, the retardation sequences quoted by Wood (1976) from Kharaka and Berry (1973) are general sequences, not specific for membrane composition. Wood also states that point recharge is not likely to control the water composition. However, tritium analyses for this study, and Ritter's (1981) indicate that point recharge areas exist, and therefore cannot be ruled out as a control of bedrock water chemistry. Thus, if Wood makes comparisons by averaging all Saginaw Formation wells regardless of location, point recharge and shaley areas are mixed, potentially masking any filtering in the shaley areas. Nevertheless, the data collected can be interpreted to indicate that the groundwater chemistry of the Saginaw Formation is controlled by shale membrane filtration. This study goes farther since two distinct study areas were Chosen in the bedrock, one favorable for filtration, and the other not favorable (point recharge areas). These two distinct areas were compared as well as drift to bedrock water. It is better to compare a dominantly shaley area to a dominantly sandstone area. Secondly, ion ratios were used for comparisons, otherwise a simple dilution model could explain the lower dissolved-solids content of the bedrock. These ratios are a more sensitive measure of a change in chemistry. This study used retardation sequences with ions being "filtered" or retarded to different degrees, not all the same. Additionally, this study used retardation sequences specific for the membrane composition (shale) found in the Saginaw Formation aquifer. This research is important because shale membrane filtration has not been verified in this or any other shallow aquifer system. The research will be presented along the following lines. The first section of this study will look at the geology and hydraulics of sandstone and drift in the study area, and will attempt to confirm suspected recharge areas within the Tri-County region. For example, previous studies (Larson, 1979; Ritter, 1980) have suggested areas where recharge is likely to occur to the bedrock Saginaw Formation. In this study, measurement of tritium levels will be used to substantiate Larson's study (1979), and to confirm suSpected recharge into the bedrock. Obviously these areas would not be good for shale membrane filtration since the presence of shales is a prerequisite for filtration. High rates of recharge would imply no shales at the surface of the bedrock acting as aquitards. The second section will present the theory of shale membrane filtration. Previous field and experimental data regarding shale membrane filtration will be discussed, and predictions concerning groundwater chemistry will be made. A field example of possible shale membrane filtration will also be discussed to see how it compares with experimental lab data, and to determine if experimental predictions might be modified under actual field conditions. The third section will detail what analysis procedures were used to analyze groundwater samples for various ions. Also explained is the computer modeling done to describe the groundwater chemistry. These data are then compared with theoretical predictions using chemical equilibrium analysis of groundwater from the Saginaw Formation and from the glacial overburden. The chemical compositions of waters from the input (drift) and output side (Saginaw Formation) will be determined, and computer compilation and data reduction processes will be done to determine dissolved ion ratios and equilibrium conditions. Selected ratios will be compared to trends based on shale membrane filtration lab experiments and field data. For example, the following ratios should increase in groundwater if shale membrane filtration is occurring: $04/HC03, Cl/HCOB, Ca/Na. The fourth section will present the results of the computer modeling and one other water analysis procedure, the Piper diagram. The data will be presented and compared to the predictions from shale membrane filtration theory, lab experiments, and field studies. GEOLOGY AND HYDROGEOLOGY OF THE STUDY AREA The study area shown in Figure 1 is located in the Michigan Basin, slightly south of the center of the basin. This basin contains about 14,000 feet of Paleozoic age sediments at the deepest point, plus some thin and scattered sediments of Mesozoic age. All periods of the Paleozoic are represented in the basin with the exception of the Permian. The sedimentary rocks forming the basin have the characteristic outcrOp pattern, with beds dipping toward the center of the basin at about one degree. The bedrock in most places is also covered by drift of Pleistocene age, ranging from a few feet to over 1,000 feet thick. Geology of Study Area The surficial deposits in the study area consist of unconsolidated glacial deposits of Pleistocene age. These deposits are generally 50 to 200 feet thick (Vanlier, et al., 1973), and are very complex. They fall into three broad categories: 1) outwash, kames, eskers, 2) glacial lake deposits, and 3) till. The outwash, kames, and eskers are composed of well-sorted mixtures of sand, silt, and gravel. This type of deposit is the best and most widely used as a water source of the glacial deposits. The glacial lake deposits are generally layered sequences of sand, silt, and Clay. The glacial till, on the other hand, consists mostly of unsorted and unstratified mixtures of sand, silt, clay, and gravels. Except for southwestern Eaton County where Mississippian sediments occur, and in northern Clinton County where the younger Pennsylvanian age Grand River Formation is present, the bedrock directly underlying the study area is the Pennslyvanian age Saginaw Formation (Kelly, 1936). The Grand River LOWER PENNINSULA OF MICHIGAN NORTH CLINTON Fl Ill EATON INCEun FIGURE 1: LOCATION OF STUDY AREA IN CLINTON, EATON, AND INGRAM COUNTIES Formation is primarily a sandstone and apparently was deposited on an erosional surface Of the Saginaw Formation. It is usually distinguished from the Saginaw Formation by its red color. The Grand River and Saginaw Formation are considered one complex aquifer (V anlier, et al., 1973). The Saginaw Formation, the principal aquifer in the area, consists predominantly Of sandstone and shale beds, but also includes minor coal and limestone beds. Both the sandstone and shale beds tend 'to be non-persistent and lenticular, and most layers are generally no more than 20 feet thick. However, in the Lansing area, some Of the sandstone beds are as much as 300 feet thick (Kelly, 1936). The sandstones are composed chiefly of fine grained quartz grains, with abundant light colored mica, and are cemented with silica and calcium carbonate. Many Of the sandstones also contain minor beds of coal and fossil plant fragments. The term shale seems to be used by drillers in the area for anything they do not consider sandstone, coal, or limestone (Wood, 1969). Thus "shale" may signify in some cases siltstone or underclay as well as shale. Wood (1969) analyzed two samples Of shale taken from a quarry in Grand Ledge where the Saginaw Formation crops out. Results of examination of the Clay fraction in the samples (less than 2 microns) showed that they are composed primarily of three types of clays. SampleA was a black fissile shale with a cation exchange capacity of 67.7 mqu 100 gr. for calcium/magnesium, and 52 mqu gr. for K/NHu. Sample B was a light brown siltstone to shale with a cation exchange capcity of 65.5 meq/100 gr. for Ca/Mg, and 48 meq/lOO gr. for K/NHl‘. The sample's clay fraction consisted of the following: A E 59% Kaolinite 142% 30% Illite 46% 1 1% Vermiculite 12% Coal was the main reason that the Saginaw Formation was first studied, but it is rarely reported in well logs in the Tri-County area. Limestone occasionally appears on well logs, but it is usually quite thin (less than 2 feet thick). Aquifer Characteristics The origin Of the water in the drift and Saginaw Formation is precipitation, which averages about 31 inches per year. According to U.S.G.S. records (1968), stream runoff averages about 7.6 inches per year. Wood (1969) states that the difference, about 23.4 inches, is the approximate loss per year to evapotranSpiration. Of the 7.6 inches that comprises stream runoff, it was found by graphical methods that 58%, or 4.1 inches, was derived from groundwater sources (Wood, 1969). Pump tests and water level data from wells throughout the study area indicate that the Saginaw Formation acts as a leaky artesian system over most of the area. Wheeler (1967) estimated the average value of leakage between the drift and the Saginaw Formation to be approximately 0.0012 gpd per square foot under existing head conditions. This value represents leakage before extensive pumping in the Lansing MetrOpOlitan area. The many pumping tests indicate the sandstone of the Saginaw Formation has a permeability of about 100 gpd per square foot, and that the shales are more variable with permeabilities ranging from 0.01 to 1.0 gpd per square foot (Wood, 1969). The transmissibility of the formation is about 23,000 gpd/ft (Stewart, 1945; Firouzian, 1963). Assuming steady-state groundwater conditions, Wood (1969) estimated that one-seventh (1/7) Of the water in a stream under base flow is from the Saginaw Formation, with the remaining water coming from the overlying drift. Also, the piezometric surface Of the Saginaw aquifer seems to be a smoothed reflection of 10 the surface topography, suggesting a hydrologic connection between the glacial drift and bedrock. The Saginaw Formation is confined below by the Bayport Formation which is a dense limestone approximately 40 feet thick. This formation acts as a base to the flow system in the Saginaw Formation; it also prevents all but small quantities of water from entering the Saginaw Formation from below (Wood, 1969). Some contamination Of the Saginaw aquifer, however, has been documented from water originating in underlying formations, but it usually is associated with abandoned, poorly cased and sealed mineral water wells that allow leakage Of salt water from lower formations (Wood, 1969). The Bayport Formation directly underlies the glacial drift in southwestern and extreme western Eaton County. Since the aquifer in this area is composed almost entirely Of limestone, it was not considered in this study. In summary, precipitation is the source of water for the Saginaw Formation which receives recharge water through the overlying drift. The pattern of flow is dependent on topography, formation thickness, and permeability contrasts (Wood, 1969). The Bayport limestone acts as a base to the Saginaw system, retarding flow into, and out of the Saginaw Formation from below. Recharge to the Saginaw Formation Figure 2 is a bedrock map of the southeast part of the Tri-County area. The map shows that eastern Ingham County is underlain predominantly by shale, whereas western Ingham and eastern Eaton County is underlain predominantly by sandstone. Figure 2 also shows various areas which are favorable for rapid recharge to Saginaw Formation sandstone through sands and gravels in the drift. All Of these areas occur in western Ingham and eastern Eaton Counties (Larson, 1979; Vanlier, et al., 1973), and would indicate areas where shale membrane filtration would not be operating. Therefore, on the basis of lithology, two study 1 |II l' I ll lllll‘lllll l l | EATON 00. new 2: NAP or GENERALI- SURFICIAL mmm LITHOLOGY CJSAIESTOIE -SHALE 12 areas were selected (Figure 3). One area (study area A) is assumed to have slow recharge rates and to favor shale membrane filtration, and the other (study area B) is assumed to have high recharge rates and be an area where filtration would not occur. Source of Dissolved Solids in Groundwater The water chemistry of the Saginaw aquifer is dependent on the source, distribution, and movement of water. Thus to understand the water chemistry, the source Of recharge water must be identified, the chemistry of the recharge water should be known, the source Of dissolved ions should be determined, and the interaction between the water and aquifer materials should be understood. According to Wood (1969), precipitation is not a major source Of dissolved solids in the water from the Saginaw aquifer. The only ion that precipitation may contribute in significant quantities is chloride. Other ions are present in precipitation, but their very low concentration cannot account for the observed concentrations found in groundwater from the Saginaw Formation and glacial drift. Since recharge to the Saginaw Formation is through the overlying drift, the drift would be the next Obvious potential source of dissolved solids. According to an experiment by Wood (1969), water leached from samples of soil-glacial material after standing 5 to 7 days was similar to the ion composition of water Obtained from the drift and Saginaw Formation. The results Of the experiment showed that most Of the dissolved ions probably were dissolved from glacial drift, and that much of the calcium and sulphate was probably derived from solution of detrital gypsum in the drift. Wood also states that the sulphate is not due to the oxidation of sulfide minerals because there is a one-tO-0ne relationship between sulphate and non-carbonate hardness, and that the major source of calcium, magnesium, and bicarbonate appears to be from the reaction l3 CLINTON CO . R514 R214 R114 R1E R23 EATON CO. INGHAM 00. FIGURE 5 : LOCATION OF TOWNSHIPS USED FOR SPECIFIC STUDY AREAS HID Hyper-filtration model study area A a Point Recharge model study area B 14 Of carbonic acid with limestone and dolomite in the drift. Chloride and sodium probably come from the solution of halite, with some sodium contributed by ion exchange with clay minerals. Leaching experiments were also performed by Wood (1969) on rock samples Of the Saginaw Formation, and produced very small concentrations of dissolved solids in the leachate. The results of the experiments show that some sodium, however, might be contributed by the Saginaw Formation through cation exchange in shales. Also, the underlying Michigan Formation does not appear tO contribute dissolved solids, as the formations (Michigan and Bayport) do not seem to be hydraulically connected to the Saginaw aquifer. This is also supported by chemical analyses of the Saginaw and Michigan Formations' waters since they are chemically different. If water from the Michigan Formation did migrate into the Saginaw and contribute dissolved ions, it would change the chemistry or composition of the Saginaw, and this phenomenon is not Observed. Tritium in the Groundwater By analyzing for tritium content Of water from the Saginaw Formation, the relative age of the water can be determined, and can provide an estimate Of the rate Of recharge to the bedrock. Groundwater samples were collected from domestic water supply wells in the drift and bedrock aquifers. One set Of samples came from study area "B" near Eaton Rapids where a study by Larson (1979) located several potential recharge areas to the bedrock aquifer through sands and gravels. Other samples from wells came from study area "A" near Williamston, and a shaley area in Clinton County where recharge to the bedrock may pass through a shale. Figure 4 shows the location of wells sampled for tritium content. 15 CLINTON CO. NORTH T3N TZN ' .’ TlN R3W RZW RIM RIE R2E FIGURE 4: LOCATION OF WELLS SAMPLED FOR TRITIUM STUDY 0 DRIFT WELL 0 ROCK WELL l6 Tritium Data The tritium content of groundwater analyzed for this study ranged from 0.0 T.U. to 118 T.U. Two samples were analyzed three times to check reproducability. One of the samples (Avery) averaged 60.8 T.U. with a standard deviation Of 5.04, while the other (Herriff) averaged 71.6 T.U. with a standard deviation of 3.61 (see Table A, Appendix). The results Of tritium analyses are summarized in Table l. "T.U." stands for tritium unit, a standard unit of 18 measure for tritium content where 1T.U. = 1 tritium atom for every 1 x 10 hydrogen atoms. Interpretation of Tritium Data The average tritium content of water from 23 wells tapping the bedrock Saginaw Formation is about 18 T.U., while water from 7 drift wells averaged about 25 T.U. Thus, the water from the drift seems to contain more tritium, and hence appears to be relatively younger. This would be expected since the drift receives recharge directly from precipitation, which today contains approximately 20-45 T.U. There appears to be no apparent correlation between tritium content and depth Of well; this probably reflects the complexity Of both the drift and bedrock aquifers. It appears that the tritium content depends on the aquifer's complexity, depth, and degree Of interconnection with surface or rain water recharge. One area that directly shows younger drift water over older bedrock water, separated by shale, is in Bath (05N01W). Here a drift well 79 feet deep contained 17 T.U., whereas three nearby rock wells completed below the shale in sandstone (200, 260, and 425 feet deep, reSpectively) all contained 0.0 T.U. Other pairs Of wells (drift and rock well close together) in the study area also indicate that water in the drift is relatively younger than in the bedrock. One pair, OlN03W27CC (drift) and OlN03W28DD (rock) contained 14 T.U. and 6 T.U., 17 TABLE 1: Location number consists of township, range, section, and quarter, with the largest quarter given first. The quarters are divided with A = northeast, B = northwest, C = southwest, and D = southeast. TRITIUM DATA Precipitation Date T.U. Dec. 1979 42.9 i- 8.4 Jan. 1980 43.1 i 6.2 Mar. 1980 34.4 i 7.7 Apr. 1980 22.8 i 8.9 May 1980 26.3 i 5 6 Surface Water Grand River 54.9 i 5.9 Red Cedar 47.0 i 6.5 Groundwater Depth Drift Wells Location (feet) T.U. Bath 05N01W28CA 79 17.6 i 7.4 Wilson 01N03W27CC 121 14.2 i 8.8 Maxey 02N 03W 12CC - 68 . 2 i 9 . 1 Iverson 03N01E33AB 25 21. 5 i 5 .1 Roadside park 04N01E29BC - 52.2 i 5.5 Smithville dam 01N03W02DB 1.2 i 6.5 Geyer 04N01E26AA - 1.9 i 4.9 3.9M Bath townhall 05N01W20AC 425 0.0 i 6.3 Bath 05N01W28BA 200 0.0 i 7 .1 Bath 05N01W20BB 260 0.0 i 5.6 Johnson 02N02W 30BC 196 3 . 4 i 7 . 2 Avery 02N02W06BD - 57 .l i 8 . 6 Herriff 02N02W20AB 150 69 .1 i 8 . 2 Listing 02N02W ZODC 139 6 . 7 i 9 . 3 Hartenburg 01N02W06BC 115 64 . 3 i 8 .6 TOpliff 02N02W20BB - 27 .1 i 8 . 4 Allen 02N02W18CC 216 4.0 it 7.2 M.S.U. tap water 04N01W - 8.9 i 7.6 Maxey 02N03W12CC - 118.2 i 5.7 Church 02N03W 27DA 100 2 . 0 i 6 . 9 Kowalski 02N03W12AD 120 0.0 i 8.5 SR 8 03N01W06BD 89 19.2 i 7.9 Battley 01N03W28DD 137 5.9 i 6.7 Ness (I...) 03N03W34CC 155 2.6 i 5.4 Ness (R.) 03N03W30CC 60 18.4 i 5.3 Iverson 03N01E33AB 186 1.9 i 3. 8 Williamston, city O3N01E02 200 7.7 i 3.3 Chubb 03N01E33AB - 0.4 i 3.9 Hayes 04N01E27BA - 13.1 i 2.8 Shaulis 04N01E26AA 240 4 . 2 i 2 . 9 * May be rock wells, no logs available, assumed to be drift. 14* 18 respectively. Another pair (03N01E33AB) separated by shale showed 21.5 T.U. in drift water, and 1.9 T.U. in bedrock water. Four of the bedrock wells tested contained 0.0 T.U.‘s, and three contained 2.0 T.U.‘s or less. All of these wells were either deep, or in areas where the surface of the bedrock contained shale. The highest values of tritium came from wells in areas where the geologic conditions favor rapid recharge to the bedrock aquifer (study area B). Figure 4 shows the location of wells sampled for tritium, and Figure 3 shows the location Of areas where conditions are favorable for rapid recharge. Comparison of the two figures clearly shows that the prOposed rapid recharge area (study area B) has the highest tritium levels in the bedrock. Bedrock wells from study area "A" averaged 5.5 T.U., and bedrock wells from study area "B" averaged 35.2 T.U. Ritter (1980) also analyzed the tritium content of water samples from wells in Meridian Township, which is thought to be a recharge area to the Saginaw Formation. Ritter's study generally shows higher tritium content in rock wells close to where sands and gravels directly overlie sandstone than in wells further distant. This suggests that areas where the drift consists chiefly Of sand and gravel in contact with sandstone of the Saginaw Formation do indeed transmit water readily to the bedrock. Some rock wells in shaley areas near Williamston contain some tritium, but generally in very low levels. These wells tap sandstones below a layer Of shale that acts as an aquitard. Thus, descending water must pass either through the shale or fractures in the shale. This is probably the path of recharge to the bedrock since there are no areas characterized by sand and gravel lying directly over sandstone. In addition, tritium values in the aquifer are low, which does not support the occurrence of rapid recharge. The presence of some tritium, however, suggest slow leakage downward through the shale. 19 In conclusion, tritium levels suggest that select areas located by lithologic considerations do seem to be rapidly recharging water to the Saginaw Formation since wells in these areas (particularly study area B) are relatively higher in tritium content. On the other hand, in the shale membrane filtration area (study area A), tritium levels and subsurface geology do not indicate any rapid recharge system and thus recharge to the Saginaw Formation is probably slow through the shales present at the surface of the bedrock. The levels Of tritium in study area A, however, do suggest some recharge is occurring, although it is much slower than in study area B. SHALE MEMBRANE FILTRATION Previous Works - Theoretical The process Of osmosis occurs when two chemically different solutions are separated by a semi-permeable membrane which allows the passage of water. A chemical potential difference across the membrane is created. The net movement Of water will be from the low concentration side to the high concentration side. This water movement causes an increase in pressure on the high concentration side, with the pressure differential across the membrane known as the osmotic pressure. When the osmotic pressure equals the chemical potential difference, the flow of water will cease. The process of reverse osmosis occurs if a hydraulic pressure greater than the osmotic pressure is applied to the high concentration side of a membrane. Water will then be forced through the membrane in the direction Opposite to osmosis flow. Reverse osmosis is called shale membrane filtration when the semi- permeable membrane involved is shale (Toerell, 1935; Meyers and Sievers, 1936). Other early papers suggesting that clays and shales act as membranes include Schlumberger, et al., 1933; Russell, 1933; Mackay, 1946; De Sitter, 1947; Korzhinskii, 1947; Wyllie, 1948, 1949, 1951, 1955; White, 1957; and others. An extensive review on the subject is found in Berry, 1969. There is substantial evidence that clays and shales act as semi-permeable membranes, due to the negative electric charges on the clay particles (Kharaka and Berry, 1973). The negative electrical sites on the surface and edges of clay particles are caused by 1) substitution of lower valence cations for higher +3 valence cations such as Al and Si”, 2) broken bonds along clay particle edges, 20 I" 21 3) removal Of the hydrogen from an exposed hydroxyl group and its replacement by an exchangeable cation, and 4) under certain conditions, structural cations other than H+ become exchangeable (Kharaka and Berry, 1973). Hanshaw and COplin (1973) demonstrated that anions, such as Cl' and 504, are excluded from passing through a clay by a layer at the surface of the Clay particle known as the Gouy-Chapman double layer. Electrostatic forces attract cations to the negatively charged clay platelet, and repel anions. This forms a "diffuse layer" between the clay platelet and the equilibrium solution in which the concentrations of cations is higher than in the equilibrium solution. The diffuse layer and the negatively charged clay surface form the Gouy-Chapman double layer. With compaction of a clay sediment, the particles of clay are brought closer together, which causes the double layers of the particles to overlap. This lowers the concentration of the anions in the pore solution since they are repelled by the negative charge on the clay surface. This anion exclusion is reSponsible for the membrane pr0perties of clays and shales (Kharaka and Berry, 1973). The flow Of ions through a geologic membrane is a function Of the following: 1) the concentration Of the ion in the pore solution, 2) the water velocity through the membrane, 3) the electrical interaction of the ion with the negative sites on the clay particles, and 4) the electrical interaction Of the ion with the "streaming potential". The streaming potential is caused by the diSplacement Of the double layer (composed mostly of cations) by water movement through the membrane. Since the cations are diSplaced, some negative sites on clay particles would be exposed, and cations that were associated with the negative sites would be diSplaced toward the direction Of water flow. The effect is the retardation Of cation flow within the membrane due to attraction with exposed negative sites on clay particles and repulsion by 22 the cations previously associated with the negative sites. The same phenomena causes the acceleration of the flow of anions within the membrane. An overall effect Of this potential is to cause the concentration of anions and the concentration of solutes to be greater in the effluent solution than its concentration in the pore fluid (Kharaka and Berry, 1973). The efficiency of a geologic membrane in retarding the flow of dissolved ions is dependent on: 1) the temperature Of surrounding solutions, 2) the composition of the membrane, 3) water velocity through the membrane, and 4) the ionic strength Of the solution. Membrane efficiency decreases with higher temperature Of the solutions, and is attributed to two factors. The first is that flow rates are higher because of lower water viscosity, and higher flow rates cause a decrease in efficiency (Milne, et al., 1964) since ions are prOpelled through the membrane faster. The second reason is that ionic association usually increases at higher temperatures, and complexing may affect the charge and hydrated radius Of ions which in turn primarily control selective adsorption Of cations by clays. Increasing ion complexing would mean more neutral ion complexes in the solution which would decrease the liklihood that the ion complex would be adsorbed. McKelvey and Milne (1962), Milne, et a1. (1964), and Kharaka (1973) show that the geologic membrane's efficiency increases as the hydraulic-pressure gradient decreases. This is important since hydraulic- pressure gradients ued in the experiments are much higher than those found in field situations. Thus the lower hydraulic-pressure gradient in the field should mean an increase in a geologic membrane's efficiency over laboratory experiments. In summary, low temperature, low flow rate, low ionic strength, and high cation exchange capacity increases the efficiency of the membrane. Kharaka and Berry (1973) predicted a retardation sequence for ions based on ionic charge, hydrated radii, and hydraulic drag. Higher ionic charge and 23 larger hydrated ionic radii should increase the degree of electrical interaction between ions and the membrane which would increase retardation Of the Ion. Hydraulic drag is the force flowing water has on dissolved ions in solution, prOpelling them in the direction Of flow. Increasing hydraulic drag should decrease retardation on an ion because the ion would be prOpelled through the membrane faster. The retardation sequences as a function of increasing retardation by a membrane predicted by Kharaka and Berry (1973) were as follows monovalent cations Li < Na < K < NH 4 < Rb < Cs divalent cations Mg < Ca < Sr < Ba anions HCO3 < I< B < 504 < Cl < Br The sequences mean, for example, that for anions the HCO3 ion is retarded the least from flowing through the membrane in comparison with other anions. Previous Works - Experimental Evidence that shales could act as semi-permeable membranes was presented by Wyllie (1948), who measured the electric potential developed across a shale separating NaCl solutions of unequal concentrations. Osmotic pressures were measured across compacted shale samples separating different solutions by Kemper, 1961, and young and Low, 1965. The first filtration experiments using geologic membranes were conducted by McKelvey and Milne (1962) who forced different concentrations of NaCl solution through bentonite and shale under 10,000 psi compaction pressure. The ratio Of input to output concentration of ions, called the filtration ratio, was reported as high as 8 with bentonite, and 1.7 with shale. McKelvey and Milne (1962), and Milne, et al. (1964) further showed that the filtration ratios increased (membrane efficiency increased) with a decrease of the input fluid pressure. Hanshaw (1962), and Hanshaw and COplen (1973) also studied the filtration of 24 NaCl solutions by using montmorillonite and illite membranes, and reported a good agreement between the filtration ratios and those ratios predicted from experimental studies, with filtration ratios increasing with solutions of lower salinity. Kharaka and Berry (1973) studied the flow Of artificial sea water and chloride solutions of alkali and alkaline earth metals through compacted bentonite, illite, and shale samples. These are important studies in that Kharaka and Berry first used solutions similar to what may be expected in the field, not just NaCl solutions. The investigations by Kharaka and Berry (1973) confirmed trends established in previous works, confirmed their own theoretical predictions, showed that the efficiency of a membrane increased with increased compaction pressure, and that geologic membranes are specific in that the retardation sequences of dissolved species are different depending on the experimental conditions (temperature, pressure) and the composition of the membrane. The experimental filtration ratios obtained in the study by Kharaka and Berry (1973) generally are the same as the qualitative predictions, and also about the same as filtration ratios from field studies. The retardation sequence obtained for monovalent cations in the 1973 study by Kharaka and Berry using bentonite is: Lioq eo>comn0 02km 29 ..<.. no.8 comumhacoaxz 3 do; room :9. 88 «828m 86.5 £8888 22.8—Oman 4on E voao~o>oe mm fioeoE 533:: so Comma ohm mecob Couofivocm .=<.. no.8 363.6. .83th ocmanoE oficm nouooamsm was :9. moan 33m omen—Loo. Coho cooBuon Loam? Swarm Bocfiwmm no 838 :2 mmoam 33m mo canton—BOO ”n mam/2. 40 lists the observed trend, and the last column shows the level of significance as calculated from a "t" test. Table 6 compares a field example discussed earlier Of possible shale membrane filtration (Kharaka and Berry, 1974) and theoretical/laboratory predictions to the results of this study. The results contained in Tables 5 and 6 are significant since the ion ratios behave as predicted by theory, while at the same time comparing favorably to the field study by Kharaka and Berry (1974). All ratios increase or decrease as predicted when comparing the two study areas. The levels Of significance are low in several cases, but this is most likely due to the leaky artesian system present in the Saginaw aquifer. The levels of significance are low in several cases, but this is most likely due to the leaky artesian system present in the Saginaw aquifer. Not all water is filtered in the Tri-County region, such as in study area "B", and mixing with filtered water in the Saginaw Formation would reduce the observable effects of filtering. Secondly where there is shale, the water may recharge to some extent through cracks and fissures in the shale, thus diluting the filtration effect. The following is a more detailed discussion of the individual ion ratios with any other controlling factors addressed besides filtration. The discussion is based on the data in Table 5 which is a comparison of Saginaw Formation water between study areas "A" and "B". sommco3 Since SO is retarded relative to HCO ratio should be 4 3’ 3 lower in water that has been filtered as compared to non-filtered water. the SO 4/ HCO Therefore, if shale membrane filtration is occurring, as water moves from the drift to the bedrock through clays and shales, the ratio should decrease in the effluent or bedrock water (study area "A") relative to the non—filtered bedrock water (study area "B") and the drift water. Saginaw Formation water from the 41 TABLE 6: Comparison of field example from California study to Observed trends found in Saginaw Formation water. "A" would be non-filtered, and "B" filtered in this case. COMPARISON OF KHARAKA‘S FIELD DATA TO PRESENT STUDY INPUT A ——————————— - -—-_-— SHALE B _____ _ ._ _ __ OUTPUT A to B A to B KHARA KA PRESENT GEN ERALL Y* * 5121.12 my PREDICTED Ca/ N a decreased decreased* either TDS decreased decreased* decrease Cl decreased decreased* decrease HCO 3/Cl increased increased increase N a/ K increased increased* increase HCO 3/ SO 4 increased increased* variable SO #/Cl decreased decreased* variable * Significant at greater than 10% level. ** Based on previous studies experimental and field data. For this study, A = drift, b = Saginaw Formation. .u’ 42 shale membrane filtration study area "A" does indeed have a smaller ratio than bedrock water from study area "B", or water from the drift (Table 5). A comparison of the average drift water ratio tO bedrock water from the shale membrane filtration study area "A" shows a significant decrease or lower ratio in the bedrock water as predicted. The bedrock SOQIHCO3 ratio from study area "B" is also less than drift water. This means that the lower ratio of bedrock water could be explained by dilute water recharging the bedrock through "clean" sands and gravels without filtration. However, a comparison of the two study areas "A" and "B" shows that study area "A" has an even lower ration than area "B". This can best be explained by shale membrane filtration where SO 4 is retarded from passing through a membrane relative to HCO3. Computer analysis Of the water data indicates that calcite is super- saturated in bedrock water. If calcite did precipitate, then the effect would be to increase the ratio. However, shale membrane filtration theory predicts a decrease in the ratio, and that is Observed. If precipitation of calcite is tending to increase the ratio in the bedrock versus the drift, it is not a strong enough effect to overcome the apparent effect of filtration that decreases the ratio. Cl/HCO3 This ratio should decrese or be smaller in water that has been filtered. A comparison of drift water to Saginaw Formation water from study area "A" shows a significantly lower ratio in the bedrock as would be predicted by shale membrane filtration. The bedrock water from study area "B" is also much lower than drift water. As well, the bedrock water from area "A" is lower than area "B" when comparing this ratio. The ratio behaves as would be predicted in all three cases. Less salt or NaCl in clean sands and gravels could account for the lower ratio since it is easily dissolved and would be "washed" out rapidly. Thus direct 43 recharge could explain the lower ratio in the bedrock as compared to the drift, but does not explain why study area "A" has a lower ratio than area "B". If calcite was precipitating and removing CO3 from the system, it would have the effect Of increasing the ratio. This is not Observed, or may not cause enough increase to counter a decrease caused by filtration. If calcite was precipitating, Ca would also be removed, affecting the Ca/ Mg and Ca/Na ratios. Calcite precipitation does not seem to affect these ratios, however, as the discussion later suggests. Cl/ SO ,‘ Determining whether this ratio should increase or decrease depends on the composition of the membrane and temperature. At 70°C, the 50,! ion is retarded with reSpect to C1. However, at room temperature, the retardation sequences obtained for anions were variable depending on membrane composition, except that HCO3 was always least retarded. The Saginaw Formation shales clay fraction is composed of kaolinite, illite, and vermiculite. Since kaolinite has a low cation exchange capacity, and the prOperties that create cation exchange also create the filtering effect, this clay probably has minor effects on filtration. Any membrane prOperties in the Saginaw Formation likely result from illite and vermiculite. In experiments, an illite sample retards SO“ with reSpect to Cl (Kharaka and Berry, 1973). Therefore, the Cl/SO,‘ ratio should increase in any filtered water with respect to non-filtered water. A comparison of drift water to study area "A" bedrock water shows the bedrock water indeed does have a larger ratio as predicted. However, when drift water is compared to bedrock water from study area "B", the bedrock water ratio is lower than drift water. Thus filtration does not seem to occur in study area "B". The lower bedrock ratio in area "B" may be due to less NaCl present in recharging waters and the bedrock. 44 A comparison of bedrock water from the two study areas "A" and "B" shows that the Cl/SO4 ratio is larger or increases in area "A" relative to area "B". Thus the lower Cl/SO,‘ ratio in bedrock water from direct recharge areas such as area "B" as compared to drift ratios can be explained by little NaCl present in recharging waters from the drift, but the increase of the ratio in study area "A" over both drift and study area "B" becrock water can only be explained by shale membrane filtration. Ca/Mg In terms of ion exchange, Ca is preferentially absorbed over Mg (Kharaka and Berry, 1973), and therefore the Cal Mg ratio would decrease if ion exchange were occurring as water passed through a clay or shale. The other major controlling factor in shale membrane filtration for cations is hydraulic drag, and this consideration suggests that the Cal M g ratio should decrease in water passing through a clay since Mg has a greater hydraulic drag and would be propelled through the membrane faster than Ca. Most clays, considering absorption and hydraulic drag, retard the movement of Ca more than Mg, and thus the Ca/Mg ratio should be smaller or decrease in filtered water. However, the 1973 study by Kharaka and Berry showed the Clay illite to be one of the exceptions. That study showed that an illite sample retarded Mg more than Ca. Thus, the Ca/ Mg ratio should increase in filtered water. Water from study area "A" bedrock has a higher ratio than drift water. Water from study area "B" bedrock wells is nearly identical to drift water, and this would be expected since in direct recharge areas such as area "B" the drift and bedrock are directly connected. Study area "A" bedrock water, where conditions are favorable for filtration, has a higher ratio than study area "B" bedrock water as predicted by shale membrane filtration theory. 45 The only possible water/ rock interaction that might affect this ratio in the Saginaw Formation system would involve calcite or limestone and dolomite. Equilibrium data show that the drift and bedrock are both supersaturated with calcite. Thus calcite would tend to precipitate and remove Ca from the water, which in turn would have the effect of decreasing the Ca/Mg ratio. But this is not Observed as the ratio shows an increase as predicted by filtration theory, and calcite precipitation may only serve tO lessen any increase caused by filtration. Calcite precipitation would only cause a difference if precipitation occurred in only one of either drift or bedrock aquifers. Since calcite is supersaturated in both aquifers it makes no difference. M According to the given retardation sequence, K is retarded with re5pect to Na. Thus the ratio should increase in waters that have been filtered. However, drift water or "unfiltered" water has a larger ratio than bedrock water from either study area "A" or "B". This is possibly due to less NaCl in the bedrock, and was mentioned earlier in connection with the Cl/HCO3 and Cl/SO,‘ ratios. The decrease from drift to bedrock may also be due to K being released in a cation exchange process as well as Na. Illite is a potassium rich clay, and therefore K may be released. One or both of the above reasons may account for the lower ratio in the bedrock relative to drift water. However, when comparing bedrock water in the two study areas, the ratio behaves as predicted by filtration theory. Saginaw Formation water from study area "A" has a larger ratio Of Na/ K than bedrock water from study area "B". Going from drift to bedrock the ratio is smaller which is the Opposite of filtration theory, so some other mechanism controls that decrease. Yet when comparing filtered and non-filtered bedrock water, there is a noticeable change in the ratio in the direction predicted by shale membrane filtration theory. 46 same; This ratio has been used in the past as an important test to distinguish between effluent (filtered) and membrane concentrated waters (Kharaka and Berry, 1973). Field investigations have indicated that Ca is retarded with respect to Na. However, experimentally Obtained results are the Opposite, and this discrepancy is attributed to the much higher pressure gradients used in the lab studies. According to field studies, Ca is retarded with respect to Na, and the ratio would decrease in filtered water. The Ca/Na ratio does decrease from drift to bedrock water in study area "A". The Ca/Na ratio increases from drift to bedrock water in study area "B" however. And the ratio decreases significantly from study area "B" to study area "A" bedrock water. The increase from drift to bedrock in area "B" may also be due to less NaCl again in "clean" sands and gravels in the drift and bedrock. With less Na, the ratio would increase in the bedrock water compared to other drift water. Supersaturation Of the bedrock water with certain minerals as indicated by the computer analyses may also affect this ratio. If Ca were being removed by calcite precipitation, this would cause the ratio to decrease. This could account for the decrease observed in study area "A" over study area "B" bedrock water instead of shale membrane filtration. But that would require that calcite not be precipitating in area "B", otherwise the effect would be the same in both study areas. The analyses show that CaCO3 is supersaturated over the entire Saginaw Formation aquifer in both study areas. SO if precipitation of calcite is occurring, it is occurring in both areas, and cannot account for the Ca/Na ratio decrease only in one Of the study areas. 47 91192 Since the data comparing anions to cations are scarce, this ratio is included only as another indicator that filtration may be occurring. This ratio should increase or be larger in water tha has been filtered (Van Everdingen, 1968). When comparing study area "A" to area "B", the ratio is larger as predicted in area "A" bedrock water. The bedrock water ratio is smaller than the drift water, however, and may be due to less NaCl in the bedrock again, or much more Ca relative to Cl since the bedrock aquifer cement is silica and calcite, and the water is supersaturated with calcite. Regardless, some proces increased the ratio as predicted in bedrock water from study area "A" relative to area "B" water, and filtration can account for it. _T_D_§ Water from the Saginaw Formation aquifer in both study areas "A" and "B" has less total dissolved solids (T DS) than drift water. As well, study area "A" bedrock water has slightly less TDS than bedrock water from study area "B". Both comparisons are consistent with predictions bsed on filtration. Direct or point recharge to sandstone of the bedrock through sand and gravels, such as in area "B", could provide recharge water lower in TDS relative to most drift water, and shale membrane filtration also could provide lower TDS recharge water to the bedrock. Although both processes appear to be occurring, shale membrane filtration can further reduce the TDS content of bedrock water in study area "A". It would be difficult to say at this point which process, direct recharge or shale membrane filtration, is dominant in controlling the TDS levels in the bedrock water. Summary When comparing drift water to bedrock water from study area "A" where shale membrane filtration is thought to occur, 6 of the 8 parameters (ratios) 48 discussed behave as predicted by filtration theory. When comparing bedrock water from study area "A" to study area "B" (proposed non-filtration area), all 8 parameters or ratios behave as predicted by filtration theory. In study area "B", where shale membrane filtration is not thought to occur, 5 of the 8 ratios do not behave as would be expected were filtering to be occurring, suggesting that filtration does not occur in this area. Another possible mechanism to control the chemistry of the Saginaw Formation aquifer's chemistry is mixing with the lower Michigan Formation waters. However, water data as plotted on a Piper Diagram do not support that hypothesis, and mixing in this manner cannot account for all the ratio changes. Thus, mixing with Michigan Formation waters is ruled out as a major control on the chemistry of groundwater in the Saginaw aquifer. The fact that some ratios change slightly is probably due to two effects: 1) some ions are retarded nearly the same by a membrane, thus any ratio comparing two similarly retarded ions would not show much change, and 2) the Saginaw Formation aquifer is a leaky artesian system having shale aquitards that may not be very efficient in all cases at preventing water from migrating through fissures and cracks. Direct recharge also seems to be occurring to the Saginaw aquifer providing water that is not filtered to mix with any filtered water. The water chemistry in the bedrock is most likely due to both shale menbrane filtration and direct recharge systems Operating simultaneously in the Tri-County region. Further evidence supporting shale membrane filtration may be provided by pairs Of wells, one completed in the drift and one in the Saginaw Formation, in close proximity. A well pair such as this in shaley areas on the Saginaw Formation should show filtering affects in the bedrock water. The following is a list of some "well pairs". 49 05N01W28: In this section, a drift well shows a static water level of 830 feet elevation near a rock well completed below shale with a static water level of 820 feet. Thus, any water movement between drift and bedrock would be downward through the shale. All ratios except Cl/ Ca increase or decrease as predicted, including TDS. 03N01E02: This well pair in Ingham County from Vanlier (1973) is located in an area that has been noted before for its "naturally" softened bedrock water. All ratios increase or decrease as predicted, except TDS was higher in the bedrock well than in the drift well. 01N03W27: A well pair in this section showed all ratios except Na/ K behaving in the predicted manner, with a well log showing the rock well completed below 44 feet Of shale. 02N03W12CC: This well pair is in study area "B", the direct recharge area. Predictably, only 3 Of the 8 ratios increased or decreased as would be expected if the bedrock water was filtered. Most ratios show no filtering, and this seems to confirm that shale membrane filtration is not occurring in this area. 03N01E34: A drift well water elevation is about 895 feet near a rock well completed below shale with a water elevation Of 880 feet. Thus the water movement if occurring is downward which is required for filtration here in study area "A". Conclusion Two study areas were chosen based on surficial bedrock lithology and location Of probable recharge areas. One represents and area where conditions are favorable for the occurrence of shale membrane filtration (study area "A"), and the other an area where shale membrane filtration is not occurring (study area "B"). 50 Tritium analyses of water from drift and bedrock aquifers showed that Saginaw Formation waters from study area "B" have the highest tritium content relative to bedrock water from study area "A". This suggests faster recharge to the Saginaw Formation in study area "B", supporting the existence of direct recharge in this area. Water from the Saginaw Formation in study area "A" show low levels of tritium. This suggests some water is recharging through shales in the bedrock from the drift. Tritium data show that direct recharge systems are found in the Tri-County area, and that a "point recharge" model is a valid explanation of how water is recharged to the bedrock. In direct recharge areas such as study area "B", shale membrane filtration could not occur. Water composition data of drift and bedrock wells were plotted on a Piper Diagram, and the wells were distributed in a distinct pattern that suggests some sort of chemical change or evolution is taking place. Wells from study areas "A" and "B" plotted in specific, different locations Of the graph, which can be explained by shale membrane filtration of waters recharging the Saginaw Formation. Experimental and field studies have shown that clays and shales behave as semi-permeable membranes, selectively filtering out ions in water flowing through the membrane. Membrane composition is important in determining which ions are filtered the most. Retardation sequences have been developed in previous investigations allowing prediction of whether certain ion ratios would increase or decrease if the water were being filtered. Ion ratios were used in three basic tests to determine if shale membrane filtration was occurring. First, drift water was compared to bedrock water from study area "A"; second, drift water was compared to bedrock water from study area "B"; and finally, bedrock water from study area "A" was compared to bedrock water from study area "B". Ion ratio variations predicted by shale 51 membrane filtration theory are Observed in bedrock water from the shale membrane filtration study area "A", and are not Observed in bedrock water from the direct recharge study area "B". This study's results also compared favorably with previous field investigations. The theory of shale membrane filtration best explains the Observed groundwater chemistry of the Saginaw Formation in study area "A", and accounts for the difference with bedrock water from area "B". Shale membrane filtration does not seem to be occurring in study area "B". It would appear that both recharge processes prOposed to explain the observed water chemistry, direct recharge and shale membrane filtration are occurring in the Tri-County study area. BIBLIOGRAPHY Berry, F. A. F., 1959. Hydrodynamics and geochemistry of the Jurassic and Cretaceous Systems in the San Juan Basin, northwestern New Mexico and southwestern Colorado. Standord University, Ph.D. dissertation. Berry, F. A. F., 1969. Relative factors influencing membrane filtration effects in geologic environments. Chem. Geol. 4, p. 295- 301. Berry, F. A. F. and Hanshaw, B. B., 1960. Geologic evidence suggesting membrane properties Of shales. let International Geologic Congress, Copenhagen, 1960. ‘ Bredehoft, J. D., Blyth, C. R., White, W. A. and Maxey, G. B., 1963. Possible mechanism for concentration of brines in subsurface formations. Bull. Amer. Assoc. Petrol. Geol. 47, p. 257-269. DeSitter, L. V., 1947. Diagenesis of Oil-field brines. Bull. Amer. Assoc. Petrol. Geol., v. 31, p. 2030-2040. Dunk, R., Mostyn, R. A., and Hoare, H. C., 1969. The Determination Of Sulfate by Indirect Atomic Absorption Spectroscopy. Atomic Absorption Newsletter, v. 8, no. 4, J uly-August. Firouzian, A., 1963. Hydrological studies of the Saginaw Formation in the Lansing, Michigan, area. Michigan State Univ., Unpub. M.S. Thesis. Graf, D. L., Meets, W. F., Friedman, 1., and Shimp, N. F., 1966. The origin Of saline formation water, 3, Calcium chloride waters. 111. State Geol. Surv. Circ. 397, 60p. Hanshaw, B. B., 1962. Membrane prOperties of compacted clays. Harvard University, Unpub. Ph.D. dissertation. Hanshaw, B. B. and COplen, T. B., 1973. Ultrafiltration by a compacted clay membrane - 11. Sodium ion exclusion at various ionic strengths. Geochim. Cosmochim. Acta 37, p. 2311-2327. Hanshaw, B. B. and Hill, G. A., 1969. Geochemistry and hydrodynamics of the Parados Basin region, Utah, Colorado, and New Mexico. Chem. Geol. 4, p. 264-294. Hill, G. A., Colburn, W. A., and Knight, J. W., 1961. Reducing oil-finding costs by use of hydrodynamic evaluation. Economics of Petroleum Exploration, Development and PrOperty Evaluation. Prentice-Hall, Englewood Cliffs. Kelly, W. A., 1936. The Pennsylvanian system in Michigan. Michigan Dept. of Conservation, Geol. Surv. Div. Pub. 40, Geol. Ser. 34. 52 53 Kemper, W. D., 1961. Movement of water as affected by free energy and pressure gradients - 11. Experimental and analysis of porous systems in which free energy and pressure gradients act in Opposite directions. 501. Sci. Soc. Amer. Proc. 25, p. 269-265. Kharaka, Y. K., 1971. Simultaneous flow of water and solutes through geologic membranes: experimental and field investigations. Univ. of California, Unpub. Ph.D. dissertation. Kharaka, Y. K. and Berry, F. A. F., 1973. Simultaneous flow of water and solutes through geologic membranes, 1. Experimental Investigations. Geochim. Cosmochim. Acta 37, p. 2577-2603. Kharaka, Y. K. and Berry, F. A. F., 1974. The influence of geologic membranes on the geochemistry of subsurface waters from Miocene sediments at Kettleman North Dome in California. Water Resources Reserach, v. 10, no. 2. Kharaka, Y. K. and Smalley, W. C., 1976. Flow of water and solutes through compacted clays. Bull. Amer. Assoc. Petrol. Geol., v. 60, p. 973-980. Korzhinskii, D. S., 1947. "Filtration effect" in solutions and its role in geology. Bull. Akad. Sci. U.S.S.R., Geol. Ser., 2:35-48. Larson, G., 1979. Study for Tri-County Regional Planning Commission. Long, D. T. and Angino, E. E., 1977. Chemical speciation Of Cd, Cu, Pb, and Zn in mixed freshwater, seawater, and brine solutions. Geochim. Cosmochim. Acta 41, p. 1183-1191. McKelvey, J. G. and Milne, I. H., 1962. The flow of salt through compacted clay. Clay Minerals 9, p. 248-259. Mackay, R. A., 1946. The control of impounding structures on ore deposition. ECO“. G801” #1, p. 13-u6o Mencenberg, F. E., 1963. Ground-water geology of the Saginaw group in the Lansing, Michigan, area. Michigan State Univ., Unpub. M.S. Thesis. Meyer, K. H. and Sievers, J. F., 1936. Helv. Chim. Acta 19, 649, 665, 987. Milne, I. H., McKelvey, J. G., and Trump, R. P., 1964. Semi-permeability of bentonite membranes to brines. Bull. Amer. Assoc. Petrol. Geol. 48, p. 103-105. Piper, A. M., 1944. A graphic procedure in the geochemical interpretation of water-analysis. Transactions, Amer. GeOphysical Union - Hydrology papers--l944. Radfar, S., 1979. Determination of recharge areas from groundwter quality data, Ingham County, Michigan. Michigan State Univ., Unpub. M.S. Thesis. Russell, W. L., 1933. Subsurface concentration of chloride brines. Bull. Am. Assoc. Petrol. Geologists, 17, p. 1213- 1228. 54 Ritter, M. H., 1980. The use of tritium for confirming areas of groundwater recharge, meridian Township, Michigan. Michigan State Univ., Unpub. M.S. Thesis. Schlumberger, C., Schlumberger, M., and Leonarden, E., 1933. Electrical coring: a method of determining bottom-hole data by electrical measurements. Trans. AOIOMOEO, 110, p. 237-2720 Stuart, W. T., 1945. Ground-water resources of the Lansing area, Michigan. Michigan Geol. Survey Prog. Report 13. Toerell, T., 1935. Proc. Soc. Exptl. Biol. Med., v. 33, p. 282. Truesdell, A. H. and Jones, B. F., 1973. Wateq: A computer program for calculating chemical equilibria of natural waters. U.S.G.S. Pb220-464, 73p. VanEverdingen, R. O., 1968. "Mobility of pair ion Speciation reverse osmosis and the modification of subsurface brines. Canadian Journal of Earth Sciences, v. 5, no. 5, October. Vanlier, K. E., 1964. Groundwater in the Tri-County region, Michigan, Chapter 1 of Natural resources problem study: Lansing, MI, Tri-County Reginal Planning Comm., p. 122. Vanlier, K. E., Wood, W. W. and Brunett, J. O., 1973. Water-supply deveIOpment and management alternative for Clinton, Eaton, and Ingham Counties, Michigan. U.S.G.S Water-Supply Paper #1969. Vanlier, K. E. and Wheeler, M. L., 1968. Ground-water potential of the Saginaw Formation in the Lansing MetrOpOlitan area, Michigan. Tri-County Regional Planning Commission, Lansing, Michigan. Walton, H. F., 1958. Principles of osmosis applicable to oil hydrology. Unpub. research report, Petrol. Reserach Corp., Denver, CO, 66p. Wheeler, M. L., 1967. Electric analog model study of the hydrology of the Saginaw Formation in the Lansing, Michigan, area. Michigan State Univ. Unpub. M.S. Thesis. White, D. E., 1957. Magmatic, connate and metamorphic waters. Bull. Geol. SOC. Ame, 68, p. 1657-1706. Wood, W. W., 1969. Geochemistry Of groundwater of the Saginaw Formation in the upper Grand River basin, Michigan. Michigan State Univ. Geology Dept., Unpub. Ph.D. dissertation. Wood, W. W., 1976. A hypothesis of ion filtration in a potable water aquifer system. Groundwater, v. 14, no. 4, p. 233-244. Wyllie, M. R. J ., 1948. Some electrochemical prOperties Of shales. Science 108, p. 684-685. 55 Young, A. and Low, P. F., 1965. Osmosis in argillaceous rocks. Bull. Amer. Assoc. Petrol. Geol. 49, p. 1004- 1007. APPENDIX 56 Tritium Analysis Ideally, a tracer used for groundwater studies should accurately describe water flow through the groundwater system under investigation for the duration of the study. This requires that the presence of a tracer in the water does not change the prOperties of groundwater or the transmission characteristics of the medium through which the flow is being traced. In particular, hydrogen isotopes serve as excellent tracers because they are incorporated directly into the water molecule, and at concentrations most often used do not change the water prOperties. Since tritium is also a naturally occurring isotope, artificial introduction of the tracer is not necessary. In addition, tritium is radioactive and can be detected at very small concentrations. It has a half-life Of 12.35 years, and therefore is particularly useful to date recent water movement in recharge areas. Tritium is continuously being produced in the upper atmosPhere by interaction Of cosmic radiation with nitrogen atoms. The reaction is: lq’N4-no + 31"I-I-12C This cosmicly produced tritium eventually reaches the earth's surface through precipitation. Early 1950's and 1960's thermonuclear bomb tests in the atmOSphere also injected large quantities Of artificially produced tritium into the atmOSphere, and hence into the hydrologic cycle. This Spike Of artificial bomb tritium from thermonuclear events supplies a means to "date" the relative age of recharging groundwater from its tritium content. Analysis for tritium is based on detection of the Beta radiation that accompanies radioactive decay of tritium (3 H) to helium (3 He). There are two types Of Beta decay: negatron and positron. Tritium Beta decay involves negatron (electron) decay. The decay scheme is: 3H + 3He + e- 57 neutron + p+ + e' + v e A tritium neutron decays into a proton, electron, and energy. The energy (Beta radiation) produced by this type of tritium decay can be emitted over a continuous range from essentially 0.0 MeV to 0.019 MeV. Tritium Analysis Procedure The most common type of radiation detection device requires radiation to pass through a window of a detection device. Once inside, the radiation ionizes certain gases in the chamber, generating an electric current which in turn can be detected and counted. However, since tritium is a very low energy Beta emitter, all or most of the Beta particles would be absorbed by the window of such a detector. Also, the Beta particles have small mass and are easily detected, particularly low energy Beta particles. Because of the inherent difficulty in detecting weak Beta radiation, special techniques were developed to measure soft Beta raidation. This method utilizes liquid scintillation, a technique which involves mixing the radioactive sample in a detecting medium such as a fluorescent organic solution. The radiation from tritium decay causes the organic solution to emit light, which can be converted to photoelectrons in a photomultiplier tube, and measured as an electric pulse. In theory, this method assumes that each tritium disintegration produces a pulse of light which can be more easily detected and counted than the radiation itself, and therefore it is much more sensitive in the detection of low energy Beta radiation than other methods. The liquid scintillation counter is set up to count the number of light pulses, and thus the number of disintegrations. The maximum efficiency using the counter available for this study is approximately 60% Of all decays detected. The procedure for sample preparation prior to liquid scintillation counting is relatively simple. Care must be used at all stages not to contaminate the 58 samples, as even atmOSpheric moisture could add tritium to a sample. Samples of well water, precipitation, and river water were collected in 500 ml prewashed plastic bottles and labeled. Each water sample was then distilled to remove suspended solids and dissolved minerals. Next an electrolysis process resulting in isotropic fractionation, facilitated by the addition Of an electrolyte after distillation, is used to concentrate the naturally occurring tritium to levels that are more easily detected. After electrolysis, the water samples are distilled out of the electrolysis cells for two reasons. First, the post-distillation removes the electrolyte added beforehand. Second, the water sample is removed from the cumbersome electrolysis cell into a smaller weighing bulb to ease the measurement of the final volume Of sample left after electrolysis. After post- distillation, a measured amount of water is placed in a boro—silicate scintillation vial, mixed with a fixed amount Of scintillation fluid, and measured in a liquid scintillation counter when a sufficient number Of samples have been processed. The Appendix contains a more detailed analysis procedure. Tritium Lab Set-up The four basic steps of tritium analysis are: l) predistillation, 2) electrolysis, 3) post-distillation, and 4) liquid scintillation counting. Pre-distillation consists of a three unit glass vacuum system mounted on an aluminum lattice rack, with boiling flask, Kjeldahl bulb, condenserto, and receiving flask. The system uses cooling water, vacuum, electric heating elements, and dried air. The electrolysis system consists of a low temperature chamber, thermostat, Ostlund electrolysis cells, and adjustable power supply, and a vacuum exhaust line. The low temperature chamber is an Open tOp freezer containing water and anti-freeze, a circulation pump, and a plexiglass rack to hold the electrolysis cells. A vacuum switch was installed on the power line to the electrolysis cells to authomatically cut off power in the event of failure in 59 the vacuum exhaust lines, minimizing the danger of explosion due to a concentration of explosion due to a concentration of explosive gases in the cells. The post-distillation is a three unit glass system mounted on an aluminum lattaice rack. Electric heating elements, liquid nitrogen, vacuum, and dried air are used in this system. It is designed to distill the water sample out of an Ostlund electrolysis cell into a weighing bulb. This study used a Beckman LS 8100 liquid scintillation counter. This as an ambient temperature, soft beta counting spectrometer provided with digital readout and printout for data recording. This unit automatically counts to a preset error or time, and calculates counts per minute. Laboratory Procedures Prior to distillation, all units are dried under vacuum before adding a water sample to the boiling flask, and each cooling water circulation is checked for leakage. When dry, the water sample is added to the boiling and immediately placed under vacuum. After the sample stops degassing, the vacuum is turned Off and heat is applied to the flask. At this time, dry air can be readmitted to the boiling flask to avoid excess bumping and to reduce the danger of implosion, however, distillation is generally more rapid under vacuum. Each distillation unit is normally connected to a baloon to indicate pressure within. If during distillation a balloon indicates excess pressure, momentary vacuum is applied to a unit or heat reduced. Just before dryness, the heat is removed from the boiling flask to allow any water in the Kjeldahl bulb to drain, and flame is applied to the bulb and connecting glass of the condenser to thoroughly dry them. Heat is then reapplied to the flask, and distillation taken to dryness, with a heating of several minutes past dryness to dehydrolyze salts. In a glass vacuum system, there is danger Of implosion, therefore, it is recommended that a face shield be used at all times. 60 In the electrolysis step, normally two volumes are used for enrichment, 100 ml and 500 m1, although this is not a restriction. The electrolyte used is NaOH in pellett form added to the electrolysis cells with the water sample and allowed to dissolve before electrolysis begins. Up to 10 cells are placed in the cooling bath where a temperature is maintained at 10 or 2°C. Vacuum exhaust lines to remove hydrogen and oxygen are then connected to the electrolysis cells which are connected in series to a variable power supply. The initial current applied to the cells is 6 amps, but is gradually reduced as sample volume decreases. The final current is 0.5 to 1 amp to reduce the danger of arcing in the cell. At times an intial current Of 4.0 amps can be used to improve the electrolysis efficiency slightly, and to produce a more consistent final volume of sample left in the cells. The electrolysis efficiency (how much tritium in the original volume of sample is concentrated in the final volume) using an initial current Of 6.0 amps is about 74%, meaning that 74% of the original tritium in a sample is concentrated in the final volume left after electrolysis. This efficiency is determined by running a standard Of known activity with each set of samples. A blank, or sample containing no tritium, is also run with each set of samples to provide the background radiation count. The next step, post-distillation, commences when the electrolysis step is completed. Each electrolysis cell is removed from the cooling bath, and a distillation head is placed on top of the cell effectively sealing it from the atmosphere. The cells then are immediately placed on the post-distillation system, put under vacuum, and heat is applied. The water from each cell is condensed in a separate weighing bulb cooled by liquid nitrogen. When distillation is complete, the weighing bulb is weighed to determine the final volume of water left after the electrolysis step. A measured amount, as close to 5.0 grams as possible in this study is then placed in a scintillation vial. "Instagel" 61 scintillation fluid is added, 6 ml to each vial, after which the vials are store in a dark area until ready for counting. Measuring the activity of the samples is the last step. The Beckman scintillation counter used in this study is equipped with pre-programmed "Library Programs for various isotopes". Library program #5 with modifications was used for this study, a program set to count tritium. The Beckman counter has two adjustable channels or detectors. Channel 1 was "Optimized to reduce the background noise or radiation to approximately 6 or 7 counts per minute instead of 18 to 20 counts per minute with the channel wide Open. On the Beckman scale, channel 1 was set at 125 for the lower end, and 225 for the upper limit of the detection window. Channel two was left wide Open, from 0 to 397, for comparison, and encompasses the entire tritium decay energy range. The amount of quenching in each sample was determined separately using a built in feature Of the Beckman liquid scintillation counter that calculates an "H" number. The "H" number, which is a function of the quenching within a sample, is related to the liquid scintillation counting efficiency using prepared standards containing variable amounts of quenching agent. As a result, counting efficiency for each sample can be calculated. The H number is a method Of quench monitoring that is done automatically by the Beckman LS counter. A graph Of counting efficiency versus H number was developed using a set Of standards with varying quench. How much a sample is quenched determines how efficiently the LS counter can detect the bursts of light produced th bye radiations interaction with the scintillation fluid. Since all samples cannot be produced exactly alike, the quench varies slightly, and therefore the counting efficiency varies. An H number is calculated for each sample by the LS counter, and this number is converted to counting efficiency by linear interpolation of the nearest two bracketing points on the graph Of H 62 number versus counting efficiency. The counting efficiency for the samples was usually between 13% and 14% with channel 1 set at 125 and 225. In each set of samples there was a counting standard used to determine electrolysis efficiency, plus a blank used to indicate the background radiation. Both standard and blank went through the same electrolysis conditions (electrolyzed in same set as samples) and were counted on the LS counter with each sample set. Thus, for each sample, the counting efficiency of the LS counter and the electrolysis efficiency are calculated. Finally the actual activity of the sample can be determined from the equations given in the following "calculation" section of the appendix. Once the activity is expressed in dpm/gm, it is converted to tritium units (T.U.) by the conversion factor Of 1 TU = 0.0071 (1me gm. The results in this study are given in TU's. Tritium Calculations In order to calculate the final activity of a sample, the electrolysis efficiency must be calculated using a standard Of known activity. The electrolysis efficiency Of the standard is calculated as follows: V A EEes = [ ill es) (vol (A 08,) final volume Of electrolysis standard f Vo = initial volume of electrolysis standard Aes = final activity of electrolysis standard Aoes = initial activity Of electrolysis standard Initial and final volume refer to the amount of standard before electrolysis and the amount remaining after electrolysis. The initial activity of the standard of 63 course takes into account decay that has taken place since the standard was prepared. The final activity of the electrolysis standard (Aes) is calculated as follows: A = cmp - bkgd cpm/ mass es . . effICIency Cpm = counte per minute bkgd cmp .-. badkground counts per minute mass = mass of electrolysis standard counted efficiency - counting efficiency (from H number) Once the electrolysis efficiency of the standard is known, it can be applied to calculate the electrolysis efficiency of each sample. The comparison between the standard and the samples is done using what is called the Electrolysis Fractionation Factor "Beta" (B). The electrolysis efficiency of each sample may vary because during electrolysis gases bubbling through the samples carry away water vapor and Spray. Although the electrolysis cells are connected in series, slight differences in vapor loss will occur due to non-uniform temperatures in the water bath, slight variation in amount Of electrolyte added, and differences between electrodes (area, spacing, etc.). Therefore, each sample has a slightly different final volume, and therefore a slightly different electrolysis efficiency from the standard. The calculation of the electrolysis efficiency for each sample is made using the Electrolysis Fractionation FA ctor "Beta" (B). B = 1n wO/vf)es -ln EE es (V oIVf) es = initial and final volume Of electrolysis standard EE es = electrolysis efficiency Of electrolysis standard Once "Beta" (B) is calculated from the electrolysis standard, the electrolysis efficiency of each sample can be calculated from: 64 EES = (vo/vffl/B EEs = electrolysis efficiency Of sample V o = initial volume of sample Vf : final volume Of sample B = Electrolysis Fractionation Factor When the electrolysis efficiency for each sample is known, then the activity Of each sample can be determined by: dme gm = cmp - bkgd cpm/ mass (EESXVo/Vchounting eff.) (1me gm = disintegrations per minute per gram cmp = counts per minute of the sample bkgd cmp = background counts per minute mass = mass of sample counted EEs = electrolysis efficiency of sample V o/Vf = samples initial volume/samples final volume counting eff. = efficiency of LS counting (from H number) The average electrolysis efficiency Of the standards during the study was 74.25%, meaning 74.25% Of the tritium originally in the sample remained after electrolysis. The average Electrolysis Fractionation Factor "Beta" was 9.27, and the average value for background radiation was 6.68 counts per minute. Error for the samples were calculated from the following equation Cpmsm " CPmb cpm counts per minute Of sample + background s+b cpmb = counts per minute of background 2 s+b96 = 2 sigma % error of sample + background 2 b% = 2 sigma % error of background 65 Cpm, or counts per minute, and 2 sigma % error for both sample and background are automatically calculated and recorded by the LS counter and are read directly from the printout. TABLE A.* Well Run 1 Run 2 Run 3 Mean Standard Deviation Avery 66.60 57.10 58.90 60.86 5.046 Herriff 69.99 69.1 75.76 71.61 3.625 *Values given in Tritium Units (T.U.) TABLE 2 “In 0 o h IA “In 0 o 00 O-‘ONOOOOOOOO 00—1006 66 Inn oa- xonnNmasnInc-nxoooa- d-OOMNéd- O O I\l\ [\l\l\l\l\l\l\l\l\l\l\l\ l\l\l\l\l\f\ NO axnnm—norxoa-woo \D'fi—‘Nwifi emn o N6 N.N Re N no 00033823 cum 0 n .N a; N.N ON mu UUUomBnozmo man 0 0.: N 4 N .m ”N om UUmmBmozNo can ES RN m6 N.N aN mm 00233ch man o w; o.N m6 N n» QQ823 Domm n4 o4 3: 52 N am Om41 O C—‘ O I O O I O O O O O O O Nil'wMN -I oo—t ar—I N . . . . . C . . . C . C . \DMNNNanwdwoowwmnééa'm-dmn nooonoooooooooooooo—a—Hnooo OOWOO O :0 2 62.5269 Ewe; mime EEO Siam N N NN om nm N NN on an n N am Nm ON ON me me N— we 3 nm N Nm ”N d N 3 an ”N £1 2: mm MO 0003 mNo Zao UU§oZmo UOQNNBSZNO QQ<<¢NBSZNO DQ15 Zmo <<<§oZmo mfi~oZno DUO/ENBaoZno 15 Zn o mmUmuN BmoZno mm2moZNo :oflmooA 69 NM“ Mlfi'filfi O (\NNNNNOONN NWNNNNNNNNNNNNNNNNNNN UK s o o O 0 —¢ n WWWd‘d'MOOON NOOOW¢$°OOON§fiNMM$¢NONW¢N \DINO\ O O O O O O O 30:] N *m '\ \O 00““ 0000000000000. l\ H—a-d- OOOOOOO In (\NMNUNfl'v-‘MO MNOWOWOOOOWMMOOO‘NNDM—‘MN—‘N O COO—100000 dONCOOOOOOO—‘O—‘fiocc—‘fio 0‘ d) 1.1.. wen me: no: 33 RN: wan own man a d a cum wnm sum nmN aNm em: mNm wNN own omm NNN o g a cum own new 0mm mam N3 eon 3n noa 001 we nNd an mm a N— on N 3 we Om “O‘COONO‘G O O O O O N—t—tN—‘6Mhfi' HOMNw'flnoooo—‘é—‘NN—‘N—‘O-‘fi Nit “0000080!“ 0-4 O O N OOVNNNOO—‘OOO 6 O O o o o o v-lNNNNfi'N—‘d Nwokwoo—‘M OOWNN—i—Ioo—‘OOOé'V‘NNONNoowcn O 2| oooooooo HO‘ c—q “\OMNfiMOOthWHONOOOO‘nh “\OMV‘NDMOWNing‘V‘d'NNNONNO‘ O O O O O N -'N u—iI-II‘ SN 0 h 2 fl Em on N mfl €852.69 :EEV mime EEO cots; “N on N g 2 on on ”N ”N NN N 2 N N NN MN 3 mm Nm Nm in on N on El mm cm on 9 em on om nm 02 en ow ow en mm on“ an we ow on 3 NO on as NO 0» m0 Q<fioZNo 0015 Zno mmmmi BMQZNO <§o Zno cofimooq 7O u-«h —t O O O O O O O O O O O O O O v—t In 0 o o o o o m “\OMNd'[\MNNOOMWd-O‘éerhW—HAUAWNIANOOth-d- O O O O O O O O O O Noahlxlx[\NNNNNNNNNNNNNNNNNNNNNNNNN HMNONWWMW ‘51 “00W O O O O O O O O O O O O O O 00 O O O O O O O [\h Nfi' [\Mh N—I Infid'[\d'd'mwKOO-'\D\DO\—'O\\DI\N@O\I\&O#O‘MN[\éoo O O O O O O O IOONOOOOOCO—‘OC-fiéOOONCOOONONO-‘OCC 0 IL nwm own cum 2n on No: man can Non NmN can cum awN ram Nwm awn wwm Nme amm nan enm a; :3 One nwm omm mNe mum Cum mun 00$ on hm o2 no" 3 ca n m an N N w; mm ma mm m— 2 n— mN 3 mm I-‘ n 0 —1 0 III! O —c—oN-cNo—4—¢—Hn|f\'r\oo—d\oonué—d—I—Hfia-MO—‘ONN ON“ 0 O I-d o—IO‘NOOOOO'A—t—‘O‘NO‘NO 0"‘1 O Z3 \Dd’\O\DOOOW\Dd'°O\D—INhd‘O‘NNDO‘Na‘Md'OMOWMNN O -‘[\-'NN§'\DW—‘fiOOMNOOOOO—‘OOhhéf-dON—t—t-flN—fi ¥| 0‘ III! Md'VNMOO—dNV‘MMO‘fi'NDONBNM o—( 111 2123 66:50:68 395 SE3 EEO c9138 Nn «— N N me on 3 N m— on on N «N 2 MN 3 on on on ENRRE on ”N £1 O$oZ¢o <000Nm32ao <<00m fl mfioZaOo fioZmo (£00333ch <$och 0000mNB—oZmo 32MM MS MM N M.M M.M NM NM NmaoMN332Mo 33 MM MN N._ NM MM M: 3833328 NMM MN N M.N N.M MN 3 80033323 N? N N. _ N ._ M .M MN MM MEMMMBMono M3 MM N.M N.M M.M NM MM noquMBMozNo MMM M M.M M .N M.M MM 2: 8583323 3 MM N.M N.M M NM MM MouMNBMozNo 82 CM 6 2 M2 34 Mo :onqu €263.89 ACMEV mqqm3 hadn— cotamqm TABLE 7 73 own.n ~N0.w 55:.0 0N0.w mmn.» wum.w Nun.n mnm.n :m:.m ::m.0 mnw.n 0w:.w mum.w ”00.n Hfiw.n 00N.w 00:.» ”00.0 wnw.: n0n.w N00.m nnw.w uNn.0 ::0.- finn.n 00m.u MMM.N 1cm. .tao w—u.n :—~.w now.0 «03.0 Nnu.w mw0.w 00w.n mmm.~ :mw.o :Nn.w 00N.: n:m.0 0:u.w awm.n 0nw.n mm:.m www.w Nm~.0 20m.: 2~N.n mmn.0fi 000.n nnm.m nun.0~ :0n.0 Nun.m wmm.n MfiMw .bau ~.wn: m.~w: 0.:Nn n.00: u.w~n 0.w0n :.n0: N.~Nn 0.:2n n.u~: 0.n2m :.-n m.:nm m.:n: 0.m0n :.~0w H.0wm :.~um w.n0m 0.2mm m.m:n 0.50s n.00n 0.20: n.~:w w.an Mina mnrr gmdo. memo. ~0n0. Nmmo. N~N0. :n:0. 00N0. :::0. Nude. Nwmo. deo. mono. :wnfi. mung. fimmo. wNwN. :mfiu. n_~0. wmmo. now». anofi. mwnfi. 0~NN.~ mono. wnwm. :0u.0 mU>U MMM.M MMM.M MMM.M MN+mM.M NMM.NM MNN.M MMM.M MN+mo.N NMM.MM MMM.M MMM.M MN+mM.2 NMM.M~ MMM.M MMM.M MMMM.M MMM.M NMN.M MMM.M MN+mM.M MMM.M NNo.M MMM.M MN+mM.M MMM.N MMM.M MMM.N MMMM.o MMN.MM MMM.N NMM.~ MN+mM.M MMM.N MMM.M MMM.M MMMM.M MMM.M MMM.M MMM._ MMMM.M MNM.N MMM.M MMM.M MMMM.N MMM.MM MMM.M MMM._ NMMMM. MMN.o_ MMM.M MMM.M MMMN.M MMM.M MN2.M NMM.~ NNMM.M _MM.M MMM.M MMM.M MMMM.N MMM.MM MMM.M MMM.M MMMM.2 MMM.M MNM._M MMM.M NMMM.M MMM.M MMM.MM MMM.N MN+mM.2 NNo.M MMM.M NMo.M NMNo.o 8232 3m? 2002 81% MMM.N MMM.MM MMM.M MMNM.NM 23M.MM MMM.M MMM.M MMMM.M MMM.M MMM.NM MMM.M MMMM.M NMM.M MMM.NM MN_.M MMoN.M NMM.MM MMM._ MMM.N MMoM.o MMM.N MMM.NN MMM.M MMMM.M MMM.M MNM.M MMM.M NNMN.M «230 232 M38 M0&6 M0250 M85 mm 83mm 3."; EEO SEEM 0:m00. Nwafio. mmuuo. 0NN~0. 00u00. nn:~0. wn000. :m:_0. 0::00. ~0n00. wwmoo. 00020. nwnno. awmno. n0-0. wowod. nm:0m. me00. 00000. 00~0m. mnnmo. n:00~. 002mm. 0~NNO. ~mww0. Nm:N0. 00000. 00000. 00000. 0Nnm0. 00000. 00000. 0N020. 00000. m:N00. «RN00. wndoo. 000m0. nmnfio. w:nm~. :0200. ofinno. «5000. 00000. aw0:m. anmo. num00. nuwdm. wnN:0. :::n0. mnmfio. wmowd. DUU:m 3m02m 0 UUUom 3m02m 0 000m 3m02N0 DUE 3m02~0 ODU 0::.0— 000.0 000.0_ :5:.5 0::.0 000.— ~05.0 N~0.N 005.0 N05.0~ —5_.N~ 00:.02 500.N 050.0d 000.— 550.0 —:0.~ 00:.5 000.0 0::.N 505.: 000.0 000.N 020.0 050.: 020.0 0N0.0 N00.0 N0:.N 000.5 mZ?U N0~.0 0~:.0N 000.0 :00.0 :N0.0 :50.0~ N00.0 :05.0~ :00.00 N00.0 5:0.02 200.: 000.02 000.: 00:.:N 000.- 000.5— 000.02 005.50 000.5N :00.0 0-.0~ 0~—.0~ 500.0~ 050.:~ N~0.0 000.0 000.: 050.0N 000.5 MHZ 005.~ 00:.~ :00.~ ~50.N 000.~ 050.N 000.2 0~0.~ :0N.N 050.2 0N0.N 0N0.N 0~N.~ 000.N 00:.~ 0:5.~ 000.2 000.2 005.~ 00:.— 200.0 00:.— 0~0.~ 000.~ 000.~ 005.~ 0_0.N ~00.N 200.~ 2:0.— 0.230 0020.0 5000.0 0000.0 NNON.0 0000.0 5020.0 N~00.0 :NN0.0 NN:0.N 050:.0 000_.0 000:.0 0005.0 :00:.0 00~N._ 50~N.0 :0:_.0 :00:.: 0000.0 0020.0 00:N.0 ::50.0 002:.0 N000.N 5505.0 ::00.0 000N.0 :0:fi.0 5000.— 000N.0 M85 83mm 4.53 HEMQ A000C ..0:cm> 0 MoM_o. MNMMM. oMcoo. NoM_o. MMNoo. MMMMM. oMoMo. MNoMN. NMMN_. MMMMM. NoMMM. M_MMM. MNMNo. MMMNM. MMNMM. MMMoo. MMNNM. MMMMM. NMoMo. MMMMM. M_MNM. MMM.M. MMNMM. MMNMM. MMMMM. MNMMM. MMMMM. MMNMN. MMNN_. MMooM. MMMMM. MMMMM. MMMMM. MMMMM. MMMMM. MNMMM. _NMMM. MNMNM. MNMMM. MNMMM. NNMoo. MNNNM. NMMNM. MMNMM. MMMoo. NMNNo. MMMMM. NNMNM. M_NMM. _MMMM. MMMNo. MMMMo. MMNMM. MNoMo. _N_Mo. MMMNM. NMMMN. MMMMM. MMMMM. MMMMM. 82:0 Mouaemm Dan—N0 30 02N0 ODDS mN02:0 UUU 000.: 005.: 00:.5 050.0 00:.0 000.00 ::0.0 000.0 000.0 :0:.0 000.:0 00:.0 000.00 000.5 000.0 000.0 000.0 500.0 000.0 000.0 055.0. 0:0.0 550.0 000.0 000.0 000.:0 005.0 0:0.0 005.0 mZWU 0:5.00 050.0 :00.0 050.5 500.0 055.0 000.0 500.0 :0:.: ::0.5 000.: 000.00 000.0 500.0 000.0 000.0 000.5 000.5 000.0 500.0 0::.0 050.0 000.0 0:0.0 500.:0 0:0.5 000.00 000.00 :00.0 MHZ 00:.0 000.0 000.0 :00.0 000.0 :00.0 000.0 000.0 0:0.0 005.0 000.0 000.0 500.0 500.0 ::0.0 000.0 000.0 000.0 000.0 000.0 005.0 000.0 000.0 000.0 000.0 000.0 50:.0 050.0 505.0 005.0 002me 0:55.0 0000.0 500:.0 00:0.0 5:00.0 0000.0 500:0. 0005.0 0000.0 0000.0 00::.0 0000.0 0000.0 5000.0 :050.0 0000.0 0000.0 0000.0 :00:.0 0000.0 00:0.0 00::.0 0000.0 0000.0 0000.0 0005.0 0000.0 :000.0 000:.0 0000.0 38:0 00000. 00000. 00500. 00:00. 00000. 00000. 00500. 00000. :0:00. 50:00. 00:00. 00000. 00000. 05000. 00500. 0:000. 00500. 0:000. 00000. 00000. 00000. 50500. :0000. 00000. 00000. 00000. 00000. :0000. 00000. 0:000. 00500. 00000. 00000. 00500. 00::0. 00000. 00:00. 05000. 00:00. 00000. 00000. 00000. 05000. :0000. 00::0. 00000. 00000. 0:000. 00000. 5:000. 00000. 00000. 50:00. 00000. :5000. 00000. 0::00. 50:00. 00050. 050:0. 835380 .8an 4.53 EEG 83: 625> 8:5 M03: .00 0<00003:0200 000<003:020 0 (00000 3:020 0 <00<00 3:020 0 000000 3:020 0 000050300200 00000 30020 0 00000300200 <<0000 300200 00<<00 30020 0 00000 0 30 020 0 000000300200 000000 30020 0 <<<00 0 30020 0 00000 30020 0 0<0<00 30020 0 0<000030020 0 (00000 300 20 0 0<0050 30020 0 000050300200 000000300200 (00000 30020 0 00000 30020 0 <<<<00 30020 0 00000 30 020 0 00(000300 2:0 00:0 300 20 0 <<000 300 20 0 0<00030 020 0 000000 3:0200 20058.0 76 0:0.0 000.5 000.0 000.5 :::.5 :50.0 000.5 005.0 000.5 000.5 000.0 0:0.0 000.: 000.0 000.5 000.0 005.0 :00.5 050.0 0::.: 005.5 000.0 00:.0 000.00 050.0 :00.0 ::5.5 00:.0 000.0 000.5 .tao 0.00: 0.000 0.000 0.000 0.000 0.000 0.000 0.0:0 :.000 0.000 0.00: 0.000 5.000 0.00: 0.500 0.5:: 0.00: 0.5:0 0.000 0.000 :.500 0.000 :.000 0.000 0.000 0.000 0.000 0.000 0.00: 0.050 0000 0000. 0000. 0000. 00:0. 5000. 0000. 0::0. 0000. 0000. 0000. 0000. 0000. 0500. 0000. 00:0. 0000. 0050. 0500. 0000. 0000. 0000. 0000. 5000. 0:::.0 0:00. 0000. 0000. 0000. 0500. 0000. 0000 :00.0 000.0 000.0 000.0 000.0 000.0 00:.0 000.00 500.0 0:0.00 500.0 050.0 000.: 005.0 00:.0 000.00 000.5 000.00 :55.:0 500.0 05:.00 :00.0 000.0 0:0.0 000.0 005.0 0:0.0 000.0 000.0 5:0.5 0230 055.00 000.0 500.0 000.00 500.00 00:.0 :00.0 000.0 005.5 0:0.0 000.5 000.5 000.5 550.0 000.0 050.0 00:.0 050.0 000.0 0::.0 000.00 005.00 000.50 000.50 500.00 :00.0 000.00 500.0 000.00 000.0 dx\070 550.0 000.0 500.0 050.0 000.0 500.0 000.0 0:0.0 000.0 000.0 00:.0 505.0 000.0 000.0 000.0 500.0 0:0.0 00:.0 000.0 000.0 500.0 000.0 000.0 000.0 000.0 000.0 0:0.0 000.0 000.0 005.0 0230 0:50.0 0500.0 0000.0 :050.0 5000.0 0000.0 0005.0 00:0.0 :050.0 0000.0 00+00.0 5000.0 0000.0 0000.0 0005.0 0000.0 0000.0 00:0.0 0500.0 000:.0 :50:.0 0005.0 0000.0 0000.0 0050.0 0000.0 0050.0 0000.0 000:.0 0:00.0 3030 0 0:000. 00000. 00000. 50000. 00000. 0:000. 00:00. 0:000. 00500. 00000. 00000. 00000. 00000. 05:00. 00000. 00000. 000:0. 00000. :0000. 00000. 00000. 00000. :0000. 50050. 000:0. 05500. 00000. :0000. 00000. 00000. 00000. 00000. 0:000. 00500. 00000. :0000. 00000. 5:000. 00000. 00050. 0000. 00000. 00:00. 00000. 00000. 00000. 50000. 00000. 00000. 00500. 50000. 05000. 00000. 00500. 0:500. 00000. 00000. 00:00. 00000. :0050. 000:0 Mo0E .0 6852000 8:30 .303 F.,—Ea AMMM: b=5> 0030000320 0 00000 00020 0 00000000020 0 0<0<0000020 0 <<<<00m00020 0 <0<<00 300200 000000000200 000000300200 000000 300200 000000 300200 000000300200 000<00300200 00<000000200 <00000000200 000000 000200 00<000300200 00<00 000200 00<<00300200 000050 300200 (0300003323 000000000200 000000300200 000000300200 00<00 0 3:0250 00003 300250 <<000 0 300250 00<00 0 3:0200 <<<000300200 000<:0300200 000<:0 3:020 0 838°.— 77 :0:.0 0:0.0 000.0 000.0 000.0 000.5 00:.0 0:0.0 0:5.5 :00.0 000.5 005.0 000.0 000.0 000.0 000.0 550.0 000.0 500.5 000.5 000.0 000.5 :00.5 500.5 000.0 000.0 500.5 000.0 000.0 .tav 00:.0 :00.0 000.0 050.0 00:.0 005.0 000.0 500.0 000.0 050.0 000.0 005.0 000.0 000.0 00:.0 :00.0 000.5 00:.0 000.5 500.0 :00.5 00:.5 :05.0 000.0 000.0 50:.0 000.5 000.0 0:0.0 .wmu .tam 0.000 0.00: 0.00: 0.00: 0.005 :.000 0.050 0.00: 0.5:0 0.:0: 5.000 5.5:0 0.00: 0.00: 0.050 0.000 5.:00 0.000 0.000 0.000 0.000 0.000 5.000 0.:00 0.000 0.000 0.000 0.050 0.:00 0.00: 0000 5000. 0000. 0500. 0000. :000. 0000. 50:0. 0000. :000. 0000. 00:0. 0000. 0000. 0000. 0000. 0000. :000. 0000. 0:50. :000. 0000. 0000. 0000. 0:00. 0000. 0000. 0000. 0:00. 0000. 5000. M000 000.0 000.0 0::.: 000.0 005.0 000.0 500.5 000.0 000.0 ::0.0 0:0.00 500.50 000.0 500.50 000.0 :00.5 000.0 50:.0 0::.00 000.00 005.00 000.0 000.: 050.50 000.0 000.0 00:.0 0:0.00 000.0 000.: «.2me 500.00 0:0.0 :00.0 005.00 :50.0 550.00 000.: 000.0 00:.0 5:0.0 000.0 000.0 :05.: 000.00 000.0 005.0 500.: :00.0 050.0 :50.0 050.5 0:0.0 000.0 00:.0 500.0 000.: 000.0 0:0.:0 000.00 V032 505.0 :00.0 000.0 :00.0 000.0 500.0 000.0 000.0 :00.0 000.0 000.0 005.0 550.0 000.0 005.0 000.0 500.0 000.0 005.0 0:0.0 000.0 005.0 000.0 000.0 550.0 000.0 055.0 0:0.0 055.0 000.0 MEWU 0:00.0 0000.0 0000.0 0000.0 0050.0 00:0.0 0050.0 0000.0 0000.0 0550.0 0500.0 0500.0 0000.0 0000.0 0:05.0 0000.0 0500.0 000:.0 0000.0 0000.0 0000.0 0000.0 :000.0 0000.0 0000.0 0050.0 5000.0 00:0.0 500:.0 00:0.0 .0006 685058 8:30 in; 0.55 8*: 6:50, 00000. 00000. :0000. 00000. 0:0:0. :5500. 00000. :0000. 00000. 50:00. 00500. 00000. 0:000. 00:00. 00050. 0::00. 00000. 00000. 50000. 05000. 05:00. 50:00. 00000. 00000. :0000. 50000. 00000. 00:00. 00000. 00:00. 05000. 00000. 00:00. 00000. 05050. 00050. 00050. :0500. 00000. 00000. 50000. 00000. :0500. 00000. 000:0. :0000. 00000. 5:000. :0000. 05050. 00000. 00500. :0000. 00000. 50:00. 00000. 0::00. 0:000. 00000. 0:000. m02.50 m8E mm <<0<50 3002.000 00000 0 332:0 0000 0 B 0 02:0 (0002 332:0 <000003002:0 0000503002:0 00(0503002:0 (00000002:0 <<00000002:0 <0<0000002:0 0<0000B00200 000<00 30020 0 00000 30020 0 0<0<0 0 30020 0 000<50 00020 0 000<0 0 00020 0 0<000000020 0 000<:000020 0 <<0<00B 00200 <<0000B00200 <<00003 0020 0 <<0<00300200 000000300200 000050 300200 <00000 300200 00903 300200 <<<<5030020 0 00000000300200 000<:0m000200 00000000020 0 coumUOA 78 000.5 000.0 00:.0 005.0 000.0 000.5 :00.5 000.0 000.5 5:0.0 0:0.0 000.0 00:.00 :00.0 000.0 000.0 000.5 000.0 055.0 :05.0 500.5 000.0 :00.5 05:.0 000.: 0:0.0 :00.0 050.5 500.0 000.5 .:00 050.5 000.0 0:0.0 005.0 000.0 000.5 0:0.5 000.0 000.0 :00.0 :00.0 500.00 000.00 5:5.0 500.0 050.0 000.5 000.0 0:0.0 005.0 000.0 000.0 000.5 000.0 005.: 050.0 500.: 500.0 000.0 000.5 .umo .tau 5.000 0.000 0.00: 0.00: 0.00: 0.500 :.050 0.00: 0.000 0.000 0.00: 0.0:5 0.0000 0.000 0.:00 5.00: 5.000 0.05: 0.0:0 0.05: 0.050 5.00: 0.0:0 0.000 0.000 0.00: 5.:5: :.000 0.00: 0.000 0% 0000 0000. :0:0. 00:0. 0000. 0000. 0000. 0000. 0:00. 00:0. 0000. 0000. 0:00. 0000. 0000. 0500. 0000. 0000. 0000. 5000. 0000. 0000.0 0000. 0000. 0000. 0000. 0000. 0050. :000. :000. 5500. m0>0 505.:0 505.00 005.00 000.00 000.0 000.0 00:.00 000.0 000.00 000.0 5:0.5 050.0 000.0 000.0 :00.5 ::0.: 005.0 000.0 050.: 000.: 000.0 000.0 500.5 5::.0 050.0 000.0 000.: 00:.00 000.0 0:5.0 allzwo 005.: 000.0 005.0 500.0 000.: 000.5 000.: 00+00.0 000.0 000.0 000.00 000.00 000.00 00+00.0 00:.: 000.5 00+0:.0 0:5.:0 000.00 000.0 000.00 050.00 :05.00 000.5 000.0 005.00 000.0 50:.0 000.00 xxmz 000.0 000.0 :00.0 000.0 050.0 000.0 000.0 000.0 000.0 :55.0 000.0 000.0 000.0 500.0 050.0 000.0 00:.0. 000.0 055.0 :50.0 000.0 :00.0 000.0 500.0 :00.0 000.0 500.: 000.0 000.0 000.0 mzwo 0000.0 5:00.0 00:0.0 0:00.0 5000.0 :005.0 0000.0 0000.0 0000.0 :0:0.0 0000.0 :000.0 0000.0 0000.0 0000.0 000:.0 0500.0 0000.0 :000.0 0005.0 0:50.0 0000.0 50::.0 0000.0 0000.0 00:0.0 0000.0 0050.0 000:.0 ::0:.0 .0006 685280 803m 3m? #55 0320 baa> 00000. 00000. :5000. 00000. 00000. 00000. 00000. 50:00. 00500. 00:00. 00000. 0000:. 000:0. 00000. 00000. :0000. 00000. 00:00. 00000. 00000. 000:0. 00000. 55000. 00000. 00000. 00000. 50000. 00000. 00000. 00000. 00000. 00500. 00000. 50050. 00050. 0:000. 00000. 00500. 0:000. 00000. 0:500. 00000. 00000. 05000. 00000. 000:0. 00000. 00000. 05050. 00000. :0000. 00000. :0500. 50500. 00000. 00000. 00000. 00000. 00000. 00000. m02.50 n8:0 NW 000<00 30 07000 00902 3:0 200 0000:03:0 200 0<0<:0 30 0200 0<000 30 0200 <0<00 3:0200 0<<000 300 2:0 (0050 3002:0 <0<<003002:0 000<:03002:0 0<0000 3002:0 <<000 300 2:0 00000 300 2:0 000000 0 3002:0 <000003002:0 0<00003002:0 0000003002:0 00000 03002:0 00000 0300 2:0 0000:0300 2:0 000000 000Z:0 0000003002:0 0000003002:0 0<0003002:0 00<0503002:0 (00003002:0 0<<0:03002:0 <<0000 3002:0 (0000 3002:0 0<000 3002:0 50030.0 79 0:0.0 500.0 00:.0 ::0.5 000.0 050.0 0:0.0 000.0 000.0 000.0 :00.5 005.0 000.5 000.5 0:0.0 050.0 000.0 000.5 000.5 000.5 000.00 mum .tam 000.0 0:0.0 500.0 000.5 00:.0 000.0 005.0 000.0 55:.0 0:5.0 000.5 005.0 ::0.5 000.5 0:0.0 000.0 000.0 000.5 000.5 00:.5 000.00 .018 .tam 0.500 0.:00 0.500 0.000 0.0:0 0.0:0 0.0:0 5.:00 0.:00 0.000 0.000 0.0:0 0.000 0.000 0.005 5.00: :.000 0.:00 0.000 0.000 0.000 00?: 0000 5000. 5000. 0000. 00:0. 00:0. :000. 0050. 0000. 0000. 0000. 0000. 0000. 0050. 0000. 00:0. 0000. 0000. 0000. 0000. 00:0. 0000. m0>0 000.00 :::.0 000.0 005.5 5:0.5 0::.0 050.: 5:0.0 000.0 000.0 550.0 000.: 500.00 05:.00 ::0.0 5:0.0 005.00 000.00 0:0.00 005.00 000.5 “1......sz 000.: 000.:0 000.0 005.00 005.00 0:5.00 05:.00 050.50 000.5 000.0 000.5 00:.0 000.5 055.0 000.50 000.0 000.: :00.00 000.: 000.0 000.00 ¥>;0 00000090000 «000mm .0003 0:000 0000 0000cm> méon.o Nogoo. a0m0o. 0n0m.o wgnoo. omm0o. mmn_.c Namoo. n~¢ao. maon.o mmn0o. onwao. omm0.o ”000°. manna. wnm~.o omnoc. anmo. w~nu.o nn-o. ammNO. nmmw.0 woNNO. mom0o. oooo.o cooeo. nnmoo. m0mm.o 0mnoo. unm0o. nmn¢.o cameo. Nam0o. 0n0m.o Romeo. omn0o. a0¢n.o nmnmo. ”000°. 0mm¢.o 00ono. amomo. ammu.o wuuao. 0NNN0. wnm~.o 0No0o. mango. un~0.0 m0noo. 00moo. umen.o u0¢mo. ”ammo. nume.o ¢¢moo. nouoo. naew.o ¢~¢0o. .m0mo. 0mom.c m00m0. m0n00. é0&6 Maura n8:0 om 00000 3002:0 0000030 02:0 00000 3:02:0 <<0<:0 3:02:0 0<00 0 0 3:02:0 (00000 3:02:0 00<<003:02:0 0<0:0 30 02:0 00<0 0 3002:0 000<503002:0 00000300200 <<<0030 020 0 0000003:0 20 0 0000003:0Z0 0 0000003:0200 0000003:020 0 0<00 0 30020 0 000000300200 0000000 30 0200 00000300200 (00400300200 50080.0 80 TABLE 8: Rock wells from the Saginaw Formation used for ion ratio comparison from study areas. OBNOIEBBABB 03N01E33ABC 04N01E27BAA 02N01E15DDBB 02N01E25CDDA OZNOIEZSDACC OZNOZEOZACDD 02N02W20ABA OZNOZWZOBBC OQNOZWZODCD OZNOZWBOBCC 02N03W lZADD 02N03W12CC 02N02W19AABA OBNOZW l lADAD 03N02W18DCB STUDY AREA A OBNOIEOZAAAA OBNOIEOSABAB OBNOIEOSDDCC OBNOIE lZBCC 03N01E22BAAD 03N01E32BBCC OBNOlEB‘IADDD STUDY AREA B O3N02W23ADBB 03N02W23CBAD OQNOZWOQDDCC 04N02WO5CCCC ' 04N02W5DCDC 04N02WO8BCDC 04N02W09BDAD 04N02W09CBCA 03N02E0l¥ACCC OBNOZE lZBCAD OBNOZE 15ABBC 03N02E17ADDD 04N01E08DADA 04N01E20CCA 04N02W11DBDD 04N02W18BDAD OlINOZW leADDD OlIN 02W25AABA 04N02W27CDA OliNOZWBGDAAD 02N03W BODAB OZN 03W BllCAB "iiiiiiiiii'iilflifliiiflfli