llllllllfllllllllll 93 00100 3650 x» LMEARY ‘ “me Michigan sm- 3 ‘2 University This is to certify that the thesis entitled AN ISOTOPIC INVESTIGATION OF GROUNDWATER IN LEELANAU COUNTY, MICHIGAN presented by DAVID PHILIP REGALBUTO has been accepted towards fulfillment of the requirements for @— degree i% Ma' professor Date a“)! 022,, /7é;7 07639 MS U it an Afimm‘ve Action/Equal Opportunity Institution M _.“.——A_ -_.~.— IV1£3l_} RETURNING MATERIALS: Place in book drop to LJBRAjJES remove this checkout from .—_- your record. FINES will be charged if book is returned after the date stamped below. - MAR 2 1 2| iii-A: I’Cw ! 05 O 9 1;. g g fix? -.“- n ‘7. w P 1' 92605 5"”? h, ¢ {ii-LL!“ (,2 R 2 'JLu'Jllsl932 [0‘0 (law’s \§ / )Cme S AN ISOTOPIC INVESTIGATION OF GROUNDWATER IN LEELANAU COUNTY, MICHIGAN By David Philip Regalbuto A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Geological Sciences 1987 @1988 DAVID PHILIP REGALBUTO All Rights Reserved ABSTRACT AN ISOTOPIC INVESTIGATION OF GROUNDWATER IN LEELANAU COUNTY, MICHIGAN By David Philip Regalbuto Tritium and the stable isotOpe ratios of hydrogen and oxygen are used to study the timing, components and rates of recharge to a glacial drift aquifer system. Bomb-tritiated waters are found from O to 25.0 meters below the water table. Dead waters are found below 19.2 meters. The bomb-tritium interface is placed from a minimum of 20.6 to a maximum of 25.2 meters below the water table. These depths are used to calculate an average recharge rate of 17 cm/year. on values ranged from -92.2 (0/00) to -82.u (0/00). also ranged from -13.03 (0/00) to -ll.8l (O/OO). These ratios indicate of recharge during the cool months of the year. A one-dimensional transport code is used to model the influence of lake-effect snowfall on tritium concentrations in groundwaters. The model shows that this input is respon- sible for low observed tritium concentrations relative to those predicted by the tritium input function. ii ACKNOWLEDGEMENTS The author wishes to extend his gratitude to those who have made this study both possible and rewarding. First of all is my appreciation of the advice and friendship extened to me freely by the thesis advisor, Grahame J. Larson, who on many occasions has gone the extra mile in supporting me. I wish to thank the other guidances commitee members, Duncan F. Sibley and Roger B. Wallace, for their insights and suggestions. A special word of thanks also to Duncan Sibley, for being there when the one who was down under could not. I am indebted to the Sentiniel, G.w. Monoghan, who created the amazing code which saved me hundreds of hours in doing science, and to Manrico Delcore, whdse guidance on the fine art of liquid scintillation counting was irreplacable. Most of all I thank Caroline Olmsted and my parents for their endless encouragement and support. The Michigan State University Foundation provided funds for stable isotope ratio analyses. iii TABLE OF CONTENTS LIST OF TABLES.............. ..... ...... ..... ... ..... ..... Vi LIST OF FIGURES.. .................. ............ ..... .....Vii INTRODUCTION... ....... ...... ....... ...................... l Tritium.......IOOOOOOOOOOOOOOOO0.0.0.0...00.0.0.0... 1 Stable Isotopes.. ......... ............. ....... ...... 2 SCOPE OF RESEARCH.... ..... ....... .......... .............. A Tritium..... .......................... ......... ..... 6 Stable Isotopes ................ . ..... ............... 9 STUDY AREA. ..... . ..... ..... .............. ................ l2 Geology... ......... . ....... . ..... . ............... ... l3 Hydrology.... ................ . ........ .............. l8 METHODS. . . . ....................................... . ...... 22 Sampling...... ........................... ........... 22 Tritiated Sample Preparation ............. ........... 23 Analysis. ............ ................. ..... .... ... 25 RESULTS..... ............... . ............. . ......... ...... 26 Tritium..... ........................... ....... ...... 26 Stable Isotopes..... ......................... ....... 28 DISCUSSION. ................. . ............................ 28 Recharge Rates ................................... ... 28 Stable Isotope Ratios ............................... 33 The Possible Effect of Lake Michigan Water on Tritium Activities in Groundwaters ............... 34 CONCLUSIONS........... ................... . ............... 39 RECOMMENDATIONS FOR ADDITIONAL RESEARCH......... APPENDICES..................... APPENDIX A - QUENCH MONITORING AND COUNTING EFFICIENCY.......... APPENDIX B — TRITIUM COUNTING DATA.......... APPENDIX C - TRITIUM ACTIVITY CALCULATIONS.......... BIBLIOGRAPHYOOOOO00...... 42 AA AA A6 A7 50 Table 1. Table 2. LIST OF TABLES Discharge characteristics of the Boardman River drainage basin.................. 20 Comparison of average observed tritium concentration with average simulated values, at varying solute exchange fractions..u.............................s..... 40 vi Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure .Figure Figure Figure Figure Figure Figure Figure LIST OF FIGURES 1. Location of study area in Leelanau County..... 2. Tritium input function for the study area..... 3. Mean annual (Lamb, 1961) amd growing season (Bernabo, l981)temperature curves for the past 1100 years in the northern Ilemisphepeo.....IOOOOOIOOOOO... - A. Stable isotope input function.............. 5. Quaternary deposits of the study area (After Martin, 1957).......... A_A’ooo 00000000000000 0 6a. Cross section 6b. Cross section 6c. Cross section 7. Till thickness in the study area..... 8. Piezometric surface of the study area..... 9. Theoretic flow, in two dimensions, at a groundwater divide, in an unconfined aquifer. (after Toth’ 1962)......OOOOOOOOOOOOOO. 10. Sampling locations ................... . ........ 11. Plot of tritium concentration vs. depth....... 12. Plot of at) and 6130 vs. depth..... ......... 13. Plot of ob vs. (3180........................ . 1A. Comparison of simulated tritium concentrations when D=O and D=10E-L}m/dayoooo 000000000000 one... ...... no 15. Comparison of maximum observed tritium concentration with simulated values, at varying solute exchange fractions........ ..... l6. Quench curve for tritium counting window 170-225... vii 10 11 114 15 16 16 17 21 22 214 27 29 3o 36 38 AA INTRODUCTION Environmental isotopes, such as deuterium, tritium and oxygen-18 are attractive as either point or. non—point tracers because they possess all the traits of the "ideal tracer", at least in shallow groundwater environments. They are essentially conservative in their reactions with the host porous medium; they experience no retardation with reSpect to groundwater flow; they are a part of the water itself and they are continuosly added to the groundwater system in time and space. There does exist exchange of oxygen isotopes, and to a lesser degree hydrogen isotopes, between the liquid and solid phases in an aquifer, above all where carbonates and shales are the dominant lithology. This has been shown, however, not to be significant below formation temperatures of about 50° (0) (Clayton, 1961). Tritium Tritium became a useful tool in the field of subsurface hydrology beginning in the late 1950's. By this time atmos- pheric nuclear testing had resulted in elevated tritium levels (termed "bomb" tritium) in the atmosphere and precipitation, much higher than existed prior to late 1952, when such testing began. The utility of bomb tritium as an age indicator lies 1 in its global distribution in meteoric waters and the ease of distinguishing between pre— and post-1953 groundwaters. For example, post-1953 tritium concentrations in recharged waters were found to be as much as four orders of magnitude higher than ambient conditions which existed prior to 1953 (Thatcher and Payne, 1965). Pre-l953 ambient background tritium concentrations have been estimated as being from 10 to 20 tritium units (TU) (Brown, 1961), with a TU being equivalent to one tritium atom per 10E18 hydrogen atoms. As the half-life of tritium is 12.3 years, background concen- trations in meteoric waters older than 1953 should not exceed 5 TU. Water of this age is commonly termed "dead". Most tritium research applied to the subsurface uti- lizes the isotope to either delineate regional flow patterns ( Rabinowitz, 1977; Wurzel, 1983; Kondoh, 1986), or to measure recharge rates (Andres, 1985; Offer, 1982; SukhiJa and Shaw, 1976; Brendenkamp et a1., 197A). Most recently, tritium has been used as an indicator of dispersion in unconfined aquifers (Egboka et a1., 1983). Stable Isotopes By the late 19A0's, research into the variability of stable hydrogen and oxygen isotopic ratios was well under way. Major advances were possible due to refined methods in mass spectrometry. The works of Craig (1961a), Dansgaard (196A) and Friedman et a1. (1964) are recommended as early summaries of stable isotopic behavior, and the mechanisms 2 3 which account for their fractionation. The reader is referred to Ferronsky and Polyakov (1982) for a more recent comprehensive reviews of stable isotopes and their applica- tions in hydrogeology. Isotopic fractionation of deuterium and oxygen-18 occurs primarily as water is evaporated and precipitated. Since oxygen-18 (180) and deuterium (D) are heavier than oxygen-16 (160) and protium (H), respectively, their vapor pressures are lower than their lighter isotOpe counterparts at any temperature. For a pair of isotOpes, however, the ratio of their vapor pressures, termed the fractionation factor, is not constant with respect to temperature. The fractionation factor of any isotopic pair increases as the temperature of evaporation decreases, such that at cooler temperature newly formed water vapor is preferentially enriched with respect to H and 160. Stable hydrogen and oxygen isotopic ratios are normally given with respect to Standard Mean Ocean Water (SHOW), where: 00(0/00) = [(D/H) - (D/H) / (D/H) J x 1000 sample SMOW SMOW and 0180(0/00) = [(180/160) — (180/160)/<180/160) J x 1000. sample snow snow (Craig, 1961b). . Negative values would indicate "light" water, or preferen- tially enriched with 16O or H. The opposite would be true for positive values. Thus, the utility of the stable isotopes D and 18O in groundwater research is that their ratios, in meteoric waters, to H and 16O are largely a function of the tempera- ture of their evaporation and precipitation. As a result, the isotopic signature of a groundwater body can be tied to the climatic conditions of its formation and recharge. Stable isotopic ratios have been used to differentiate between connate and meteoric waters of recent age (Clayton et a1., 1966; Hitchon and Friedman, 1969; Graf et a1., 1965). Since they are sensitive to temperature, stable iso- topes can be used to identify climatic events in ground- waters whose flow patterns are understood. Of particular interest in temperate regions is the study of residence times of groundwaters since the end of the Pleistocene Epoch. At several locations in the Great Lakes region, groundwaters with very negative isotopic ratios have been interpreted as being of Pleistocene age (Clayton et a1., 1966; Desaulniers et a1., 1981; Long et a1., 1986; Sklash et a1., 1986). SCOPE OF RESEARCH The~ research presented in this paper uses tritium and stable isotopic ratios largely to understand the nature of recharge and the age of groundwaters in a heterogeneous drift aquifer system underlying part of Leelanau County, which is located in northwestern lower Michigan (Figure 1). 10km LEELANAU COUNTY Northport r Figure 1. Location of study area in Leelanau County. Tritium The use of tritium to determine recharge rates involves sampling at different depths in a groundwater column in order to "sandwich" the depth of the interface between bomb—tritiated water and the dead water beneath it. The bomb-tritium interface method was first applied to studies of the unsaturated zone (Smith et a1., 1970; Dincer et a1., 1974; Zimmerman et a1., 1967) and has been used within the 1980's to study recharge rates in the saturated zone (Andres, 1985; Maloszewski, 1980). The first studies of recharge in the saturated zone using tritium were performed on simple, relatively homogeneous, highly permeable aquifers (Offer, 1982; Larson et a1., 1987). More recently, this method has been applied to more complex aquifer systems (Delcore, 1985; Delcore and Larson, 1987; Kondoh, 1986). One of the goals of this research is to attempt to extend the tritium interface method to a complex and heterogeneous drift aquifer system. In order to properly interpret tritium activities in groundwater profiles it is necessary to have a history of tritium levels in the meteoric waters available for recharge. The tritium input function for the Leelanau County area is shown in Figure 2. The years indicated on the input function are rearranged as water years, with a water year beginning in October. The input function gives mean tritium concentrations in precipitation occurring during the recharge season within a water year, and the depth of recharge season precipitation. Examples of tritium input _ Average TU's in precipitation during recharge season (Oct— May) ‘Allll - Above values. corrected for decay — Depth of precipitation during recharge season 10- LL] 1 I- _-———__ _-—————— _ . — l — — l TRITIUM CONCENTRATION (TU'S) m PRECIPITATION (cm) 1950's 1960's 1980's Figure 2. Tritium input function for the study area. 8 functions are found in Rabinowitz et a1. (1977), Delcore (1985) and Offer (1982). Tritium activities used in the input function are those reported for precipitation by the International Atomic Energy Agency (1969 through 1986) at its Ottowa, Canada, station. Data is also available from the Chicago, U.S. station, but the record isn't continous from 1953 to the present. - One problem which has perplexed isotOpe hydrologists working with tritium is that tritium activites measured in groundwaters are commonly less than those measured in precipitation (Egboka et a1., 1983; Delcore, 1985). The pos- sibility of analytical error is an unnacceptable explanation for this observation, for it is possible to accurately and precisely measure the activities of tritium standards in the laboratory. One explanation is that dispersion is respo- nsible (Egboka et a1., 1983). While this is certainly plaus- ible, the input of nearly dead water from Lake Michigan in this study area is believed to be partly responsible for the observed low tritium values. One of the goals of this research, then, is to take the tritium activities from the input function, account for decay and dispersion, and attempt to match the observed tritium concentrations with a simulation by varying the amount of the suggested mixing in of the relatively dead water derived from Lake Michigan. Stable Isotopes D and 18O in this study serve three purposes. First, the D/H and 180/160 ratios of groundwater samples will be compared to those of present day snowmelt and rainfall, in order to get an idea of whether rainfall or snowmelt domin- ates the recharge. The second usage is to test for the recharge of pre- Holocene (>10,000 yrs.) waters. The existence of pre-Holo- cene water in a shallow drift aquifer system would indicate very long residence times within the regional flow system and the existence of nearly impervious atrata overlying the aquifer. At a few sites within Michigan and Ontario, particularily beneath thick clay sequences, both deep and shallow groundwaters have shown extremely light (or nega- tive) values for hydrogen and oxygen ratios (Clayton et a1., 1966; Desaulniers et a1., 1981; Long et a1., 1986; Sklash et a1., 1986). It is believed that these waters represent recharge during Pleistocene times, as the extremely cool climates during glacial events would have a profound frac- tionation effect. Stable isotopes will be used thirdly to test whether the effect of a less severe but more recent climatic episode can be identified. A host of studies dealing with climatic changes over the past few thousand years are in close enough agreement to suggest that a significant cool period, termed the "Little Ice Age", persisted from about the 16th through 19th centuries (Williams and Wigley, 1983). Of par- ticular interest is a growing season temperature curve 10 (Figure 3) established by Bernabo (1981) using 1“Cudated Y 033 Lamb Bernabo (.1961) ( 1981) 1960 1 1800 . ' 1400 1 - 1200 1 - 1000 - - 1800 --4 le 10. 5 16.5 5°C °c17 Figure 3. Mean annual (Lamb, 1961) and growing season tempe- rature (Bernabo, 1981) curves for the past 1100 years, for two locations in the northern hemi- sphere, including the study area in Leelanau County. palynomorphs from a lake some 15 miles east of the study area in Leelanau County. Based on Webb and Clark's (1977) transfer function, which relates pollen deposition to summer temperatures, Bernabo has calculated that tempera- tures during the late 17th to early 18th centuries were about 1°(C) cooler than the 30-year interval of 1931-1960 in northern Michigan. Bernabo's curve agrees qualitatively with 11 that prOposed by Lamb (1961) for central England. If real recharge rates are small enough in the study area, that is, on the order of a few centimeters per year, it should be possible to sample water two to three hundred years old, for the at this rate the isotopically light water recharged during this cool period would not have exceeded the depths of the existing well coverage in the study area, from about 10 to 30 meters below the water table. In this case, a curve relating isotopic ratios to depth will mimic Bernabo's temperature curve, as ratios should become increasingly negative with depth to the point in time when the Little Ice Age reached its zenith. This relationship is demonstrated in a stable isotope input function (Figure A). 1960a g: 1800‘ 1600- 1400- 1200- 1000- BOOJ oS—fi ' A00 0 Figure 4. Stable isotOpe input function. 12 The stable isotope input function integrates Bernabo's temperature curve and Dansgaard's theoretical calculation (19611) of a 0.5 0/00 00 and 0.70/00 0180 shift per 1° (0) in temperature. By knowing the depth at which the ground- water is lightest, it might be possible to calculate a recharge rate for the past two to three centuries. It is expected that actual recharge rates to the drift aquifer system in the Leelanau area are much too high for Little Ice Age water to be sampled. The drift in this area, however, is known to contain clay-rich strata. It may be possible that, given a till with a high enough clay frac- tion, hydraulic conductivities are low enough to limit recharge rates to fractions of an inch per year. STUDY AREA The study area (Figure 1) is the peninsula of Leelanau County, located in the northwest part of Michigan's lower peninsula. The study area on the Leelanau Peninsula covers approximately 500 square kilomters. The peninsula is bounded on the west by Lake Michigan and on the east by the west arm of Grand Traverse Bay. The Leelanau Peninsula was selected as a study area because while if offers a relatively complex glacial geological environment in which to test the bomb tritium interface method, local flow patterns are simplified by the presence of major discharge boundaries (Lakes Leelanau, Michigan and Grand Traverse Bay) close to one another. By 13 knowing precisely where the piezometric surface is lowest, the groundwater divides between Grand Traverse Bay and Lake Leelanau, and between Lakes Leelanau and Michigan can be accurately defined. Geology The surficial Quaternary deposits of the Leelanau Peninsula, shown in Figure 5, are of Woodfordian age, being formed most likely during the Great Lakean readvance (Even— son et a1., 1976). The southern end of the peninsula east of Lake Leelanau is underlain by a segment of the Manistee Moraine. The till associated with the moraine consists chiefly of a poorly sorted mixture of coarse-grained sand and stones, most of which are carbonates. In most areas the till is poorly sorted, although well-sorted sands were observed during field reconnaissance. Some clay-rich areas are present in the region between the town of Bingham and the southern boundary of the study area. Ice-contact struc- tures, chiefly slumped and convoluted beds of till and outwash, are exposed in a few roadcuts and gravel pits along the eastern flank of the moraine. Highly drumlinized till plains are developed west and north of the Manistee moriane. Between Lakes Leelanau and Michigan some very localized outwash deposits cover the till plain along its western and eastern flanks. The texture of the till is fairly consistent from north to south. It is predominantly stony and sandy but with a much higher clay content than the till observed on the Manistee moraine. 14 QUATERNARY DEPOSITS Scale 10 km EXPLANNHON Dune Sand Lacustrine Manistee Moraine Drumlinized Till Plain (after area deposits of the study Quaternary 5. Figure 1957). Martin, 15 Geologic cross sections of both till plains and the Manistee Moraine are shown in Figure 6. Correlations based on drill- ers logs are tenuous at best, given the variable nature of the drift, the distances between control wells and the interpretive nature of formation logging during drilling. Lacustrine clays are confined to the areas immediately adjacent to the lakes, although much of the area from Cathead Bay to the end of the peninsula is underlain by sand, gravel and stones associated with the Algonquin and Nipissing strandlines (Martin, 1957). 300- DRIFT TYPES A--3oo Sandy Clay-rich -275 Sandy wz stones g Clay-rich w/ stones 275 " C --250 g E w 2 225—1 . 225 I; 3 W a " ' ‘-‘ 3'3.-‘.-':::-'7:-:2:'7‘ ":?3.':-'~,1.-:;:-:1- '.~.'-.‘;':;'=;':§:;-,- 20°.— 1 km - a}- .. ~"-" -—200 175 175 Figure 6a. Cross section A-A . DRIFT TYPES Sandy Clay -rich Sandy a Stony E Clay - rich 4 Stony B Figure 6b. DRIFT T Sandy Sandy a Stony Cross section B—B’. YPES Clay - rich g Clay-rich 81 Stony C 25° Figure 60. Cross section C-C’. Elevation (mast) M N (I 200 Elevation (mast) 17 Till thickness, shown in Figure 7, increases from north to south. North of Northport it is in the range of 50 to 75 meters, increasing to about 75 meters near Omena, then thic- kens to over 13)nwters Just two to three kilometers south of Omena. This marked increase lies at about the same E 60-90 90-150 I 10km Figure 7. Till thickness in study area. 18 subcrOp location of the contact between the Traverse Lime- stone and the Ellsworth Shale. It is thought that the increase in till thickness results from the relatively easier scouring of the Ellsworth by glacial ice. Based on drillers logs isolated lenses of clay-rich drift are known to exist in the subsurface, to the base of the drift, throughout the peninsula. Hydrology Annual precipitation decreases from west to east across Leelanau County, from about 90 centimeters/year to 76 cm in the immediate Traverse City area. Precipitation on the pen- insula, as recorded for the period of 1953 to 1985 (corres- ponding to the time covered by the tritium input function) is approximately 81 cm/year (Nurnberger, 1952-85). As seen by the tritium input function (Figure 2), the recharge season during these years experienced an average of 44 cm of precipitation per year. The nearest drainage basin whose discharge is gaged is that of the Boardman River, which drains some 530 square kilometers of mostly outwash, develOped between the Manistee and Port Huron moraines, to the southeast of the study area. For the 33-Year interval from 1953 to 1985, the mean annual discharge of the Boardman was equivalent to 36.6 cm of total runoff from the basin (USGS, 1953-1986). Hydrograph sepera- tions, using three different techniques, were performed on flow data from the water years of 1957, 196A, 1971, 1978 and 1985. These techniques - the sliding interval, the fixed 19 interval and the local minimum are described in detail by Petrie (1985). Annual mean dischages of the Boardman river and the percentages of the baseflow components are summar- ized in Table 1 for the years of records mentioned above. It was found that that average base flow, based on the five sets of hydrograph seperations, is about 82 percent of the total runoff, or about 30.0 cm per year for the drainage basin. Aquifer conditions on the Leelanau Peninsula are mostly unconfined although semiconfining conditions do exist local- ly where lenses of clay-rich till occur in the saturated zone. The many flowing wells located close to the shores of Lake Leelanau and Sutton's Bay are strong evidence that confined conditions exist in areas underlain by lake clays. A water table elevation map of the peninsula is shown in Figure 8. It is based on a total of 414 well logs whose locations could be verified with confidence. As might be expected, the two major groundwater divides follow closely the traces of topographic divides. The major uses of groundwater in the study area, besides domestic consumption, are for irrigation and cherry cooling. These large capacity wells, which account for only a small percentage of the wells on record in the area, produce up to 500 gallons a minute. Cherry cooling is limited chiefly to the month of July. Table Water Year 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 Means: Baseflow of total runoff = recharge Runoff 20 Discharge characterisics of the drainage basin. Boardman River Percent of Total Runoff as Baseflow Total Local Sliding Fixed (in.) 7-day 15-day 7-day 15—day 7-day 15-day 37.8 37.8 3709 34.9 35.1 84.4 79.3 84.6 80.6 84.2 81.0 30.4 32.0 15.2 34.1 38.3 32.7 30.4 79.0 71.0 80.1 73.2 80.2 72.3 32.3 36.8 42.6 36.5 39.8 38.7 40.7 82.2 73.3 84.3 77.6 85.3 77.3 34.9 33.0 34.9 36.8 43.7 33.6 35.5 87.5 84.8 91.7 86.9 91.5 86.1 37.0 35.6 27.8 30.1 37.4 36.6 43.6 84.3 80.4 85. 7 80.2 85.1 80.4 36.6 83.5 77.8 85.4 79.7 85.3 79.4 Mean of three methods = 81.9% (36.6 cm/yr)(0.82) 30.0 cm/yr. 21 PIEZOMETRIC SURFACE CONTOURINTERVAU 10m SCALE 10km Figure 8. Piezometric surface of study area. 22 METHODS Sampling Sampling locations for all isotopic analyses were limited to the wells which fall on or immediately adjacent to the two major groundwater divides discussed above. Where an environmental tracer is being used to determine recharge rates, it is essential that sampling be done on a divide for it is here that movement in an aquifer is essentially verti- cally downward, as suggested by Toth (1962) (Figure 9). Sampling off a groundwater divide results in values which are less than actual since a component of flow is horizontal. Groundwater DhNde ——) Flow Line ----- Equipotential Line Figure 9. Theoretical flow, in two dimensions, at a groundwater divide in an unconfined aquifer (after Toth, 1962). 23 O:‘ 1 SAMPLE A -a LOCATIONS a 0. I :2 .' .9 .6. I _ 3H, 2H, 180 O o J 0 - 3H only , ,- o - Surface . fa. . j. - , ll 0 - Snow ..1. o. g a: -. l A - Porosity ‘: . 5b - Groundwater 5 0' . ' Divide c] I I 9 . - Scale ‘ 9 10km \ o 0.0 1 Figure 10. Sampling locations. 24 Twenty nine domestic wells were sampled for tritium. Depths of domestic wells ranged from 7 to 28 meters below the water table. Well screens are normally 1.2 to 1.6 meters (4 to 5 ft) long. Eighteen wells, including one high production well, were sampled for stable isotopes. Their depths ranged from 8 to 91 m feet below the water table. Five of these are deeper than 27 m. Samples were taken by simply turning on a tap or spiget served by the domestic well, and letting it run for about two minutes. This serves to stablize the well system temper- ature and ensure that the sample hasn't undergone exchange with any of the well components. Samples were delivered to 500 m1 Nalgene plastic bottles, which were oven dried and capped before sampling. Four surface water locations were sampled for stable isotOpes - Grand Traverse Bay, south Lake Leelanau, Lake Michigan and a small creek which discharges to south Lake Leelanau. A composite sample of about 0.3 meters of snow was sampled from the Northwest Michigan Agricultural Extension Station, south of Bingham. All sampling locations are shown in Figure 10. Tritiated Sample Preparation Samples to be analyzed for tritium activity were prepared for decay measurement by the U.S.G.S. electrolytic enrichment method (Thatcher et a1., 1977) of pre-distillation, electrolysis, and post-distillation. The 25 method is also described in detail by Wyerman (1976). Other descriptions can be found in Offer (1982), Slayton (1983) and Delcore (1985). Pre-distillations are run with a small quantity of powdered copper, which ensures that any sulfer present is sequestered. The radioisotope of sulfer, like tritium, is a beta - emitter, whose energy emission spectrum overlaps that of tritium. Electrolyses were run in sets of six to eight samples. Each set was run with one standard and one blank consisting of dead water, which is taken from a 500-meter deep well at the fish hatchery in Thompson, Michigan. Electrolysis reduces the sample volume from 200 ml to between 5 - 10 ml, therby enriching the sample with respect to tritium. Post-distillations were run using liquid nitrogen as the coolant. Such extreme temperatures are necessary due to the small mass of sample at this point. Upon completion, approximately 5 to 6 ml of sample is delivered to a 20 m1 borosilicate glass vial, and mixed with a liquid scintil- lation counting cocktail. Analysis Tritium activities were measured by means of liquid scintillation spectroscopy, which is an indirect process. The liquid scintillation spectrometer counts not the actual decay events, but rather, detects photons produced by the bombardment of the cocktail by beta radiation from the 26 tritiated water molecules. Produced at a rate preportional to the frequency of decay, the detectable photons are photo- multiplied and recorded electronically as decay events. For a thorough treatment of the liquid scintillation technique and theory, the reader is referred to Long (1978) and Florkowski (1981). All samples were counted for at least 270 minutes. Several 'of the deeper well samples were counted for over 1000 minutes or longer. The deeper samples require longer counting times in order to reduce their 2-sigma errors to the point where they can be distinguished as dead or bomb- tritiated waters. All tritium counting data is presented in Appendices A and B. The method for calculating tritium activities is presented as Appendix C. Stable isotopic ratios were measured by means of mass spectroscopy, at the Isotope Research Laboratory of the University of Waterloo in Waterloo, Ontario. RESULTS Tritium A plot of tritium activity as a function of depth is shown in Figure 11. Depths are given in meters from the water table to the bottom of the well screen. Assuming any activity over 5 TU indicates the presence of bomb tritum, the deepest penetration of bomb—tritum lies between 25 and 25.5 meters. The average depth of all screens, however, is about 0.7 m less, since they are about 27 TU [X104] 1 3456789101112131415 C 0 l T J i l l l l I .1 l l l l J T7 I 5_ l I O _ I . . E ‘ o o a . - - ' a P O . t 15" I . C ,2 l a o 3 I . . 20- - 0 I m 0 3’ Lowest Bomb Tritium C) l I 30 I I 35 I Figure 11. Plot of tritium concentration vs. sample depth. 1.2 meters long. This places the deepest bomb-tritium at an average depth of 25.2 m. The shallowest dead water (less than 5 TU), on the other hand, is found between average depths of 19.2 and 22 m (again, accounting for screen lengths), or an average depth of 20.6 m. The dead water sampled at a depth of about 10 m is not used to define the depth of the interface, for the well from which this sample 28 was taken falls about 2 km laterally off the groundwater ivide in the northern part of the peninsula. It is believed that at this location the groundwater flow is being strongly affected by horizontal and perhaps upward flow components. Stable Isotopes Plots of stable isotope ratios in the groundwater samples, as a function of depth, are shown in Figure 12. It is immediately apparant that there is no correlation between depth and deviation from SMOW for either isotopic ratio. The plot of 0D vs 01%), 1uflxfliirmludes both surface and groundwater samples, is shown in Figure 13. As might be expected the snow sample gives the most negative values. Groundwater values are slightly more positive, while the creek which discharges to Lake Leelanau yields values in the same area as the groundwaters. Lake Michigan gives the most positive value; Lake Leelanau is intermediate between Lake Michigan and groundwater. DISCUSSION Recharge Rates The formula used to calculate recharge rates is simply an expression of velocity: R = (D X Ne) / T (Delcore, 1985; Andres, 1985; Offer, 1982), where R is the recharge rate in meters per year, D is the depth of the Depth Below Water Table (m) 29 00 (smow) 0/00 -ap -815 -9p ‘915 I 0:0 -18 54 .- O o o 10- 0 O P . O o 000 . o o 15- O . o o 20- 0° 0 ’ 25. . o oo 30 I t 1 I I r _' -1175 -120 -1225 42.5 42.75 43.0 -13.25 b180(SMOW) 0/00 Figure 12. Plot of 0D and 0180 vs. sample depth. 30 -15 —14 -13 -12 -11 -10 -9 -8 -7 l l l 1 l L g l f I / —55- / O :55 / , Lk. / Michigan _. .. /~ .- 65 / 65 / «63” -75. (93/ :75 Creek discharging into *0; 0 Lk. Leelanau bD% Lk.Leelanau / ‘ o, (SMOW) -85- / f35 Groundwaters - .. ’ ‘95 95 / I' I I ‘ / -105- I :105 / I .Snow -115 , I I I I I I I T -115 —15 —14 -13 -12 -11 -10 -9 -8 -7 180 0/00 (SMOW) Figure 13. Plot of 0D vs. 0180. 31 bomb—tritium interface in meters below the water table, Ne is the effective porosity of the aquifer and T is the amount of time, in years, between the first input of bomb-tritium into the aquifer system (1953) and the date of groundwater sampling (1985). When solving for R, the two values for depth, 20.6 and 25.2 m must be used, since the lowest depth of bomb-tritium, shown in Figure 11, does not coincide with the highest dead water. Porosity of drift materials was measured on a total of six samples, whose locations are shown in Figure 10. The samples were collected in undisturbed till about one meter below the land surface. The average porosity measured in these is 31.5% (+/- 3.5%). The effective porosity, however is estimated to be near a nominal value of 25%. The recharge rates to the aquifer system are calulated to be a minimum of: R (20.6 m)(0.25) / 33 yrs 0.16 m/yr and a maximum of: R (25.2)(0.25) / 33 Yrs 0.19 m/yr. The average recharge rate would thus be 0.17 meters per year, or about 17 cm (6.5 inches). That this value is only 60% of the depth of baseflow for the Boardman River Basin (30.0 cm), for the same 33-year interval, raises the ques- tion of how realistic the calculated recharge rates are, and hence if the interface method is applicable to a hetero- geneous aquifer system. 32 A number of arguments can be made to account for the apparently low values for R calculated above. First of all is the difference in drift textures. The drift on the Leela- nau Peninsula, while being predominantly sandy, contains much more clay than the outwash system underlying much of the Boardman River basin. This results in a lower infiltra- tion capacity, which increases surface runoff on the penin- sula. The relief on the Leelanau peninsula must also be considered. Much of the peninsula is underlain by slopes exceeding 25% (Weber, 1973). Dunne and Black (1971) and Price and Hendrie (1983) demonstrated that over 50% of the equivalent water depth of a snowpack can be lost as surface runoff, even over sandy soils, during a snowmelt period when soils are still frozen. Petrie (1985) demonstrated within the greater Grand River drainage basin in south-central Michigan that relief exerts a greater influence in reducing baseflow than does infiltration capacity. It is suggested by the author, then, that an average recharge rate of 6.5 inches per year is realistic for the study area, given the nature of the drift and the relief developed on the peninsula due to drumlins and the Manistee moraine. The value of 17.5 centimeters per year is close to the nominal value of 20.3 cm (8 inches) per year which is com- monly used by the USGS in Michigan, for recharge to till plains when calibrating groundwater flow models (Granneman, 1987, personal communication). 33 Stable Isotope Ratios Values for D in groundwater samples range from a low of -92.2 to ahigh of -82.4. For 0130, the minimum and maximum are -13.03 and -11.81, respectively. These "light" waters are indicative of recharge during cool temperatures, as ratios for 0D in precipitation during the summer months are normally 30 to 50 0/00 more positive than the observed ratios. Values for 0180 in summertime precipitation are commonly 4 t0 6 0/00 more positive than the observed values (IAEA, 1953-1986). The snow sample from the study area yielded values of -15.29 and -112.0 for C980 andbD, respect- ively. It is suggested, then, that the source of groundwater in the study area is local precipitation, being a combina- tion of rainfall and snowmelt. This conclusion corroborates measurements of evapotransporation from northern lower peninsula and the eastern upper peninsula, where it is found that ET exceeds precipitation from May to August (Nurnberger, 1982). From Figure 13 it is seen that there is no systematic decrease in isotopic ratios with depth, contrary to the proposed stable isotope input function (Figure 4). This is one piece of evidence that none of the groundwater samples reflect recharge during the Little Ice Age. Besides the stable isotopic ratios, the average vertical displacement of the tritium interface - 22.6 meters in 33 years - would place water 300 years old at a depth of more than 200 m below the water table. This would make it impossible to sample the isotopically lightest water, recharged during the 34 height of this cool period, with the existing well coverage. The Possible Effect of Lake Michigan Water on Tritium Activities in Groundwater According to the tritium input function (Figure 2), some of the deeper samples from the Leelanau Peninsula should have activities over 200 TU, even exceeding 400 TU. The highest activity found in this study, however, is only 136 ,TU; in fact only four wells yielded water of 100 TU's or more. There are a number of possible explanations for this. It may be that the zone of highly tritiated water, recharged from 1962 - 1965, was not sampled, for well screen coverage in the bomb-tritiated zone is not complete. This notion is rejected, however, because at an average seepage velocity of 22.6 m/33 yrs, the highest value used in the input function (426 TU in 1964) should occur where there is adequate well screen coverage (14.4 m). It can be argued that the low measures values are a result of averaging values over the length of a well screen. This explanation, too, is rejected, for at an average recharge rate of 16.7 cm/yr and an effec- tive porosity of 25%, the vertical displacement in the four years time mentioned above exceeds the length of any four or five-foot well screen by a factor of nearly two. A third possibility is that that original tritium activities, upon entering the aquifer, have been altered by dispersion. The input function does not account for hydrody- namic dispersion, for it assumes piston flow in the 35 saturated zone. Hyrodynamic dispersion would result in decreased maximum values, for it causes mixing due to varia- tions in groundwater pore velocities (Ogata, 1970). Piston flow, as described by Nir (1964), assumes that essentially no mixing occurs as a recharge event (rainfall or snowmelt) adds a discrete increment of water to a groundwater column which is contiuously translated downward, as is the case on a groundwater divide (Figure 9). To test the influence of hydrodynamic dispersion, the values of tritium activities from the input function were modeled in the groundwater flow regime in the study area. The transport model used (Javandel et a1., 1984) is one- dimensional; it was modified to account for decay, retard- ation, hydrodynamic dispersion, and exchange of the solute (tritium). Exchange, in this case, refers to the additional input of nearly dead water, in the form of lake-effect snow, derived from Lake Michigan. Hydrodynamic dispersion also causes a solute front (in this case the first bomb-tritium from the autumn of 1952) to advance faster than the average groundwater seepage velocity. It was suspected, though, that the effect of dispersion would be minimal, for two reasons. First of all, hydrodynamic, or mechanical, dispersion is proportional to groundwater seepage velocity. The average downward velocity in the sampled area, 0.68 m/yr, in terms of meters per day, is 1.87 x 10E-3. This is relatively small in comparison to other field studies of dispersion (Sudicky et a1., 1983; 36 Fried, 1975). Secondly, dispersion is a scale-dependent parameter (Anderson, 1979); diSpersion coefficients tend to increase as a function of distance from the solute source. In a comprehensive review of dispersion coefficients meas- ured in field studies, Gelhar et a1. (1983) found that for distances of several to a few tens of meters, dispersion coefficients rarely exceed one meter, and are commonly on the order of centimeters. In Figure 14 are plotted two curves of simulated tritium activities. One assumes no dispersion and no exchange, the other assumes no exchange but uses a dispersivity of 10E-4 m2/day. When dividing this value of dispersivity by the seepage velocity of 1.87 x 10E-3 m/day, the dispersion coefficient is 0.05 m, or 5 cm, which is a 50 100 150 200 250 300 350 400 450 L 1 1 1 1 l 1 4 U‘ l D :10 m Iday /D20 01 L DEPTH BELOW WATER TABLE (m) M .1 O O 1 1 M U" I I f I I T T fi 50 100 150 200 250 300 350 400 450 TRITIUM CONCENTRATION (TU) F‘igure 14. Comparison of simulated tritium concentrations when D=0 and D=10E-4 m(2)/day. 37 realistic value for such a shallow penetration of the tritium front. As seen in Figure 14, there is little difference in the depth of the tritium front between the two curves. The other remaining possibility is that input of relatively dead water, from Lake Michigan, is entering the aquifer system during recharge events. The proposed addi- tional input is, of course, in the form of lak effect snowfall. The influence of the Great Lakes on increased snow depths along their lee borderlands is well documented (Eichenlaub, 1970). Chagnon et al. (1972) estimate that the lake effect increases winter precipitation in lee areas by as much as 45%, compared to areas outside of the lake effect. As the tritum activity of Lake Michigan is around 10 TU (Begemann and Libby, 1957), lake snow is thought to decrease values predicted by the input function. This is not accounted for by the input function, since the tritium values used are from Ottowa, which does not experience such a pronounced lake effect as the Leelanau Peninsula. Using a value of 10E-4 m2/day for dispersivity, 'a number of simulations were run, using values‘for exchange ranging from 50 to 75%. An exchange factor of 75% means that 25% of the input water is being replaced by dead water. The results are plotted in Figure 15, along with the measured values. It is observed that at an exchange value of about 55%, the maximum simulated value matches the maximum meas- ured activity of 136 TU. This is not considered to be an 38 59 100 1:130 200 250 l O O O O O O ' 0 o o I 5. : .. .. lMax. obs. . . . lvalue=136 TU o . o I O O O . | + ° ° ‘ 0 Ex. :o,55 10 O 0 Q I o o o . I .155 O : simulated Q : observed DEPTH BELOW WATER TABLE (M) A ~¢ . -—-- 25 O so 150 1gb 250 250 TRITIUM CONCENTRATION (TU) Figure 15. Comparison of maximum observed tritium concen- trations with simulated values, at varying solute exchange fractions. accurate simulation, however, because the field measurements give average values over the length of a well screen, while the 1-D transport model simulates values at points in depth, at one-meter intervals. To achieve a more realistic match, mean values for the observed groundwater column and the simulated values were compared. Since the highest depth any well screen in this study could have sampled formation water (assuming lateral flow into a well screen) is about 6 meters, no simulated samples above 6 meters were used for averaging simulated 39 tritium activites. The model, as it was run, gives simulated values at one-meter intervals. Field sample means, for a given depth corresponding to the depth of a simulated value, were calculated by averaging the measured activities of each sample whose well screen lies at the same depth as a simulated value. As all well screens are at least four feet long, the same well sample could be used in determining mean measured activities corresponding to two simulated values. The method for determining mean field sample activies is demonstrated in Table 2, which includes the profile of well coverage within the bomb-tritiated interval of the ground- water column. The mean value for all field measurements is then calculated by averaging all the means which correspond to depths at which values were simulated. Table 2 shows the plots of all tritium measurements, averaged measurements (corresponding to the depth of a sim- ulated activity) and simulated values using the exchange values ranging from 0 to 40%. Taking the total mean of the corresponding well sample activitiesas 74 TU, it is seen that the more realistic value for exchange is around 20%. CONCLUSIONS The geological, hydrological and analytical data pre- sented in this study, combined with the tritium and stable isotope input functions, permit the following conclusions to be drawn for the drift aquifer system underlying the study Table Simulated Depth Below Water Table (m) 2 7— 8—1 10--l 11-1 12- 13"I 14--I 15-r 16-l 17- 16- 19-- 20- 21- Comparison of average observed trations with average simulated values, 40 tritium concen- at vary- ing solute exchange fractions. Well Screen Covetage 22..., Average TU 40.1 73.0 64.0 119.6 60.4 Simulated TU Values With Vatiable Exchange JELHJQL.§&.£E&HAEL 41 37 33 29 25 41 37 33 29 24 45 40 36 31 2 7 52 47 42 36 31 65 59 52 45 39 79 71 63 55 47 93 63 74 65 55 127 115 102 69 76 193 174 154 135 116 244 219 195 171 146 150 135 120 105 60 19 17 15 13 11 11 10 6 7 6 69 60 71 62 53 41 area on the Leelanau Peninsula: 1) The aquifer system is completely heterogeneous; composed chiefly of a sandy stony till which is more clayey in the areas underlain by till plains than on the Manistee Moraine, and which contains many clay-rich zones. In the recharge area it is unconfined to semi-confined with depth. Confined conditions exist in the discharge areas, near the peripheries of Lake Leelanau and Grand Traverse Bay, which are underlain by lake clays. 2) Waters in the upper 90 meters ofthe aquifer system are less than 200 years old, based on an average vertical seepage rate of 22.6 m/ 33 yrs, as determined by average depth of the tritiiun interface. 3) Stable isotopic ratios also indicate that recharge is a combination of snowmelt and rainfall, being mostly snowmelt. The ratios also indicate that recharge occurs during the cooler months. 4) The lack of decrease in OD andmme values with depth supports the above conclusion that the sampled groundwater has all been recharged since the height of the Little Ice Age in the early 1700's. 5) Recharge rates in the study area are between 0.16 and 0.19 meters per year. The average recharge rate is 0.17 m/yr, or 17 cm. 6) An average recharge rate of 17 cm per year is prob- ably accurate for the Leelanau Peninsula, when considering the influence of relief on runoff during snowmelt over a 42 temporarily frozen soil surface. 7) Input of water to the system by lake effect snow results in a substantial decrease in the tritium activities of groundwater as compared to areas where a lake effect is not present. SUGGESTIONS FOR ADDITIONAL RESEARCH All sampling for isotOpic analyses was performed using existing well coverage in the study area. This distribution of sampling points is not sufficient to thoroughly charac- terize the inherent heterogeneity of the aquifer system. The drawback associated with this is best seen in Figures 11 and 12. In Figure 11 it is seen that different tritium concen- trations were measured at nearly the same depth in the groundwater profile. It is likely, given the heterogeneity of the aquifer, that this is a result of differing recharge rates across the peninsula, due to the variable nature of the drift. The same effect is observed in Figure 12, where at the same depth, different isotopic ratios are observed, for the same reason mentioned above. Based on the conclusion that the tritium interface method has been used successfully to determine an accurate recharge rate to the study area, the method could be best applied with well control which is concentrated in a few areas where different drift types are known to exist. With sufficient vertical well screen coverage at a few locations, accurate recharge rates could be determined as a function, 43 in part, of drift lithology. This is, in fact, the utility of the tritium interface method, as compared to the use of a hydrologic budget, groundwater models or basin analysis. The tritium method is most valuable for determining very site specific recharge rates, while the other more traditional method are suitable on a regional scale. APPENDICES APPENDIX A QUENCH MONITORING AND COUNTING EFFICIENCY Liquid scintillation counting of tritium decay events is an inefficient process. It is a function of several parameters, including the counting window, the tritium activity itself, the ratio of sample to cocktail in the counting vial and the type of scintillation cocktail used. It is rendered even more inefficient by the presence of foreign material in the sample/cocktail mixture. Any hinder- ance to the liquid scintillation detection process is known as a quenching effect. Without accounting for quenching, the counting rates obtained from the liquid scintillation spec- trometer may contain gross error. Quenching is monitored on the Beckman LS 8000 by what is known as the H-number (H#), which is the instrument's indirect mesure of efficiency. The H# is an eXpression of the extent of quench of a sample versus that of a cesium-137 standard internal to the scintillation spectrometer. The reader is referred to Long (1978) for a thorough discussion of the H# technique for quench monitoring. A quench curve, which gives counting efficiency as a function of H#, is created by calculating the counting efficiencies of a set of tritiated standards of known activ- ities and comparing them to the H-numbers generated by the LSC. The counting efficiency of a standard is defined as: CE = [ CPM(qs)/ml - CPM(b) J / DPMo(qs)/ml; II II 45 where: CPM(qs)/ml = counting rate of the quench standard CPM(b) = background counting rate DPMo(qs)/ml = actual activity of quench standard. The quench curve standards were prepared by adding varying amounts of carbon tetrachloride to the sample/cock- tail mixture. The actual mass of quenching agent was very small; only a few drops from a Pasteur pipet are necessary. Quench can Just as easily be manipulated by varying the amount of cocktail added to a standard. The quench curve used for calculating counting eficien- cies in this study is plotted in Figure 16. Five quench standards were counted eight times each, with a blank sample (dead water) accounting for background. 110 1go 130 ‘19 1§o iso 120 e s« ‘ \ 90 46.5 0 \ \CP EFF : -O.407(H at) + 11.26 6.0“ \ o .6023 QE\ > o ‘2’ LI] 5.5. o ‘ b o .555“; 80‘? t o \ W 5.0-1 0 \ Cb .500 08 E \ 1- 8, =5 4 5.. \o o 14.5 8 ox 010g: 5 v 1 j ' 110 120 150 140 150 160 170 H-NUMBER Figure 16. Quench curve for tritium counting window 170-225. Sample ESl 250A 406 188 220 223 407 E82 100 362 370 146 413 E83 60 248 296 371 E84 115 144 213 84 114 E85 77 410 58 216 368 H# 124. 132. 133. 139. 134. 135. 130. 114. 117. 136. 127. 123. 130. 110. 125. 117. 130. 127. 127. 120. 131. 123. 123. 129. 1140. 132. 129. 12s. 130. 126. 130. WNCNWNNWOWWNOOOWOONWWNNQNNWNONN APPENDIX B TRITIUM COUNTING Eff ONC‘U‘I O\O\C\UTC\O\U'IO\O\U1U1UTU1UTUTl-‘ HFJH UHfiv1ownUmencxoc>C2 .26 .80 .77 .49 .71 .66 .88 .53 .61 .00 .17 .86 .72 .10 .42 .89 .01 .76 .71 .75 .18 .92 :80 .92 :89 .03 .88 OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO .710 .434 .004 .356 .412 .291 .238 .694 .001 .356 .005 .031 .254 .085 .710 .076 .378 .883 .682 .708 .525 .800 .587 .594 .433 .703 .981 .229 .147 .863 -195 ...: NO e H WOWUTOGDNKO (13w ONO O‘iUT swan» O\UU 4:.“le CDO\O\\1 OJ: ...: 46 DATA OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO EE .00474 .00967 .01044 .00914 .00797 .00845 .00993 .00691 .00649 .01206 .01293 .01176 .00963 .00647 .00761 .00178 .00419 .00239 .00585 .00437 .00281 .00368 .00667 .00787 .00647 .00836 .00587 .00778 .00678 .00586 .01047 g e e e e e e e e HmtmmemmHOt-flmt 0‘ LA) NUTNU‘Il—‘tWU‘lU’T-tm Error ....1 \OHt-fithNW-fimmwwwm e e e e e e e e e e e e e e e e 9 U'IUUNthNO‘H—‘mwaomfiONSNIU'TWNNUTF-‘thwmw TAEJH unocnOMOC£~quq~qu>OHuovx O 0 9 9 O 9 O O O O O I O 0 U1 APPENDIX C TRITIUM ACTIVITY CALCULATION The formula used to calculate tritium activities from counting data is that used by Wyerman (1976). The counting parameters which are required for determining an activity include those obtained for electrolytic standards as well as the samples themselves. For each electrolytic standard the folloeing values must be known: a) initial volume (ml).....................Vo(es) b) final volume (ml).......................Vf(es) c) initial activity..... ............ .......DPMo(es)/ml d) counting rate.. ............ .............CPM(es) e) counting efficiency.....................CE(es) For each sample, the same data must be known, except for the initial activity, of course, which is what is being mesured. For the background sample (dead water), only the counting rate (CPM(b)) must be known. Calulation Tritium activities are expressed as disintegrations per minute per milliliter (DPM/ml) in the formula used by Wyerman (1976). Calculated values are given as: DPM/ml = [ CPM(s)/ml - CPM(b)] / (CE)(EE)(V/Vo); [1] where: 47 48 CE = counting efficiency of the sample EE = electrolytic efficiency of the sample V = mass of sample used during counting Vo = mass of sample before electrolysis CPM(s)/ml = counting rate of sample CPM(b) counting rate of background Electrolytic efficiency is a measure of the fraction of tritium remaining in the liquid phase of a sample upon completion of electrolysis. It is defined as: -(l/B) EE = (Vo/V) ; [23 where: B = fractionation factor of the sample. The fractionation factor for all samples within an electro- lytic set are assumed to to be equal, for B , in turn, is a function of the electrolytic efficiency of the electrolytic standard. The fractionation factor,.B, is defined as: B = l-log [ Vo(es)/V(es) ] / log EE(es); [3] where the electrolytic efficiency of the standard (of an electrolytic set), EE(es), is defined as: EE(es) = [ DPM(es)/ml ][ V(es) ] [ DPMo(es)/ml ][ Vo(es) J. [4] 49 The unknown in equation [4] is DPM(es)/ml, which is the final activity of the electrolyic standard. It is defined 8.8: DPM(es)/ml = [ CPM(es)/ml - CPM(b) ] / CE(es). The counting efficiency, CE, of the standard, as well that of individual samples, are calculated from the empiric- ally derived quench curve function previously discussed in Appendix A. To convert values of DPM/ml to tritium units (TU), the following relationship is used: 1 TU = 0.0071 DPM/ml. [SJ BIBLIOGRAPHY BIBLIOGRAPHY Anderson, M. P., 1979. Using models to simulate the movement of contaminants through groundwater flow systems. C.R.C. Critical Reviews in Environmental Control, 9:97—156. Andres, G. and Egger, R., 1985. A new tritium interface method for determining the rate of recharge of deep ground- water in the Bavarian Molasse Basin. Journal of Hydrology, Begeman, F. and Libby, W. F., 1957. Continental water bal- ance, groundwater inventory and storage times, surface ocean mixing rates and world-wide water circulation patterns from cosmic ray and bomb-tritium. Geochimica et Cosmochimica Acta, 12:277-296. Bernabo, J. C., 1981. Quantitative estimates of temperature changes over the last 2700 years in Michigan based on pollen data. Quaternary Research, 15:143-159. Bredenkamp, D., Schutte, J. and Dutoit, G., 1974. Recharge of a dolomitic aquifer as determined from tritium profiles, In Proceedings of a symposium, Vienna, International Atomic Energy Agency, p. 73-94. Brown, R. M., 1961. Hydrology of tritium in the Ottawa Valley. Geochimica et Cosmochimica Acta, 21:199-216. Chagnon, S. A. and Jones, D. M. A., 1972. Review of the influences of the Great Lakes on weather. Water Resources Research, 8:360-371. Clayton, R. N., 1961. Oxygen isotOpe fractionation between calcium carbonate and water. Journal of Chemistry and Physics, Clayton, R.N., Friedman, 1., Graf, D.L., Mayeda, T.K., Meents, W.F. and Shimp, N.F., 1966. The origin of saline formation waters. 1. Isotopic composition. Journal of Geo- physical Research, 71:3869-3882. Craig, H., 1961a. Isotopic variation in meteoric waters. Science, 133:1702—1703. Craig, H., 1961b. Standard for reporting concentration of deuterium and oxygen-18 in natural waters. Science, 133:1883-1884. rn 51 Dansgaard, w., 1964. Stable isotopes in precipitation. Tellus, 16:436-468. Delcore, M.R., 1985. Rate of recharge to a heterogeneous drift aquifer: a test of the bomb-tritium method. M.S. thesis, Michigan State University. Delcore, M.R. and Larson, G.J., 1987. Application of the tritium interface methodfor determining recharge rates to unconfined drift aquifers. II. Non-homogeneous case. Journal of Hydrology (in press). Desaulniers, D.E., Cherry, J.A. and Fritz, P., 1981. Origin, age and movement of pore water in argillaceous Quaternary deposits at four sites in southwestern Ontario. Journal of Hydrology, 68:231-257. Dincer, T., Al-Mugrin, A. and Zimmerman, U., 1974. Study of the infiltration and recharge through the sand dunes in arid zones with special refernce to the stable isotopes and thermonuclesr tritium. Journal of Hydrology, 23:79-109. Dunne, T. and Black, R.D., 1971. Runoff processes during snowmelt. Water Resources Research, 7:1160-1172. Egboka, B.C.E., Cherry, J.A., Farvolden, R.N. and Frind, E.0., 1983. Migration of contaminants in groundwater at a landfill: a case study. 3. Tritium as an indicator of dispersion and recharge. Journal of Hydrology, 63,51-80. Eichenlaub, V., 1970. Lake effect snowfall to the lee of the Great Lakes: its role in Michigan. American Meteorological Society Bulletin, 51:403-412. Evenson, E.B., Farrand, W.R., Eschman, D.E., Mickelson, D.M. and Maher, L.J., 1976. Greatlakean Substage - a replacement for Valderan Substage in the Lake Michigan basin. Quaternary Research, 6:411-24. Ferronsky, V.I. and Polyakov, V.A., 1982. Environmental isotopes in the hydrosphere. Translated by S.V. Ferronsky. Wiley, New York, 466 pp. . Florkowski, T., 1981. Low-level tritium assay in water samples by electrolytic enrichment and liquid scintillation counting in the International Atomic Energy Agency labor- atory. In: Methods of low level counting and spectrometry. Proceedings of a symposium, International Atomic Energy Agency, 9- 335-351. Fried, J.J., 1975. Groundwater pollution. Elsevier Scientific, 330 p. 52 Friedman, I., Redfield, ~A.C., Schoen, B. and Harris, J., 1964. The variation of the deuterium content of natural waters in the hydrologic cycle. Review of GeOphysics, 2:177- 2214. Gelhar, L.W., Mantoglou, A., Welty, C. and Rehfelt, K.R., 1985. A review of field scale physical solute transport processes in saturated and unsaturated porous media. Rpt. RP-2485-05, Elect. Power Res. Institute, Palo Alto, Calif. Graf, D.L., Friedman, I. and Meents, w.F., 1965. The origin of saline formation waters. 2. Isotopic fractionation by shale micrOpore systems. Illinois Geologic Survey Circular 393, p. 32. Hitchon, B. and Friedman, I., 1969. Geochemistry and origin of formation waters in the western Canada sedimentary basin. 1. Stable isotopes of hydrogen and oxygen. Geochimica et Cosmochimica Acta, 33:1321-1349. International Atomic Energy Agency, 1953-1986. Environmental isotope data nos. 1, 2, 3, 4, 5, 6, 7. Worldwide survey of isotope concentrations in precipitation. Technical report series 96, 117, 129, 147, 165, 192, 260. Javandel, I., Doughty, C. and Tsang, 1984. Groundwater transport: Handbook of applied models. American Geophysical Union Water Resources Monograph No. 10. 228 p. Kondoh, A., 1986. Study on the groundwater flow system by environmental tritium in Ichihara region, Chiba prefecture. Environmental Research Center Bulletin, Ibaraki, Japan. Lamb, H.H., 1966. The changing climate. Methuen, London. Larson, G.J., Delcore, M.R. amd Offer, S.A., 1987. Appli- cation of the tritium interface method for determining recharge rates to unconfined drift aquifers. I. Homogeneous case. Long. D.L., Rezabek, D.H., Takacs, M.J. and Wilson, T.P., 1986. Geochemistry of groundwater in Bay County, Michigan. Michigan Dept. of Public Health Research Rpt. ORD38553, 3 volumes. Long, E., 1978. Liquid scintillation counting theory and techniques. Beckman Instruments, Irvine, California. Maloszewski, P., Rauert, W., Stichler, W. and Herrman, A., 1983. Application of flow models in an alpine catchment area using tritium and deuterium data. Journal of Hydrology, 66:319-330. 53 Martin, H., 1957. Outline of the Geologic History of the Grand Traverse Region. Michigan Dept. of Conservation, Geological Survey Division Publication 49. Nir, A., 1964. On the interpretation of tritium "age" measurements of groundwater. Journal of Geophysical Research, 69: 2589-2595. Nurnberger, F., State Climatologist, 1953 - 1985. Precip- itation records for Northport - Sutton's Bay station. Michigan Dept. of Agriculture, Climatology Division, Michigan State University. Nurnberger, F., State Climatologist, 1982. Summary of evapo- ration in Michigan. Michigan Dept. of Agriculture, Climatology Division, Michigan State University. Offer, S.A., 1982. Determination of recharge rates to a drift aquifer using bomb-tritium in the saturated zone. M.S. thesis, Michigan State University. Ogata, A., 1970. Theory of dispersion in a granular medium. U.S. Geological Survey Professional Paper 411—A. Petrie, M.A., 1984. Morphologic and Lithologic influences on recharge in a glaciated basin. M.S. thesis, Michigan State University. Price, A.G. and Hendrie, L.K., 1983. Water motion in a deciduous forest during snowmelt. Journal of Hydrology, 64:339-356. Rabinowitz, D., Gross, G. and Holmes, C., 1977b. Environmen- tal tritium as a hydrometeorologic tool in the Roswell Basin, New Mexico. III. Hydrologic parameters. Journal of Hydrology, 32:35-46. Sklash, M., Mason, 8., Scott, S. and Pugsley, C., 1986. An investigation of the quantity, quality and sources of groundwater seepage into the St. Clair River near Sarnia, Ontario, Canada. Water Poll. Res. J. Canada, 21:351-367. Smith, D., Wearn, P., Rowe, P. and Richards, H., 1970. Water movement in the unsaturated zone of high and low permea- bility strata by measuring natural tritium. In: IsotOpe Hydrology; proceedings of a symposium, Vienna, International Atomic Energy Agency, p. 73-85. Sudicky, E.A., Cherry, J.A. and Frind, E.0., 1983. Migration of contaminants in groundwater at a landfill: a case study. 4. A natural-gradient dispersion test. Journal of Hydrology, 63:83-108. 54 Sukhija, B.S. and Shaw, C.R., 1976. Conformity of groun- dwater recharge by tritium method and mathematical modeling. Journal of Hydrology, 30:167-178. Thatcher, L.L., Janzer, V.J. and Edwards, K.W., 1977. Methods for determination of radioactive substances in water and fluvial sediments, Chapter A5. In: Techniques of water- resources investigations of the U.S. Geological Survey. Toth, J., 1962. A theory of groundwater motion in small drainage basins in central Alberta, Canada. Journal of Geophysical Research, 67:4375-4387. Toth, J. 1963. A theoretical analysis of groundwater flow in small drainage basins. Journal of GeOphysical Research, 68:4795-4811. U.S. Geological Survey, Water Resources Division, Lansing, Michigan. Streamflow data for Michigan, 1953-1985. Weber, H.L., 1973. Soil surey of Leelanau County, Michigan. U.S. Dept. of Agriculture, Soil Conservation Service. Webb, T. and Clark, D.R., 1977. Calibrating micropaleon— tological data.irxclimatic terms: A critical review. Annals of the New York Academy of Science, 288:93-118. Williams, L.D. and Wigley, T.M.L., 1983. A comparison of evidence for late Holocene summer temperature variations in the northern Hemisphere. Quaternary Research, 20:286-307. Wurzel, P., 1983. Updated radioisotope studies in Zimbabwean ground waters. Groundwater, 21:597-605. Wyerman, T., 1976. Laboratory facility of the analysis of natural levels of tritium in water. U.S.G.S. Open File Report. Zimmerman, U., Munnich, K.O. and Roether, W., 1967. Downward movement of soil moisture by means of hydrogen isotopes. In: Isotope techniques in the hydrologic cycle. American Geophysical Union Monograph No. 11: 28-36. "711111111111111115