ISGS CONTRACT/GRANT REPORT: 1983-6 • UILU-WRC-83-0180 557.09773 [L6cr 1983-6 Hydrogeologic and Geochemical Studies of Selected Natural Radioisotopes and Barium in Groundwater in Illinois Robert H. Gilkeson Keros Cartwright Illinois State Geological Survey Champaign, Illinois James B. Cowart Florida State University Tallahassee, Florida Richard B. Holtzman Argonne National Laboratory Argonne, Illinois May 1983 University of Illinois WATER RESOURCES CENTER Urbana-Champaign, Illinois Illinois Department of Energy and Natural Resources STATE GEOLOGICAL SURVEY DIVISION Champaign, Illinois Gilkeson, Robert H. Hydrogeologic and geochemical studies of selected natural radioisotopes and barium in groundwater in Illinois / Robert H. Gilkeson and others. - Urbana- Champaign, III. : University of Illinois. Water Resources Center, May 1983. 93 p. : ill. ; 28 cm. - (UILU-WRC-83-0180) (Illinois- Geological Survey. Contract/Grant report; 1983-6) 1. Water, Underground— Illinois. 2. Water, Underground- Analysis. I. Title. II. Seriesl. III. Series 2. Printed by authority of the State of Illinois/1983/1000 HYDROGEOLOGIC AND GEOCHEMICAL STUDIES OF SELECTED NATURAL RADIOISOTOPES AND BARIUM IN GROUNDWATER IN ILLINOIS Robert H. Gilkeson Associate Geologist Hydrogeology and Geophysics Section Illinois State Geological Survey Champaign, Illinois 61820 Keros Cartwright Geologist and Head Hydrogeology and Geophysics Section Illinois State Geological Survey Champaign, Illinois 61820 James B. Cowart Assistant Professor Geology Department Florida State University Tallahassee, Florida 32306 Richard B. Holtzman Chemist Radiological and Environmental Research Division Center for Human Radiobiology Argonne National Laboratory Argonne, Illinois 60439 FINAL TECHNICAL COMPLETION REPORT Project No. B-108-ILL The work on which this report is based was supported in part by funds provided by the United States Department of the Interior as authorized under the Water Research and Development Act of 1978, CONTRACT /GRANT REPORT: 1983-6 Illinois Department of Energy and Natural Resources State Geological Survey Division 615 East Peabody Drive Champaign, Illinois 61820 CONTENTS ABSTRACT 1 ACKNOWLEDGMENTS 2 INTRODUCTION 4 HYDROGEOLOGIC SETTING AND MAJOR IONS IN GROUNDWATER 6 NATURAL RADIOISOTOPES IN GROUNDWATER 19 BACKGROUND 19 RADIUM ISOTOPES IN GROUNDWATER 23 The Relation Between Gross Alpha Radiation and Radium-226 in Groundwater 26 Regional Concentrations of Radium-226 and Radium-228 in Groundwater 32 Mechanisms That Control the Concentration of Radium Isotopes in Groundwater 40 RADON-222 IN GROUNDWATER 51 URANIUM ISOTOPES IN GROUNDWATER 54 THORIUM ISOTOPES IN GROUNDWATER 60 BARIUM IN GROUNDWATER 62 BACKGROUND 62 BARIUM IN GROUNDWATER FROM BEDROCK OF THE CAMBRIAN AND ORDOVICIAN SYSTEMS IN NORTHEASTERN ILLINOIS 63 CASE HISTORIES— PUMP FAILURES DUE TO SECONDARY PRECIPI- TATION OF BARIUM SULFATE 72 BARIUM IN GROUNDWATER AT LOCATIONS OUTSIDE NORTHEASTERN ILLINOIS 81 SUMMARY 84 NATURAL RADIOISOTOPES IN GROUNDWATER 84 BARIUM IN GROUNDWATER 86 REFERENCES 89 i i Digitized by the Internet Archive in 2012 with funding from University of Illinois Urbana-Champaign http://archive.org/details/hydrogeologicgeo19836gilk TABLES 1. Stratigraphic units of the Ordovician System in northeastern Illinois that were dewatered in 1980 14 2. Gross alpha radiation and radium-226 in groundwater for repeated measurements at selected wells 28 3. Uranium-234 + uranium-238, radium-226 and gross alpha radiation in groundwater from selected wells 31 iii FIGURES 1 . Map of the study area 7 2. Generalized column of rock-stratigraphic units in northern Illinois 8 3. The relation between Eh and the concentration of total dissolved sulfides in groundwater from the Cambrian and Ordovician bedrock 10 4. The potentiometric surface in a region of northeastern Illinois for wells that produce groundwater from the Ironton-Galesville Sandstone and overlying Cambrian and Ordovician rocks 12 5 . The radioactive decay series of uranium-238 20 6. The radioactive decay series of thorium-232 21 7. The low correlation coefficient between the concentrations of radium-226 and radium-228 in groundwater from the Cambrian and Ordovician bedrock 33 8. The relation between the concentration of radium-226 and total dissolved minerals in groundwater from the Cambrian and Ordovician bedrock 46 9. The low correlation coefficient between the concentrations of radium-226 and radon-222 in groundwater from the Cambrian and Ordovician bedrock - 53 10. The relation between Eh and the concentration of uranium in groundwater from the Cambrian and Ordovician bedrock. The ^ 3t +U/ 238 U activity ratio is shown for each analysis 56 11. The distribution of dissolved sulfate in groundwater from the Cambrian and Ordovician bedrock for a region of northeastern Illinois 64 12. The distribution of dissolved barium in groundwater from the Cambrian and Ordovician bedrock for a region of northeastern Illinois 65 13. The relation of the activities of (Ba 2+ ) and (SO^ 2- ) in groundwater from the Cambrian and Ordovician bedrock to the solubility of BaS0 4 70 iv ABSTRACT Dissolved concentrations of the natural radioisotopes 222 Rn, 226 Ra, 228 Ra, 230 Th, 232 Th, 234 U and 238 U and the element Ba 2+ were investigated in a study of high concentrations of 226 Ra, 228 Ra, and Ba 2+ in groundwater from the Cambrian and Ordovician bedrock of northern Illinois. The high radium and barium concentrations are naturally present in the major aquifers — the sandstone bedrock; therefore, reme- dial well construction measures will not lower the concentrations. The combined concentrations of 226 Ra and 228 Ra range from 2.3 to 50.2 pCi/L; the majority of analyses exceed the limit in the drinking water regula- tions of 5.0 pCi/L. The 226 Ra/ 228 Ra activity ratio ranges from 0.2 to 41.0; a 226 Ra analysis has no validity for predicting the 228 Ra concen- tration. Important controls on dissolved 226 Ra concentrations are secondary U on the sandstone matrix and the ionic strength of ground- water. The distribution of accessory minerals that contain 232 Th is believed to be an important control on dissolved 228 Ra concentrations. Locally the dissolved 226 Ra and 228 Ra concentrations are affected by dissolution-precipitation of secondary barite. Dissolved concentrations of 2 Rn and U are less than proposed drinking water regulations. The extreme enrichment in 234 U that occurs widespread in groundwater from the Cambrian and Ordovician bedrock is unique and problematic. The enrichment may reflect recharge of uranium to the bedrock by glacial processes. Dissolved Ba 2 concentrations range from <0.01 to 22.4 mg/L. The high Ba + concentrations occur in groundwater that is depleted in dissolved SO4 by anaerobic microbial reactions. A map presents the distribution of Ba + in groundwater from the Cambrian and Ordovician bedrock in northeastern Illinois. Gilkeson, Robert H., Keros Cartwright, James B. Cowart, and Richard B. Holtzman HYDROGEOLOGIC AND GEOCHEMICAL STUDIES OF SELECTED NATURAL RADIOISOTOPES AND BARIUM IN GROUNDWATER IN ILLINOIS Final Technical Completion Report to Bureau of Reclamation, U.S. Depart- ment of the Interior, May, 1983, Washington, DC 20240, 93 p. KEYWORDS — aquifer proper ties /barite /barium/ chemical proper ties/ glacial events /groundwater /radium/ radon /sedimentary rocks /sulfate /uranium/ thorium -2- ACKNOWLEDGMENTS The authors of this report are deeply grateful to the superin- tendents of public and industrial water supplies in Illinois and their staffs for their cooperation and assistance, which was invaluable to our research. Among those who merit special recognition for their support are Ronald E. Zegers, Superintendent of the Elgin Water Depart- ment, and two members of his staff, Scott Brenner and Keith Kijawa; and Rick Dime, Superintendent of the Or land Park Water Department. The authors thank the staff of the analytical laboratories of the Illinois State Water Survey and the Illinois Environmental Protec- tion Agency for performing chemical analyses of groundwater samples for this research. The efforts of Kenni James and James Whitney of the Illinois State Water Survey and Dorothy Bennett of the Illinois Environmental Protection Agency are especially appreciated. Robert Sasman of the Illinois State Water Survey was a valuable source of Information on groundwater supplies in northeastern Illinois. Curtis R. Benson of the Illinois State Water Survey assisted in the collection of groundwater samples from wells in northeastern Illinois. Dorothy Woller of the Illinois State Water Survey did an outstanding job of compiling information on municipal wells in Illinois. Many staff members at the Illinois State Geological Survey pro- vided assistance in this research. Dorothy C. Stouffel and Susan S. Wickham were instrumental in initiating the research program. Randall E. Hanson, Elizabeth Elischer, Kenneth Potter, William Constantine, James B. Risatti, Paul E. Heigold, and Hildegard 0. Mine provided technical assistance . Ross Brower of the Illinois State Geological Survey assisted in the design and construction of a large-capacity downhole sampler that was used to collect groundwater samples from selected boreholes. Dennis Coleman of the Illinois State Geological Survey assisted in the design and construction of a flow cell for taking measurements of the Eh, pH and temperature of groundwater. -3- Howard M. Prichard of the School of Public Health, The University of Texas at Houston, performed most of the radon-222 analyses. Henry lucas of the Argonne National Laboratory provided historical analyses on radium-226 and radium-228 in groundwater. The high disequilibrium between uranium-234 and uranium-238 in groundwater from the Cambrian and Ordovician bedrock in some localities of Illinois was first noted by J. K. Osmond, Geology Department, Florida State University, Tallahassee, Florida. The authors also appreciate the administrative assistance of Glenn E. Stout, Director of the Water Resources Center, University of Illinois, and Donald E. Perrero, Manager, Departmental Business, Insti- tute for Environmental Studies, University of Illinois. -4- INTRODUCTION This report presents findings from a study of the geochemical mechanisms responsible for the occurrence of two isotopes of the radio- active element radium (radium-226 in the uranium-238 decay series and radium-228 in the thorium-232 decay series) and the element barium in groundwater of Illinois at concentrations that exceed the limits imposed in the National Interim Primary Drinking Water Regulations (U.S. Environ- mental Protection Agency, 1975). The barium is not radioactive. The high radium and barium concentrations are largely restricted to ground- water from bedrock of the Cambrian and Ordovician Systems in northern Illinois. Radium and barium are both alkaline-earth cations with simi- lar chemical properties. The occurrence of these elements in ground- water is discussed in separate sections of this report. The present research has generated several reports. A report by Gilkeson et al. (1978) presented analyses of rock samples for barium and radioactive elements and discussed general mechanisms important to the occurrence of barium and radium in groundwater. Perry, Grundl , and Gilkeson (1982) reported on the distribution of stable isotopes of oxygen, hydrogen, and sulfate in groundwater from the Cambrian and Ordovician bedrock. Gilkeson, Perry, and Cartwright (1981) discussed evidence that the anomalous depletion of sulfate in groundwater with high barium concentrations was due to bacterially mediated chemical reactions. Gilkeson and Cowart (1982) described the extreme enrichment in uranium-234 in groundwater from the Cambrian and Ordovician bedrock. This report presents analyses for chemical constituents, including barium and radium, for raw groundwater samples from wells in communi- ties of northern Illinois. It is important to note that the concen- trations of radium and barium in the analyses may not reflect the con- centrations of these constituents in the finished water supplied to the consumer. The efficient removal of radium and barium by two con- ventional water treatment processes (ion exchange softening and lime softening) was determined in a study by Bennett, Bell, and Markwood (1976). Several communities use these treatment processes. -5- Many communities in northern Illinois produce groundwater from several wells that are finished in different aquifers. In general, groundwater produced from aquifers other than the Cambrian and Ordovician bedrock has very low concentrations of radium and barium. The blending of groundwater from these other aquifers with groundwater from the Cambrian and Ordovician bedrock may result in a finished groundwater that is in compliance with the drinking water regulations. Surface water supplies also have very low concentrations of radium and barium. Chicago and many other communities are supplied water from Lake Michigan. The City of Elgin in northeastern Illinois has recently completed con- struction of a facility to treat water from the Fox River. Several other communities in northern Illinois use rivers or reservoirs for water supplies. -6- HYDROGEOLOGIC SETTING AND MAJOR IONS IN GROUNDWATER Public groundwater supplies are produced from the Cambrian and Ordovician bedrock in a fairly extensive area of northern Illinois (fig. 1) . South of the study area water quality in the Cambrian and Ordovician is undesirable because of the high concentrations of dis- solved minerals in the groundwater. The radioactive element section of this report is concerned with the entire area shown in figure 1. The barium section of the report is primarily concerned with the Cambrian and Ordovician bedrock in restricted localities in the northeastern and northwestern regions of figure 1. A generalized column of rock-stratigraphic units in northern Illinois is shown in figure 2. The Cambrian and Ordovician bedrock is composed of thick sections of sandstone and dolomite interbedded with shale. In downward order, the stratigraphic units are the Maquoketa Shale Group, Galena and Platteville Dolomite Groups, Glenwood-St. Peter Sandstone and Prairie du Chien Group of Ordovician age, the Eminence Formation, Potosi Dolomite, Franconia Formation, Ironton-Galesville Sandstone, Eau Claire Formation, and Mt . Simon Sandstone. Shale in the upper part of the Eau Claire Formation is a significant aquitard hydraulically separating the Ironton-Galesville Sandstone from the Elm- hurst Sandstone Member (basal Eau Claire) and the Mt . Simon Sandstone. On a regional scale the significant aquifers are the sandstones, including the St. Peter, the Ironton-Galesville, the Elmhurst Member in the basal Eau Claire Formation, and the Mt . Simon Sandstones. The Ironton-Galesville Sandstone is the most productive aquifer over a major part of the study area. Production of groundwater from wells is often increased through the detonation of explosives in the Ironton-Galesville section of the open borehole. The yield of groundwater from the dolomite bedrock units is variable, depending upon the number and areal interconnections of joints, fractures, and crevices. The contribution of groundwater from the shale bedrock to the open borehole is generally small. -7- . MSCONSIN I STEPHFNSON I wiNNEBAGi" ISGS 1983 Maquoketa Shale Group absent -older Ordovician rocks 1 form the bedrock surface, except in a locality along the south side of the Sandwich Fault Zone where Cambrian rocks form the bedrock surface Maquoketa Shale Group absent -shale in the Pennsylvanian System unconformably overlies and confines Ordovician rocks ' Maquoketa Shale Group forms the bedrock surface and • confines underlying Ordovician rocks Ordovician and Cambrian units overlain and confined by a thick section of bedrock including shales of the Maquo- keta, Devonian, Mississippian, and Pennsylvanian FIGURE 1. Geologic map of the area of northern Illinois where groundwater supplies are produced from the Cambrian and Ordovician bedrock. Sources of information are Willman et al. (1967) and Willman et al. (1975). -8- Tertiary and Cretaceous Pleistocene GROUP OR FORMATION LOG A; >."*'■ •J THICKNESS (m) DESCRIPTION Unconsolidated glacial deposits - pebbly clay (till), silt, and gravel: loess (windblown silt), and alluvial silt, sand, and gravels. Sand and silt. Ji § ^^^■^ Mainly shale with thin sandstone, limestone and coal. ' . I St. Louis Salem Warsaw Keokuk- Burlington Kinderhookian Limestone; cherty limestone; green, brown and black shale; sllty dolo- ml te. / / Shale, calc thin. z^z Niagaran Alexandrian Kankakee Bdgewood iz Dolomite, silty at base, locally cherty. £^ Ctnclnnatlan '— ' Shale, gray or brown; locally dolo- mite and/or limestone, argillaceous. B rT: Plattevi He Champ lainian S Dolomite and/or limestone, cherty! Dolomite, ahale partings, speckled; Dolomite and/or limestone, cherty, sandy at base. Sandstone, fine- and coarse-grained; . little dolomite; thaln at toy. Sandstone, fine- to medium-grained) locally cherty red shale at base. Shakopee New Richmond Eminence Potosi TZJEF Proviso Lombard Elmhurst rh~-' Ht. Simon Dolomite, sandy, cherty (oolitic), sands tone. Sandstone, lnterbedded with dolomite. Dolomite, white to pink, coarse- grained, cherty (oolitic), sandy at base. Dolomite, white, flno-grained , geodlc quartz, sandy at base. Dolomite, sandstone, and shale, glauconltic, green to red, mlcaceoiiH Sandstone, fine- to medlum-qrained, well sorted, upper part dolomltic. Shale and siltstone; dolomite, glauconi tic; sandstone, dolomi tic glauconittc. Sandstone, coarse-grained, white, red in lower half; lenses of shall* and siltstone, red, micaceous. Precambrian FIGURE 2. Generalized column of rock-stratigraphic units in northern Illinois. -9- The oxidation-reduction potential (Eh) of groundwater in the Cambrian and Ordovician bedrock was investigated through measurements on groundwater in an airtight flow chamber that contained two platinum electrodes coupled to a calomel reference electrode. The measurement technique is discussed by Back and Barnes (1965). In the present study, the potentials were monitored using two separate platinum elec- trodes as a check for spurious values. The pH of groundwater was also measured in the airtight flow chamber. Whitfield (1974) performed thermodynamic analyses to quanti- tatively define the chemical reactions that constrain the measurement of Eh with the platinum electrode. In well-oxygenated aqueous systems the platinum metal is sensitive to formation of platinum oxide coatings. In anaerobic aqueous systems that contain sulfide ion, the metal is sensitive to formation of sulfide coatings. The calculations by Whit- field (1974) indicate that for the pH range of groundwater in the Cam- brian and Ordovician bedrock the platinum electrode would respond as an inert metal over the approximate range in potential of +300 to millivolts. Figure 3 presents the relation of Eh values measured with the platinum electrode to the total sulfide concentration in groundwater at wells that produce from the Cambrian and Orodovician bedrock. The range in Eh is from +400 to -180 millivolts. The values greater than +300 millivolts were measured in primary recharge zones where the Cambrian and Ordovician bedrock is unconfined. The high values may be due to the presence of dissolved oxygen in the groundwater. The total concentration of dissolved sulfide species in ground- water was measured at wells with the titrimetric (iodine) method (Stan- dard Methods, 1971). The method has a lower limit of detection of 0.1 mg/L. At many wells, sulfide was detected by odor at concentrations too small to be measured. Figure 3 shows that the potentials measured with the platinum electrodes were sensitive to the concentration of dissolved sulfide present in the groundwater. Sulfide was not present in groundwater samples with Eh values greater than 100 millivolts and was rarely present at concentrations greater than 0.1 mg/L in groundwater -10- 400- Well sites Sulfide not present in groundwater O Sulfide present at concentration less than 0.1 mg/L Ml Sulfide present at measured concentration 300- 200- AA A A A ^ 2.o 3!o Zo ?o ao" total dissolved sulfide (mg/L) 7.0 FIGURE 3. The relation between Eh (platinum electrode) and the concentration of total dissolved sulfides in groundwater from the Cambrian and Ordovician bedrock. -11- with Eh greater than +50 millivolts. Sulfide concentrations greater than 1.0 mg/L poised the Eh of groundwater in a range from -50 to -100 milli- volts; these Eh values may not accurately represent the oxidation- reduction potential of the groundwater. The groundwater pumped from the majority of wells is a mixture of groundwater produced from several stratigraphic units open to the well bore. The sulfide concentration and the oxidation-reduction potential in groundwater produced from each of the discrete units may be very different. A general finding for the eastern part of the study area is that where sulfide is present in groundwater, the high concentrations are in the Galena and Platteville Groups and the Glenwood-St. Peter Sandstone. The lowest Eh value of -180 millivolts was measured at a well where the concentration of sulfide in groundwater was less than 0.1 mg/L; a concentration too low to poise the potentials measured with the platinum electrodes. The relation between the platinum electrode values of Eh and the dissolved sulfide concentrations in the groundwater is consistent with the relation ob- served by Berner (1963) for natural and artificial sulf ide-bearing sediments . Figure 1 shows that the Maquoketa Shale Group is absent over much of the north-central and northwestern part of the study area, and that older rocks form the bedrock surface. In the northwestern region the Maquoketa Shale Group is present on some uplands and confines under- lying rocks. In some localities where the Maquoketa is absent, fine- grained glacial drift and/or dense unfractured dolomite of the Galena and Platteville Groups serve as confining units. The terrane setting is an important control on the configuration of groundwater flow systems in the north-central and northwestern re- gions. Uplands are zones of recharge to the Cambrian and Ordovician bedrock, and river valleys are zones of discharge. Water level measure- ments have defined a north-south groundwater divide through the eastern part of the region (fig. 4) . A portion of the groundwater recharge that occurs eastward of the divide enters a regional groundwater flow system in the northeastern part of the study area. ■12- Lake Michigan WATER LEVEL MEASUREMENTS October 1 980 Area where Maquoketa Shale Group is absent ^40(X. Mean sea level elevations in feet Present regional groundwater divide --■0.— Regional groundwater divide in natural flow system before large withdrawals ' — ^ Approximate predevelopment groundwater flow path FIGURE 4. The potentiometric surface in a region of northeastern Illinois for wells that produce groundwater from the Ironton-Galesville Sandstone and overlying Cambrian and Ordovician rocks. Water level measurements were taken in October 1980. The figure also illustrates the effect of the large groundwater withdrawals on groundwater flow in the regional system (modified from Sasman et al., 1982). -13- In the north-central and northwestern regions groundwater is produced from rock-stratigraphic units from the Galena through the Mt. Simon Sandstone. The concentration of total dissolved minerals in the groundwater ranges from approximately 250 mg/L to 600 mg/L. Ground- water with concentrations of total dissolved minerals of less than 400 mg/L is produced from many wells across this part of the study area that are drilled deep into the Mt . Simon. The pH of groundwater pro- duced from wells ranges from 7.0 to 7.5. The major cations in the ground- water are calcium and magnesium and the major anion is bicarbonate. The concentration of sulfate is generally less than 30 mg/L. The Eh of groundwater measured with platinum electrodes ranged from +400 to +15 millivolts. The highest values were measured at wells located in pri- mary recharge zones and may reflect the presence of dissolved oxygen in the groundwater. The lowest values were measured at wells where the aquifer units are confined. Hydrogen sulfide was often present in the confined groundwater, but generally at concentrations less than 0.1 mg/L. In the northeastern part of the study area, the Cambrian and Ordo- vician bedrock is confined by the Maquoketa Shale Group. When the first wells were drilled in the 1860s, the artesian pressure in the Cambrian and Ordovician bedrock was sufficient to cause wells to flow in the Chicago-Joliet-Fox River valley area. As a result of large withdrawals of groundwater, the potentiometric surface over the period from 1864 to 1980 has declined more than 900 feet in localities in northern Cook and northeastern Du Page Counties (Visocky, 1982) . By 1980 substantial de- watering of the upper stratigraphic units in the Ordovician bedrock had occurred in major pumping centers (table 1) . Figure 4 shows the potentio- metric surface in the fall of 1980 for wells in northeastern Illinois that produce groundwater from the Ironton-Galesville Sandstone and overlying stratigraphic units in the Cambrian and Ordovician System. The -100-foot contour outlines the major pumping centers. Groundwater flow is from all directions toward these centers. The shape of the po- tentiometric surface indicates that groundwater transmitted to these pumping centers is primarily from the west and north and to a lesser extent from the south and east. -14- TABLE 1 Stratigraphic Units of the Ordovician System in Northeastern Illinois That Were Dewatered in 1980 (Modified from Visocky, 1982) Pumping Center Aurora More than 2/3 of the Galena and Platteville Groups Elgin More than 2/3 of the Galena and Platteville Groups E. Du Page-W. Cook Co. All of the Galena and Platteville Groups and less than 1/4 of the Glenwood-St. Peter Sandstone Joliet All of the Galena and Platteville Groups and nearly 1/4 of the Glenwood-St. Peter Sandstone N. Cook Co. All of the Galena and Platteville Groups and more than 1/3 of the Glenwood-St. Peter Sandstone Most of the wells in the northeastern region have open boreholes in the stratigraphic interval from the Galena Group through the Ironton-Galesville Sandstone. Some wells are also open to the Eau Claire Formation and the Mt . Simon Sandstone. Care has to be taken in completing wells in the Mt . Simon because of the presence of highly mineralized groundwater at depths greater than several hundred feet into the sandstone. The large withdrawals of groundwater have re- sulted in the Ironton-Galesville Sandstone, the unit with the highest transmissivity, having a lower hydraulic head than the overlying and underlying bedrock. During periods when a well is not pumped, a cross- formational flow of groundwater occurs in the open borehole in response to differences in hydraulic head between the formations. The lower head in the Ironton-Galesville results in recharge of groundwater to this stratigraphic unit from overlying bedrock in the open borehole. If the well bore is also open to the underlying Eau Claire or to the Mt . Simon, then groundwater will flow up the open borehole from these stratigraphic units and out into the Ironton-Galesville during periods of nonpumpage. The exact configuration of the potentiometric surface of the Cambrian and Ordovician before extensive groundwater withdrawal is unknown. Interpretation of data from early studies indicates that under -15- natural conditions the surface was relatively featureless and sloped gently to the southeast (Suter et al., 1959). Terrane features such as highlands on glacial moraines and major drainageways such as the Fox, Des Plaines, Kankakee, and Illinois Rivers must have had a significant effect on hydraulic gradients of the natural potentiometric surface. The present location of the regional groundwater divide (fig. 4) along the western side of Boone and De Kalb Counties is a result of the change in hydraulic pressure caused by the large withdrawals of groundwater from the Cambrian and Ordovician bedrock. Estimated flow lines for the regional groundwater flow in bedrock of the Cambrian and Ordovician of northeastern Illinois before the large withdrawals of groundwater are shown on figure 4. The flow lines are based on an in- terpretation from the earliest water level measurements (Suter et al., 1959) and from consideration of regional terrane relations. The flow lines are intended only as an approximation of groundwater movement in the regional flow system and do not show the effects due to anisotropy or heterogeneity in the rock strata. The axis of the regional groundwater divide in the natural flow system trended north to south through McHenry County and the northwestern part of Kane County. The north to south axis also extends north of McHenry County across Walworth County, Wisconsin. In Kane County the axis of the divide turned westward across De Kalb County. The high topographic elevations along the Marengo Moraine are an important con- trol of the position of the regional groundwater divide in McHenry and Kane Counties in Illinois, and Walworth County in Wisconsin. The end moraine was constructed in Woodfordian time (late Wisconsinan) during the first major glacial event following the period of cool climatic con- ditions and weak soil development that occurred during the Farmdalian Substage (22,000 to 20,000 years before the present) (Kempton and Gross, 1971). Pre-Woodfordian terrane relations (mapped by S . S. Wickham and W. H. Johnson, 1981) indicate that prior to construction of the Marengo Moraine the natural regional groundwater divide was located farther west approximately at the position of the present divide. During this -16- earlier time (before construction of the Marengo Moraine) primary re- charge to the regional flow system of northeastern Illinois occurred over a large area of Boone and northern De Kalb Counties where the Ma- quoketa Group is absent. However, after construction of the Marengo Moraine recharge to the regional flow system no longer occurred in this area of Boone and De Kalb Counties and recharge was limited to the slow infiltration of groundwater through a thick sequence of fine-grained geologic materials including shale of the Maquoketa and glacial till. In the northeastern region the concentration of total dissolved minerals in groundwater ranges from approximately 250 mg/L to 2050 mg/L. The lowest concentrations are present along the western part of the region; mineralization increases to the east and southeast along the direction of groundwater flow. The pH of groundwater produced from wells in the region varies from 7.0 to 7.5. Calcium, magnesium, and bi- carbonate are the dominant ions in the dilute groundwaters. The anoma- lous depletion of sulfate ion in groundwater in the northwestern part of the region is discussed in the barium section of this report. Down- gradient from the depleted zone, sulfate concentrations increase rapidly (fig. 11) and are present at concentrations greater than 800 mg/L in southeastern Cook County. The concentrations of sodium and chloride are also significant in the mineralized groundwater. Hydrogen sulfide is commonly present in groundwater from wells within the zone of sulfate depletion and ranges in concentration from <0.1 mg/L to 6.0 mg/L. At these well sites the high sulfide concentra- tions are present in the groundwater from the Galena and Platteville Groups and the Glenwood-St. Peter Sandstone. Eh values for these wells range from +50 to -100 millivolts. Downgradient from the anaerobic environment there is a rapid increase in the Eh of groundwater, with values ranging from +150 to +300 millivolts occurring in Du Page, Cook, and Lake Counties. Over a major part of the southeastern region of the study area (fig. 1, east and south of the Illinois River and south of the Kankakee River) the Cambrian and Ordovician bedrock is confined by the Pennsyl- vanian System and/or the Maquoketa Group. The Pennsylvanian is absent in the eastern part of the region. -17- The Maquoketa Group is absent because of erosion in the north- central part of the region over an area that includes the western part of Grundy County and all but the southwestern part of La Salle County. In most of the area where the Maquoketa is eroded, the Cambrian and Ordo- vician bedrock are confined by shales of the Pennsylvanian System. An exception is a restricted area in southern La Salle County (fig. 1) where the Pennsylvanian is eroded and the Galena and Platteville or the Glenwood-St. Peter form the bedrock surface beneath the glacial drift. Groundwater in the Cambrian and Ordovician bedrock in the south- eastern region is highly mineralized within a short distance of con- finement. Only along the northern part of this region are wells con- structed to produce groundwater from the Ironton-Galesville . No wells in the region produce groundwater from the Eau Claire or the Mt . Simon. In this region the concentration of total dissolved minerals in ground- water from the Cambrian and Ordovician bedrock ranges from approximately 300 mg/L to 1500 mg/L. The pH of groundwater produced from wells ranges from 7.1 to 7.6. Sodium and chloride are the major ions in the more mineralized groundwater. At some wells the chloride concentration ex- ceeds 500 mg/L. Other ions that occur at appreciable concentrations in the groundwater are calcium, magnesium, sulfate, and bicarbonate. The groundwater at most wells contains hydrogen sulfide with the concentra- tion of total dissolved sulfides ranging from <0.1 mg/L to 4.5 mg/L. Eh measurements with the platinum electrode method vary from +30 to -180 millivolts. In the southwestern part of the study area (the region west of the Illinois River on fig. 1) the Cambrian and Ordovician bedrock is confined by a thick section of bedrock including shale of the Maquoketa, Mississippian, and Pennsylvanian. The Pennsylvanian forms the bedrock surface except where absent along the western margin. The Cambrian and Ordovician bedrock in the southwestern region is part of a large regional groundwater flow system that has zones of primary recharge in northwestern Iowa (Horick and Steinhilber, 1978). Most wells in the southwestern region are constructed to produce groundwater from stratigraphic units from the Galena through the -18- Glenwood-St. Peter. Only In the larger communities are the wells con- structed to also produce groundwater from the Ironton-Galesville. No wells in the region produce groundwater from the Eau Claire or the Mt. Simon. Groundwater in the Cambrian and Ordovician bedrock in the south- western region is highly mineralized. The concentration of total dis- solved minerals in groundwater ranges from approximately 1,000 mg/L in the northern part of the area to concentrations greater than 3,000 mg/L at a few wells in the southern part. The pH of groundwater produced from wells ranges from 7.1 to 7.5. The major ions in the groundwater are calcium, magnesium, sodium, bicarbonate, sulfate, and chloride. At many wells the concentration of dissolved sulfate exceeds 1,000 mg/L. Groundwater from wells in the southern part of the area commonly has fluoride concentrations that exceed 2.0 mg/L, the limit set in the drinking water regulation (U.S. EPA, 1975). Dissolved sulfide species are ubiquitous in groundwater from the Cambrian and Ordovician bedrock in the southwestern region; concen- trations range from less than 0.1 to 0.9 mg/L. The groundwater has a low Eh (values for platinum electrode measurements range from -10 to -50 millivolts) . -19- NATURAL RADIOISOTOPES IN GROUNDWATER BACKGROUND The study of geochemical mechanisms that control the occurrence of radium-226 and radium-228 in groundwater requires knowledge of the distribution of other radioactive elements in the decay series. This report presents information on the occurrence of the following isotopes in groundwater: uranium-238 ( 238 U), uranium-234 ( 23tf U), thorium-230 ( 230 Th), radium-226 ( 226 Ra) , and radon-222 ( 222 Rn) in the 238 U decay series and thorium-232 ( 232 Th), radium-228 ( 228 Ra) , and radium-224 ( 22t+ Ra) in the z32 Th decay series. The information presented on the occurrence of z32 Th, 230 Th and 22t+ Ra is limited. The naturally occurring long-lived radioactive elements 238 U (99.3 percent of the natural uranium) and 32 Th (effectively 100 percent of the natural thorium) are the parents of the radioactive decay series shown in figures 5 and 6. The nuclides in the two decay series, along with the long-lived radioactive element, 1+0 K (potassium-40) , account for practically all of the natural environmental radiation of terrestrial origin. 235 U (uranium-235) is the long-lived parent nuclide of the actinium decay series. Because of the very small amount of 235 U relative to 238 U, the natural environmental concentration of radioisotopes in the actinium decay series is insignificant. The rate of disintegration of a radioactive element is constant and is commonly expressed as the half-life, the period of time required for the disintegration of half the atoms in a sample of a specific radioactive element. As figures 5 and 6 illustrate, the half -lives for specific radioactive elements represent an extreme range of values. The long-lived parent nuclides of the two decay series were formed in a primary stellar nuclear synthesis process and were in existence at the time of formation of the earth; their abundances in the earth have been decreasing through geologic time because of their radioactivity. The daughter nuclides in the 238 U and 232 Th decay series are too short- lived to be remnants of the earth's formation. The nuclides are re- generated in nature by decay of the parent nuclides. -20- 1.17m ( 238 U) Uranium 238* 4.5 x 10 9 y** ( 234 U) Uranium -234* ( 234 Pa) Protactinium -234* 24.1 d ( 234 Th) Thorium -234* 26.8 m - , 2.48 x 10 s y ( 230 Th) Thorium -230* i 7.7x10 4 y ( 226 Ra) Radium-226* < 1.60x10 3 y ( 222 Rn) Radon -222 < 3.82 d ( 218 Po) Polonium -218 > 3.05 m ( 214 Pb) Lead -21 4* ( 214 Bi) Bismuth-214* 19.8 m ( 214 Po) Polonium-214 5.01 d > f 1.6x1(T'i ( 210 Po) Polonium-210* ( 210 Bi) Bismuth 210 22.3 y ( 210p b) Lead-210* > 138.4 d t ( 206 Pb) Lead -206 (Stable) Note: Vertical direction represents alpha decay; horizontal direction represents beta decay. * Also gamma emitters ** Times shown are half- lives : y = years; d = days; h = hours; m = minutes; s = seconds ISGS 1983 FIGURE 5. The radioactive decay series of uranium-238. -21- ( 228 Th) 6.13 h Thorium -228* 1 1.91 y i ( 224 Ra) Radium-224* ■ 3.64 d ( 220 Rn) Radon-220 1 55 s ( 216 Po) Polonium-216 i 0.16 s ( 212 Pb) 10.64 h Lead-212* ( 228 Ac) Actinium -228* 5.75 y ( 232 Th) Thorium-232* 1.41 x 10 lo y ( 228 Ra) Radium-228 ( 212 Bi) Bismuth-212* 60.5 m -64%— ► ( 212 Po) Polonium-212 36% 1 60.5 m ' 0.30 x 10' 6 s ( 208 TI) Thallium-208* 3.10 m (208p b) Lead -208 (Stable) Note: Vertical direction represents alpha decay; horizontal direction represents beta decay. * Also gamma emitters ** Times shown are half -lives : y = years; d = days; h = hours; m = minutes, s = seconds ISGS 1963 FIGURE 6. The radioactive decay series of thorium-232. -22- The radioactivity of an unstable isotope is by definition A = NX (1) where A is activity in disintegrations per unit time, N is the number of atoms in the system and A is the decay constant — the fraction of atoms decaying per unit time. In a closed system in which radioactive elements are neither added or removed except by radioactive decay, a state of secular equilibrium will eventually be achieved between radio- active parents and daughters. At equilibrium the activities of parent and daughter nuclides are equal. Although the activities are the same at equilibrium, the mass concentrations of the parent-daughter nuclides are greatly different. For example, at secular equilibrium, the mass ratio of 231+ U to 238 U is 0.000056. In open systems, such as groundwater flow systems, secular equili- brium between parent and daughter nuclides is seldom found because of different mobilities due to the chemical properties of the radioactive elements. The disequilibrium that is measured between two radioactive elements of a decay series is commonly expressed as the ratio of their activities. At equilibrium the activity ratio equals 1. The disintegration of a radioactive nuclide is accompanied by the release of energy in the form of alpha, beta, or gamma radiation. Alpha and beta are forms of particulate radiation; gamma radiation is composed of photons (extremely minute packets of radiant energy) . An alpha particle is composed of the nucleus of a helium atom (the two orbiting electrons are missing) that is emitted from the nucleus of the decaying nuclide at a high velocity. Ultimately, all alpha particles emitted from radioactive elements find two electrons in the environment and become helium gas. A beta particle is an electron emitted at a high velocity from a decaying nuclide. The decay series in figures 5 and 6 present information on the type of radiation emitted by specific radioactive elements. The curie, named in honor of Madame Marie Curie, is the unit used as the standard rate of disintegration. Originally, the curie was defined as the number of disintegrations occurring per second from one gram of pure 226 Ra. Because this number could change over time because of increased counting accuracy or improvements in purifying 226 Ra -23- samples, the decision was made to define the curie at a stable standard value that was close to, but not tied to, the decay rate for 226 Ra. The curie is now defined as that quantity of radionuclide that is dis- integrating at a rate of 3.7 x 10 10 disintegrations per second. Re- fined measurements have determined the activity of one gram of 6 Ra to be 3.61 x 10 10 disintegrations per second (0.98 curies). The units used in this report to express the content of many of the radioactive elements in water are picocuries per liter (pCi/L) . One picocurie is equal to 1 x 10~ 12 curies or 2.22 disintegrations per minute. Radiation produces damaging effects in living tissue through a process called ionization. Ionizing radiation introduces energies to living tissue that are huge in comparison with ordinary chemical reac- tions. The radiation has the potential to remove electrons from atoms that comprise living tissue. These electrons have a great capacity to break chemical bonds as they travel through living tissue causing the re- lease of other electrons. The atoms of living tissue that lose electrons become ions with such a high state of energy that they can produce chemi- cal reactions that would never have been possible without the ionizing radiation. All three radiations (alpha, beta, and gamma) have the potential to cause ionizations. High-energy gamma radiation may pass through living tissue with few ionizations. The efficiency of alpha and beta particles for inducing ionizations is related to the charge on the particle and the velocity of the particle. The higher the charge and the more slowly a particle is travelling, the greater is the chance for ionization. Therefore, the highly charged, slow moving alpha particles have the greatest potential to cause ionizations. RADIUM ISOTOPES IN GROUNDWATER Four natural isotopes of radium may occur in groundwater: 6 Ra (1600-year half-life) in the 238 U decay series, 228 Ra (5.75-year half- life) and 22t+ Ra (3.64-day half-life) in the 232 Th decay series, and 223 Ra (11.4-day half-life) in the 235 U decay series. The immediate parent nuclides of all four natural radium isotopes are isotopes of the element thorium. -24- The concentrations of 223 Ra present in the natural environment are insignificant because of the very slight concentrations of its long-lived parent, 235 U. No analyses for 223 Ra were performed in this research. No drinking water regulation relative to 2 Ra has been promul- gated. There is little public health concern for this isotope because of its short half -life of 3.64 days. The source for 2 4 Ra in groundwater is 228 Th that is present on the aquifer matrix; the concentration of 228 Th in groundwater is insignificant. Groundwater pumped from a well is removed from the parent thorium nuclide, and depending upon the resi- dence time of the water in the distribution system before consumption, a significant reduction of the ' 24 Ra concentration may occur because of the short half-life. Also, because of its short half-life, 22l+ Ra will not accumulate in the human skeleton, but a portion of the 221+ Ra that is ingested will be present on the surface of bones, delivering alpha radia- tion to sensitive tissue. The longer lived radium isotopes ( 226 Ra and 228 Ra) have residence in the interior of bones where tissue is less sensitive to ionizing radiation. The short half-life of 22i+ Ra presents difficulty to the design of an effective monitoring program for a large number .of water supply wells. Analyses for this isotope in groundwater of Illinois are re- stricted to a limited number of measurements reported by Stehney (1955). In groundwater samples collected in 1951 and 1952 at eight wells in northern Illinois that produce groundwater from the Cambrian and Ordo- vician bedrock, the concentration of 22Lf Ra ranged from 4.9-7.3 pCi/L. The values are similar in magnitude to the concentration of 228 Ra measured in groundwater at well sites in the same localities. The relation between Ra and Ra concentrations requires further investi- gation. If a close relation is found, then 228 Ra measurements may be an indicator of 224 Ra concentrations. If the values for 224 Ra and 228 Ra are found to be similar, then the activity concentration of 22t+ Ra in potable groundwater from the Cambrian and Ordovician bedrock will be expected to range from approximately 1.0 to 20 pCi/L. Activity concen- trations of this magnitude will not be of concern for public health. -25- Of the radioactive elements present in groundwater, 226 Ra and 228 Ra are of most concern with respect to public health. Radium is often de- scribed as a "bone seeker." Because the human body metabolizes radium and calcium similarly, the continuous ingestion of trace concentrations of 226 Ra and 228 Ra over a long period of time will result in an accumula- tion of radium in the skeleton. After deposition in bone tissue, radium leaves the body very slowly. The daughter nuclides of 226 Ra and 228 Ra have short half-lives. Therefore, the disintegration of either radium nuclide initiates a series of disintegration events over a short inter- val of time. As shown in figures 5 and 6, many of the short-lived daughter nuclides emit alpha particles upon disintegration. The high specific activity (alpha-particle emission per unit weight) of radium is an important concern of the nuclides to public health. The incidence of bone-cancers related to ingestion of large quantities of 226 Ra and 228 Ra is well known (Argonne National Laboratory, 1981). However, the relation between chronic ingestion of low concen- trations of radium and health risk is not well established. Nevertheless, the U.S. Environmental Protection Agency (1975) promulgated the following National Interim Primary Drinking Water Regulations concerning radio- activity in drinking water: 1. The average annual gross alpha contaminant level shall not exceed 15 pCi/L. 2. The combined 6 Ra and/or 228 Ra contaminant level shall not exceed 5 pCi/L. The interim regulations allow substitution of a gross alpha con- taminant level of 5 pCi/L or less for the radium analyses and do not require a ' 8 Ra analysis unless the 226 Ra activity concentration exceeds 3 pCi/L. A weakness of these provisions is that in the absence of alpha emitters, the contaminant level of 228 Ra present is not determined. U.S. Environmental Protection Agency recognized this deficiency and therefore provided for augmented monitoring by individual states. The National Interim Primary Drinking Water Regulations recommend that in localities where 228 Ra may be present in drinking water, the states re- quire D Ra and ZZ0 Ra analyses when the gross alpha particle activity exceeds 2 pCi/L. Compliance with this recommendation would require -26- 226 Ra and 228 Ra analyses at nearly all well sites in Illinois where groundwater is produced from the Cambrian and Ordovician bedrock. 9 9 A The presence of high activity concentrations of ^ D Ra in ground- water from bedrock of the Cambrian and Ordovician Systems was reported in papers by Stehney (1955), Lucas and Ilcewicz (1958), Emrich and Lucas (1963), and Holtzman (1964). Analyses for both 226 Ra and 228 Ra in groundwater from bedrock of the Cambrian and Ordovician were presented in reports by Krause (1959, 1960). The analytical data from these historical studies have been incorporated into this report. In addition, 9 9ft this report incorporates previously unpublished analyses on Ra in groundwater provided by Dr. Henry Lucas of the Argonne National Laboratory. In the present research, groundwater samples were collected from carefully selected wells distributed across the study area. Unfiltered and filtered (0.22 micron pore size membrane) samples were collected at the well head, and preserved with nitric acid or EDTA. The radon emanation method of Lucas (1977) was used for analyses of 26 Ra. The analytical procedure for 228 Ra is described in Krieger and Whittaker (1980). The Relation Between Gross Alpha Radiation and Radium-226 in Groundwater The Illinois Environmental Protection Agency performed analyses for gross alpha and gross beta radiation on untreated groundwater samples from municipal wells in Illinois; most of the analyses were performed from 1971 to 1976. Over the period of study the analyses were repeated several times for most wells. The gross radiation analysis was a useful screening method to identify regions where high concentrations of radio- active elements may occur in groundwater. Potable groundwater with a very low gross alpha concentration will also have a very low 226 Ra con- centration. For wells with low gross alpha radiation concentrations (<3 pC/L) , all repeated analyses over the 6-year study measured low concentrations. However, at wells where gross alpha concentrations of 5 pCi/L or greater were measured, concentrations found by repeated analyses were variable. The most variation occurred in groundwater samples with the highest concentrations. A general trend indicated in the gross alpha analyses for many wells was an increase in concentration -27- in the later analyses. Therefore, a concern developed that over time the groundwater was increasing in radioactivity. An objective of the present research was to identify the mechanisms responsible for the variation over time in the gross alpha concentration. Table 2 presents gross alpha radiation analyses and 226 Ra analyses for selected wells where historical analyses on 6 Ra were available. The gross alpha radiation analyses were performed by the Illinois En- vironmental Protection Agency. All of the 226 Ra analyses were performed at the Argonne National Laboratory. A large variation in gross alpha radiation values was found for most of the wells, although for most wells the variation in 226 Ra concentrations is small or insignificant. On the basis of a few analyses performed by the Illinois Environmental Protection Agency, Gilkeson et al. (1978) reported in a preliminary study that filtration (0.22 micron pore size membrane) had a significant effect on reducing the gross alpha radiation in groundwater having high values. Two examples of this apparent effect are shown in table 2 for the duplicate samples collected at municipal well Lynwood-2 (7/27/77) and at municipal well Lostant-A (7/25/77). The lower alpha radiation values in each set were for filtered groundwater. Actually, filtration does not have significant effects on the concentration of gross alpha radia- tion: the variation in gross alpha radiation concentrations at the well sites in table 2 was primarily due either to the method of analysis or to analytical error. The method for analysis of gross alpha radia- tion and the implication of the analysis to the concentration of 226 Ra in groundwater are discussed in the following paragraphs. When radioactive equilibrium with respect to 226 Ra is attained in the sample prepared for gross alpha particle analysis four alpha particles are emitted for every 226 Ra atom that decays. This is illu- strated in the 238 U decay series presented in figure 5, starting with o o r Ra. The vertical downward sequence indicates that a nuclide disinte- grates with emission of an alpha particle while the horizontal sequence indicates that a nuclide disintegrates with release of beta radiation. As illustrated in the figure, the 22b Ra decays to 222 Rn (radon-222) , an inert gas with a 3.82-day half -life, which in turn decays to 2i8 Po (polonium-218) , and then to 21t+ Pb (lead-214) . At that point two beta -28- TABLE 2 Gross Alpha Radiation and Radium-226 in Groundwater for Repeated Measurements at Selected Wells County Collection Date Gross Alpha Radiation 226 Ra Owner, Well No. (pCi/L) (P< :i/L) Cook 12-05-72 52 + 15 _ Lynwood-2 03-27-75° 149 + 27 14.7 04-02-75° 156 + 26 14.4 04-10-75° 58 + 16 14.9 10-08-75 07-27-77^ 07-27-77 d 78 + 19 - 140 + 20 - 74 + 17 - 08-04-79 - 15.6 Kane 03-21-74 13 + 4 _ Elgin-6 03-07-75° 55 + 7 13.3 12-05-79 - 13.6 06-03-80 - 11.5 La Salle 1958 6 _ 9.9 Lostant-4 08-05-72 24 + 7 - 05-05-75 34 + 10 - 07-25-77 56 + 12 - 07-25-77 12 + 6 - 12-04-79 - 10.6 Livingston 1958 6 _ 11.2 0dell-3 10-05-72 33 + 10 - 06-02-75 24 + 8 — 08-21-80 - 10.7 Peoria 1958 e _ 8.2 Elmwood-3 02-27-73 32 + 10 _ 10-08-75 39 + 11 - 03-28-76 43 + 12 - 11-05-80 — 7.9 a. b. c. d. e. Gross alpha radiation analyses performed by the Illinois Environ- mental Protection Agency. The coefficient of variation is shown for each analytical value. o n r Ra analyses performed by Argonne National Laboratory, coefficient of variation — ± 3%. Analytical data from Bennett, Bell, and Markwood (1976). Analytical data from Gilkeson et al. (1978). Analytical data from Krause (1960). -29- decays form 211+ Po (polonium-214) , which then decays by alpha emission to 2l0 Pb (lead-210) . For purposes of this analysis the series ends with 210 Pb because the nuclide has a long half-life (22.3 years) and thus decays very slowly. In the analysis of gross alpha radiation, a measured quantity of water is evaporated to dryness, forming a thin residue of the dissolved constituents on a counting planchet. The evaporation is done at high heat to drive off 222 Rn gas present in the water. If all of the z22 Rn were driven off and a period of 4 hours elapsed between preparation of the planchet and counting of alpha activity, then none of the short- lived daughter products 218 Po, 21t+ Pb, 21Lf Bi and 211+ Po would be present in the residue, since the longest half -life in the series is only 26.8 minutes. Thus, if the sample is counted 3 to 4 hours after preparation, only 6 Ra will be present at a significant concentration for the series 6 Ra to 210 Pb, and only one alpha particle is emitted for each 6 Ra atom decaying. The 222 Rn grows into the residue with a 3.82-day half- life; therefore, after about 4 days the 222 Rn would build up to one-half its maximum activity by the equation **. " V- 6 ^ (2) where A^ is the activity of 222 Rn, A is activity of the 226 Ra, e is the base of the natural logarithms with a value of 2.7183, A is the radioactive constant of the isotope 222 Rn (for 222 Rn A = 0.693/3.82 = 0.1814 day ) , and t is the time since the evaporation. When the decay products of 222 Rn reach equilibrium with 222 Rn three alpha particles will be emitted for each atom of 222 Rn that decays; one from the 222 Rn and two from its daughters, 218 Po and 214 Po. Con- sequently, if we assume that all of the 222 Rn is retained in the residual film of the evaporated sample (actually, some of the 222 Rn will escape from the sample) and that the sample has one disintegration per minute o o c of * D Ra, then one alpha particle per minute will be emitted immediately after preparation. After 4 days about 3 alpha particles will be emitted per minute, one from the 226 Ra and two from the 222 Rn series; after 40 days (10 half-lives of the 222 Rn when the value of e ' ' in equation (2) becomes insignificant) , the 222 Rn will be in radioactive equilibrium -30- with the 22b Ra and there will be 4 alpha particles emitted for each 226 Ra nuclide that decays. Additionally, a portion of the alpha parti- cle activity may not be detected because the particles are absorbed by the matrix of the residue. Table 3 presents repeated analysis for gross alpha particle activity for eight groundwater samples collected from wells that produce groundwater from stratigraphic units in the Cambrian and Ordovician Systems. Analyses were performed at Argonne National Laboratory with the interim radiochemical procedure recommended by the U.S. Environmental Protection Agency (Krieger, 1976). Unfiltered and filtered samples of groundwater were collected at each well. Table 3 also presents analyses for 226 Ra, 238 U, and 23t+ U in the groundwater samples. All of the gross alpha radiation analyses show increase in alpha activity over time due to the grow-in of 222 Rn and the 222 Rn daughters in the residue on the counting planchets. For most of the samples the later gross alpha radiation analyses approached values that were between 3 and 4 times greater than the 226 Ra concentrations. Filtration did not have significant effects on either the 226 Ra concentrations or the gross alpha radiation concentrations; the differences noted on most samples are within the limits of the analytical method. Howeyer, the 226 Ra con- centrations were slightly lower in filtered samples from municipal wells Orland Park-6, Homewood-12, and Lockport-4. The combined activity concentrations of 23t+ U and 238 U in the groundwater samples ranged from 0.02 to 2.8 pCi/L. 238 U and 23k \J disintegrate through emission of alpha particles. Therefore, the gross alpha radiation concentrations presented in table 3 include a contribution from 234 U and 238 U that is equivalent to the concentrations of these nuclides reported in the table. In the procedure recommended by the U.S. Environmental Protection Agency for analyses of gross alpha radiation in drinking water (Krieger, 1976) there are no specifications concerning the storage time of the planchet between preparation and counting. The procedure does not account for the grow-in of the short-lived daughter nuclides of 226 Ra into the residue on the counting planchet. It is obvious from the above discussion that a large range in the measured concentration of gross alpha radiation may occur in planchets with appreciable concentrations -31- TABLE 3 Uranium-234 + Uranium-2 38, Radium-226 and Gross Alpha Radiation in Groundwater from Selected Wells County Owner- Well No, 234 U + 238 U * 26 Ra b Unfil- tered Fil- tered (pCi/L) (pCi/L) (pCi/L) Gross Alpha Radiation Planchet Unfil- Planchet Fil- residence tered residence tered time (days) (pCi/L) time (days) (pCi/L) Cook, Or- land Park-6 2.8 8.1 Cook, Home- wood-12 2.2 3.4 Cook, Lyn- wood-2 2.7 15.6 Fulton, Ta- ble Grove-1 0.4 McDonough , Macomb-2 Will, Lock- port-4 15.0 Kane, Elgin-6 0.02 12.4 Kendall, Yorkville-4 0.6 7.9 0.34 18.7 2.7 10.7 6.8 2.3 15.6 15.2 13.0 8.1 20.6 9.4 4 13.3 4 7.3 12 11.1 12 7.9 20 11.7 20 8.6 10.2 4 6.7 6 12.1 12 11.3 14 15.6 20 8.5 2 37.1 30.5 14 48.4 7 49.7 22 48.6 14 52.9 21.9 29.3 12 59.1 12 57.6 14 55.4 14 64.5 1 27.8 2 36.3 13 30.8 14 39.7 21 32.7 22 44.2 2 16.6 17.6 14 17.8 7 18.6 22 18.3 18 21.7 21 20.6 4 68.8 2 35.5 12 77.4 14 68.9 20 80.0 22 83.5 4 20.9 1 15.4 12 22.8 13 24.1 20 23.7 15 24.3 Radium-226 and gross alpha radiation analyses were performed on unfil- tered and filtered samples. The data illustrate the variation in gross alpha radiation concentration that occurs as a function of counting planchet residence time. 5 Ra analyses performed by Argonne National Laboratory; coefficient of variation — ± 3%. Gross alpha radiation analyses performed at Argonne National Laboratory with analytical procedure given in Krieger (1976); coefficient of variation — ± 30%. Samples were filtered at the well (0.22 micron pore size membrane) before the addition of preservatives. -32- of 225 Ra that are counted on a random time basis after preparation. A significant part of the variation in the Illinois Environmental Protection Agency data set of gross alpha radiation analysis of water samples from specific wells over the period of 1971 to 1976 may be due to differences in storage time between preparation and counting of the planchets. However, in some of the analyses, the concentration of gross alpha radiation is significantly greater than the equilibrium value four times greater than the 226 Ra concentration. For example, three analyses for municipal well no. 2 at Lynwood (table 2) show a range of 140 to 156 pCi/L. These high values are difficult to explain and may represent laboratory error. Gross alpha radiation analysis is obviously a valuable tool for rapidly screening the radioactivity of a water sample. Samples with low concentrations of gross alpha radiation will also have low concen- trations of 226 Ra and other nuclides such as 231+ U, that emit alpha radia- tion. However, for water samples that contain high concentrations of gross alpha radiation, no accurate finding can be made concerning the concentration of 2 6 Ra present. Because of the inaccuracy of the ana- lytical method of performing the gross alpha radiation measurements, the existing data base of the Illinois Environmental Protection Agency cannot be used to investigate changes in radioactivity of the groundwater that may have occurred over the 6-year period during which the analyses were performed. Regional Concentrations of Radium-226 and Radium-228 in Groundwater In the analyses presented in this report on samples of potable groundwater from the Cambrian and Ordovician bedrock, the concentration of 226 Ra ranges from 0.1 pCi/L to a maximum value of 36.9 pCi/L, and the concentration of 228 Ra ranges from 0.5 pCi/L to a maximum value of 32.7 pCi/L. Figure 7 presents the relation of 225 Ra and 228 Ra concen- trations in groundwater from 90 wells that produce groundwater from the Cambrian and Ordovician bedrock at wide scattered locations throughout the study area. For these wells the combined concentrations of the two nuclides range from 2.3 to 50.2 pCi/L. Only 4 of the combined analyses were in compliance with the maximum limit of the drinking -33- 15 228 Ra(pCi/L) FIGURE 7. The low correlation coefficient between the concentration of radium-226 and radium-228 in groundwater from the Cambrian and Ordovician bedrock. For a least square line fit to the analyses, the correlation coefficient, r, = 0.275. Isolines show the combined concentration of the two nuclides and the I26 Ra/"*Ra activity ratio. -34- water regulations of 5.0 pCi/L for the combined concentration of 226 Ra and 228 Ra. The combined concentrations were 10.0 pCi/L or greater in 55 of the analyses. The ratio of 226 Ra/ 228 Ra for the 90 combined analyses ranges from 0.2 to 41.0. The large range in values indicates that the 228 Ra con- centration in groundwater cannot be predicted from a 226 Ra analysis. A least square line fit to the analyses determined a low correlation coefficient, r, of 0.275. In the analyses presented in this report, concentrations of 22b Ra of 5.0 pCi/L or lower were usually accompanied by concentrations of 228 Ra of 5.0 pCi/L or lower. However, analyses of groundwater from the Cambrian and Ordovician bedrock in Illinois, Iowa, Missouri, and Wisconsin by Lucas (1982) determined that high 228 Ra con- centrations were occasionally present in groundwater with low "Ra concentrations. Analyses of groundwater samples in South Carolina by Michel and Moore (1980) and in Finland by Asikainen (1981) also found a low correlation between the concentrations of 22(:) Ra and 228 Ra. Data available from analyses on 226 Ra and 228 Ra are too limited to be used to determine accurately the regional occurrence of the nu- clides in potable groundwater from the Cambrian and Ordovician bedrock. However, some general trends in concentrations are apparent. Combined concentrations of 22b Ra and 228 Ra in groundwater from the Cambrian and Ordivician bedrock exceeding the maximum contaminant level of 5.0 pCi/L are widespread in Illinois. The lowest combined concentrations for the two nuclides occur in groundwater samples from well sites in the north- central and northwestern regions of the study area in zones of primary recharge where the Maquoketa Group is absent and groundwater has had a short residence time in the Cambrian and Ordovician bedrock. The lowest °Ra concentrations in groundwater from the Cambrian and Ordovician bedrock (values of less than 2.0 pCi/L) are in zones of primary recharge in the north-central and northwestern regions of the study area where the Maquoketa Shale Group is absent. Downgradient of the primary recharge zones, the 226 Ra concentration in groundwater in- creases; however, for wells in the north-central and northwestern regions where the Maquoketa Shale Group is absent, the values seldom exceed 5.0 pCi/L with most values less than 4.0 pCi/L. Higher 226 Ra -35- concentrations occur in groundwater where the Cambrian and Ordovician bedrock are confined with the highest concentrations occurring in the mineralized groundwater zones. In the southwestern, northwestern and north-central regions of the study area the Ra concentrations in groundwater seldom exceed 3.0 pCi/L; the majority of the concentrations are less than 2.0 pCi/L. The highest concentrations of Ra occur in groundwater from wells lo- cated in the eastern part of the study area. Areas where the 228 Ra con- centrations commonly range from 5.0 to 8.0 pCi/L are along the Fox River valley in Kane County, the southern part of Cook County, Will County, Grundy County, the southern part of La Salle County, and the eastern part of Marshall County. At a few wells in Kane, Cook, and Will Counties the 228 Ra concentrations range from 10 to 19 pCi/L. Comparison of the values determined in previous radium analyses of groundwater with the values determined in the present research in- dicates that no significant variation has occurred over time in the 228 Ra or 228 Ra concentrations in groundwater from the Cambrian and Ordovician bedrock. The analyses indicate that the radium concentra- tion in native groundwater produced from a discrete stratigraphic unit at a specific well has remained essentially constant, but that at a specific well site the discrete stratigraphic units open to the well bore may each produce groundwater with markedly different radium concen- trations. The analyses also established that in groundwater from the Cambrian and Ordovician bedrock the high radium concentrations occur in groundwater from the significant water producing strata — the sandstones. Analytical data for the eastern part of the study area indicate that the highest 226 Ra and 228 Ra concentrations in potable groundwater are from the regionally most productive stratigraphic unit, the Ironton- Galesville Sandstone. Specific examples are presented in the following discussion of municipal wells at Elgin in northeastern Kane County, Orland Park in southern Cook County, Coal City and Morris in Grundy County, and Peru in La Salle County. -36- Municipal wells at Elgin are constructed with open boreholes to receive groundwater from a large section of bedrock. All of the wells are open to stratigraphic units from the Galena and Platteville Dolomite Groups downward through the Ironton-Galesville Sandstone. In addition, several wells at Elgin are also open to the Eau Claire Forma- tion and the upper section of the Mt . Simon Sandstone. The wells at Elgin that produce groundwater with the highest dissolved radium con- centrations are those that were developed with explosives to increase production of groundwater from the Ironton-Galesville. Examples are Elgin-6, with 22b Ra concentrations ranging from 11.5 to 13.6 pCi/L and a 228 Ra concentration of 13.4 pCi/L, and Elgin 5A, with 226 Ra concentra- tions ranging from 7.1 to 8.8 pCi/L. Lower radium concentrations occur in groundwater from municipal wells at Elgin that were not developed with explosives and therefore produce a smaller component of groundwater from the Ironton-Galesville. Examples are Elgin-4A, with a 226 Ra con- centration of 4.6 pCi/L and a 228 Ra concentration of 4.6 pCi/L and Elgin-5 , with 226 Ra concentrations ranging from 4.3 to 6.1 pCi/L and a 228 Ra concentration of 4.2 pCi/L. Note that groundwater from Elgin-5 and Elgin-6 has significantly different 226 Ra concentrations. The wells are located 198 meters (650 ft) apart and have open boreholes over the same stratigraphic interval. The significant difference between the two wells is that explosives were used in Elgin-6 to increase the pro- duction of groundwater from the Ironton-Galesville. Elgin municipal wells no. 5 and 6 are in a well field that is located along the Fox River on the north side of Elgin. Three other wells located in this well field are constructed with the borehole open to stratigraphic units from the Galena and Platteville through the upper section of the Mt . Simon. The °Ra concentrations in groundwater from two of these wells has been slightly less than 5.0 pCi/L for samples collected on two occasions. For these two wells explosives were not used to develop production of groundwater from the Ironton-Galesville Sandstone. The Ironton-Galesville section in the other municipal well that is open to the Mt. Simon (Elgin municipal well no. 1) was developed with ex- plosives. The ' :26 Ra concentration in three groundwater samples col- lected from this well have ranged from 5.2 to 8.5 pCi/L. One analysis for 228 Ra determined a concentration of 4.6 pCi/L. During rehabilitation -37- of this well in 1960 it was discovered that the borehole below the Ironton-Galesville was obstructed, reducing the quantity of groundwater from the lower stratigraphic units. The highest 226 Ra concentration was present in a groundwater sample collected before the rehabilitation, when because of the obstruction the well produced a greater component of groundwater from the Ironton-Galesville. After rehabilitation the Ra concentration in groundwater produced from the well decreased, with the lowest concentration, 5.2 pCi/L, being measured for the most recent sample collected. The Elgin analyses indicate that potable groundwater from the deeper stratigraphic units has lower concentrations of 6 Ra and 2 8 Ra than groundwater from the Ironton-Galesville. Radium analyses were performed on groundwater samples from four municipal wells located at Orland Park in the southern part of Cook County. 226 Ra concentrations of 15.7 pCi/L and 17.8 pCi/L were mea- sured in groundwater samples from two wells that were constructed to produce groundwater only from the Ironton-Galesville ( 2 Ra concentra- tions in the two wells were 6.5 and 7.4 pCi/L, respectively). Signifi- cantly lower concentrations of 8.1 pCi/L for 226 Ra and 4.9 pCi/L for poo Ra were measured in a groundwater sample collected from a well that is open to stratigraphic units from the Galena through the Ironton- Galesville. The low radium concentrations in the groundwater sample are believed to be related to an anomalous change in the chemical compo- sition of groundwater produced from this well. The available chemical analyses indicate that during the two year period after the well was put into service there was a gradual decrease in the concentration of calcium (from 69 to 29 mg/L) , magnesium (from 23 to 13 mg/L) , and sulfate (from 228 to 162 mg/L) in groundwater produced from the well. Over this period the concentration of total dissolved minerals has decreased from 847 to 720 mg/L. The mechanism responsible for the change in the chemical composition and the possible significance of this change to the concentration of radium in the groundwater is not understood at the present time. The municipal well (Orland Park no. 10) is a main supply well for the community and is pumped on a regular basis. If the change in chemical composition of the pumped groundwater were due to precipitation reactions occurring because of the mixing of incompatible groundwaters in the open borehole, then it is possible that the yield -38- of the well may decrease in the future because of a decrease in trans- missivity. The downhole geophysical and sampling methods discussed in the case history for Elgin well no. 3A in the barium section of this report would have application in investigating the mechanism responsible for the anomalous chemical composition of groundwater produced by this well. The Village of Coal City (situated in east-central Grundy County) has three wells that produce groundwater from the Cambrian and Ordovician bedrock. The borehole of municipal well no. 5 at Coal City is open to stratigraphic units from the Prairie du Chien Group through the Ironton- 226 Ra Galesville. A groundwater sample collected from this well has a concentration of 18.6 pCi/L and a " :8 Ra concentration of 9.9 pCi/L. The two other public supply wells at Coal City do not produce ground- water from the Ironton-Galesville and have much lower radium concentra- tions. Well no. 4 at Coal City has an open borehole to stratigraphic units from the Galena and Platteville through the St. Peter. Groundwater collected from this well had a 226 Ra concentration of 9.2 pCi/L and a 228 Ra concentration of 8.2 pCi/L. The borehole of well no. 3 at Coal City is open to the Galena and Platteville. However, fractures and crevices in the dolomite may have open connections to the underlying Glenwood- St. Peter and therefore a large component of the groundwater produced from the well may be from the sandstone. A groundwater sample collected from this well had a 226 Ra concentration of 8.9 pCi/L and a 228 Ra con- centration of 6.0 pCi/L. Although the highest concentrations are present in groundwater from the Ironton-Galesville the combined concentrations of the two radium isotopes in groundwater from the three wells ranges from 14.9 to 28.5 pCi/L. Radium analyses on groundwater samples are available for three municipal wells at Morris. The village is located along the Illinois River in northeastern Grundy County. Morris well no. 4 has an open borehole to stratigraphic units from the Prairie du Chien Group through the Ironton-Galesville. 226 Ra concentrations in groundwater samples collected from this well in 1957 and 1980 are 8.2 and 8.4 pCi/L, respectively. Morris well no. 5 is constructed with an open borehole to stratigraphic units from the Galena and Platteville through the ■39- Ironton-Galesville, and explosives were used to increase production from the Ironton-Galesville . A groundwater sample collected from this well in 1980 had a 226 Ra concentration of 8.2 pCi/L. The lowest ra- dium concentrations are 3.2 pCi/L for 226 Ra and 5.8 pCi/L for 228 Ra in a groundwater sample collected in 1957 from Morris well no. 3 when the borehole was open to stratigraphic units from the Galena through the Glenwood-St. Peter. In 1975 this well was deepened and constructed with an open borehole to stratigraphic units from the Galena through the Ironton-Galesville. A 226 Ra concentration of 4.4 pCi/L was measured in a groundwater sample collected in 1980 after the well was deepened. Although this concentration is greater than the value measured before the well was deepened, it is significantly lower than the 26 Ra con- centrations measured in municipal wells no. 4 and 5. A possible reason for the lower concentration in well no. 3 is that explosives were not used to increase production of groundwater from the Ironton-Galesville. 226 Ra analyses on groundwater samples are available for four municipal wells at Peru. The city is located along the Illinois River in western La Salle County. The highest concentration, 6.5 pCi/L, was measured in a groundwater sample from Peru well no. 8. This well has an open borehole to stratigraphic units from the Galena through the Ironton-Galesville. Explosives were used to increase the production of groundwater from the Ironton-Galesville. The other three Peru municipal wells (no. 5, 6, and 7) are located in close proximity to one another. All three wells are constructed with open boreholes to stratigraphic units from the Galena through the Ironton-Galesville. The significant difference in the construction of the three wells is that explosives were used in well no. 7 to increase production of groundwater from the Ironton-Galesville. Groundwater produced from well no. 7 has a higher 26 Ra concentration, 6.1 pCi/L, than groundwater from the other two wells. The concentration of radium-226 in Peru wells no. 5 and 6 ranges from 3.1 to 4.8 pCi/L. The only 228 Ra analysis available for Peru wells is a concentration of 2.6 pCi/L measured in a groundwater sample from well no. 6. The public supply wells that produce groundwater with the highest D Ra concentrations are located in the southwestern region of the study -40- area. These wells produce highly mineralized groundwater from the stratigraphic interval Galena through Glenwood-St. Peter. 226 Ra con- centrations are commonly greater than 10.0 pCi/L, with values greater than 20.0 pCi/L at a few well sites. It is notable that the 228 Ra concentration in groundwater at well sites in the southwestern region are commonly less than 2.0 pCi/L. The precise concentration of 226 Ra in groundwater from the Ironton-Galesville in the southwestern region is not known. The highest concentration of 226 Ra measured in the present re- search was 40.2 pCi/L for a sample of brine collected with a downhole sampler from the lower section of the Mt . Simon at a stratigraphic test boring located in the northeastern corner of Lake County in northeastern Illinois. Mechanisms That Control the Concentrations of Radium Isotopes in Groundwater Because of the differences in radium concentrations that may exist in groundwater from discrete stratigraphic units at a specific well, alterations in the well construction or a change in the rate and dura- tion of pumpage may result in a change in the concentration of radium in groundwater produced from the well. The fact that the highest radium concentrations are found in groundwater from the Ironton-Galesville in the northeastern region indicates that an increase in the radium concen- tration in groundwater may occur in the future at wells located within the major pumping centers (fig. 4; table 1). As dewatering of the Galena and Platteville and the Glenwood-St. Peter proceeds in the pumping centers, the component of groundwater produced from the Ironton- Galesville will be increased. In addition to the differences in radium concentrations in ground- water from discrete stratigraphic units open to a well bore, there may also be significant differences in the chemical composition of major ions in the groundwater. The mixing of dissimilar groundwaters in the well bore may result in secondary precipitation of mineral species. Some of these chemical reactions result in an efficient coprecipitation of radium. Study at several wells in northeastern Illinois has estab- lished that precipitation of secondary barite in the well bore may -41- result from the mixing of incompatible groundwaters from discrete strati- graphic units. At these wells the precipitation of secondary barite is often affected by the rate and duration of pumpage. The coprecipitation of radium in the barite may result in a significant decrease in the con- centration of radium in groundwater produced from the well. The effi- cient coprecipitation of radium with barium to form the sulfate is a well-known phenomenon (Gordon and Rowley, 1957). The radium concentra- tion is higher during periods when secondary barite precipitation is not occurring. Part of the increase in the radium concentration during these periods may be due to the mobilization of radium from the secondary barite. Several geochemical and radiochemical mechanisms may be signifi- cant controls on the concentration of a radium isotope in groundwater. An inherent difficulty in determining the significant mechanisms are the ultratrace concentrations of radium isotopes present in solution. For the analyses presented in this report on potable groundwater, the activity concentration of 226 Ra ranges from approximately 0.1 to 36.9 pCi/L and is equivalent to a radium mass concentration that ranges from 1.0 x 10 ° to 3.8 x 10 -8 mg/L. For comparison, the maximum value of 32.7 pCi/L measured for 228 Ra in potable groundwater is equivalent to a radium mass concentration of only 1.2 x 10 -10 mg/L. The combined mass concentrations of 228 Ra, 22Lf Ra, and 223 Ra that occur in groundwater are insignificant relative to the mass concentration of 226 Ra. The study of the distribution of radioactive elements in ground- water flow systems must include consideration of the residence time of groundwater in relation to the half -life of the radioactive elements. Laboratory experiments by Starik (1963) have demonstrated that nuclides of the short-lived isotopes, 224 Ra and 223 Ra, are released from the solid matrix of crushed minerals to solution at a rate that is dependent on their radioactive half -life. The accumulation of the nuclide into solution is described by equation (2) written in the following form: A t = A e (l-e~ At ) (3) where A is the concentration in disintegration units at time t, and A is the equilibrium concentration in disintegration units. -42- Equation (3) illustrates that if disintegration of the thorium parent on the matrix is of primary significance to supplying the radium nuclide to solution (Starik, 1963; Cherdyntsev, 1971), the concentration of radium nuclides in groundwater may increase with an increase in the residence time of the groundwater in the aquifer. Therefore, the presence of low dissolved 228 Ra concentrations in groundwater in pri- mary recharge zones may be due to the short residence time of ground- water. However, the concentrations of 228 Ra and 228 Ra present in groundwater with long residence time in confined regions of groundwater flow systems would reflect the abundance of their parent sources in the aquifer matrix (Cherdyntsev, 1971) . The 2 6 Ra/ 228 Ra activity ratio for 90 groundwater samples from the Cambrian and Ordovician bedrock in the study area ranges from 0.2 to 13.1, excluding an anomalously high value of 41.0 for a public supply well at Toulon in Stark County. (The radium concentrations in ground- water from this well are discussed later in this report.) As shown on figure 7 most analyses are enriched in z ' 26 Ra and have activity ratios ranging between 1 and 5. Fifteen analyses show values greater than 5. The high ratios are evidence of an enrichment on the matrix of the 226 Ra parent nuclides ( 238 U and 234 U) relative to their normal abundance in rocks. The accumulation of 23i+ U and 238 U on the sandstone matrix is due to the transport of the nuclides in groundwater. Study of the occurrence in groundwater of the nuclides in the 238 U decay series that precede 226 Ra and of the 222 Rn nuclide generated by disintegration of 226 Ra yields information on significant mechanisms that control the occurrence of 226 Ra in groundwater. In all groundwater samples analysed in this research the 222 Rn concentration was greater than the < :25 Ra concentration by at least one order of magnitude. At wells where ' :26 Ra concentrations in groundwater were low, the 222 Rn concentrations were commonly 2 to 3 orders of magnitude greater than o o a the ' Ra concentrations. Some of the highest 222 Rn concentrations were measured in groundwater with the lowest 226 Ra concentrations. The short half-life of 222 Rn (3.82 days) precludes significant migration in groundwater from its site of origin. The short half-life also precludes a significant contribution of 222 Rn to solution by diffusion -43- from a site of generation within minerals (Tanner, 1964A) . The 222 Rn concentrations in groundwater are evidence that 226 Ra is ubiquitous on the surfaces of the aquifer matrix; however, geochemical mechanisms strongly limit the concentration of 22b Ra present in groundwater. In the present research, unfiltered and filtered groundwater samples were collected at numerous wells to determine the 226 Ra con- centration on fine particulate matter that may be present in the ground- water. The analytical results for the majority of the filtered and un- filtered samples were identical within the limits of analytical accuracy. The 226 Ra was present in groundwater as dissolved species; very little of the 226 Ra was present adsorbed on particulate matter. The filtered and unfiltered concentrations for a limited number of samples are pre- sented in table 3. Equilibrium solubility with a mineral phase is not a controlling mechanism on the concentration of radium in groundwater from the Cambrian and Ordovician bedrock. Radium sulfate (RaS0 4 ) with a solubility product of 10~ 10,t+ is the most insoluble of the alkaline earth sulfates and is also the most insoluble radium salt known (Kirby and Salutsky, 1964) . However, the maximum concentrations of radium measured in groundwater from the Cambrian and Ordovician bedrock are many orders of magnitude less than allowed by the solubility of RaSO^. Although radium does not form pure mineral species in natural sys- tems, there is great potential for the element to precipitate from solu- tion during the formation of other mineral species. For example, radium is very efficiently coprecipitated with barium in the formation of the mineral barite (BaSO^) (Gordon and Rowley, 1957). Data gathered in the present study at several wells have determined that in groundwater that has a significant concentration of barium the precipitation-dissolution of barite may be an important control on the radium concentration. Barite does not emanate 222 Rn generated by disintegration of coprecipitated 226 Ra (Hahn, 1936). Therefore, coprecipitation of 226 Ra in secondary barite is not a suitable mechanism to explain the high 222 Rn concentrations that are present in groundwater with very low 226 Ra concentrations in primary recharge zones in the north-central region of -44- the study area. Geochemical mechanisms that may be responsible for the low 226 Ra concentrations in groundwater in the recharge zone are copre- cipitation of radium with calcium or magnesium to form carbonates or coprecipitation of radium with iron or magnesium to form hydrous oxides. Researchers who have studied groundwater in hydrogeological environments similar to primary recharge zones in north-central Illinois have at- tributed low 226 Ra and high 222 Rn concentrations to these coprecipi- tation mechanisms (Miholic, 1958; Tanner, 1964B) . The immediate parent of all 226 Ra is 230 Th present on the aquifer matrix; however, the ultimate source of all 226 Ra is 238 U, the long- lived parent isotope of the decay series. The precursor nuclides to 226 Ra, in order of occurrence, are 238 U, 234 Th, 234 Pa, 234 U, and 230 Th. Significant disequilibrium between the nuclides may occur in groundwater flow systems because of differences in the chemical and radiological properties of the nuclides. Analyses in this research have determined the widespread existence of extreme disequilibrium between 231+ U and 238 U in groundwater from the Cambrian and Ordovician bedrock. The ground- waters are greatly enriched in 231+ U. The enrichment is thought to re- flect the widespread presence of 238 U and 234 U in secondary accumulations on grain surfaces in the sandstone bedrock. The mechanisms that may be responsible for the disequilibrium are discussed in the section of this report that deals with uranium isotopes in groundwater. An important source of 226 Ra in groundwater in the confined re- gions of the Cambrian and Ordovician bedrock is the accumulation of nuclides of 238 U, 231+ U, and 230 Th on the sandstone matrix due to the groundwater transport of uranyl ions. Two chemical mechanisms that re- move uranyl ions from groundwater in the confined bedrock are precipi- tation due to a decline in the oxidation-reduction potential of the groundwater and sorption onto the surface of silicates in the sandstone bedrock. The radioactive disintegrations of 238 U to 234 Th and 231+ U to o o o Th are also important mechanisms in the secondary deposition of the nuclides. The uranium nuclides disintegrate to thorium through emission of alpha particles. The recoil energy from the alpha particle emission may propel the newly formed thorium nuclides into solution from the residence sites of the uranium parent nuclides on the matrix. The -45- disintegration of dissolved uranium will also contribute thorium nuclides to solution. Thorium is a highly insoluble element, and 30 Th nuclides placed in solution by the disintegration of 234 U will not remain in solution. The primary mechanism by which 30 Th is removed from solu- tion is the rapid sorption of the nuclides onto the sandstone bedrock. Thorium ions, because of their +4 valence, have a strong affinity for the surface of silicates in the sandstone bedrock (Stumm and Morgan, 1970) Because of their +2 valence, radium ions are much less strongly sorbed to silicate surfaces than are either thorium ions or uranyl ions. Therefore, there is potential for desorption of 26 Ra formed by the disintegration of 230 Th on the matrix. The disintegration of 230 Th to 226 Ra is by alpha particle emission, so there is also potential for the newly formed 226 Ra nuclides to be physically ejected from surfaces into solution. Even if the new nuclide is not propelled into solution the disintegration energy will weaken the chemical bonds that hold the nu- clide to the matrix and thus increase its potential for leaching into solution. The desorption of 226 Ra from silicate surfaces is affected by the ionic strength of the groundwater. In dilute groundwaters, exchange sites on the matrix may selectively sorb radium ions out of solution. However, in more mineralized groundwater the selectivity of the matrix surface for specific ions decreases due to competition from the greater concentration of ions in solution. Therefore, as the ionic strength increases there is potential for more radium to be desorbed from the matrix. The relation between 225 Ra concentration and the concentration of total dissolved minerals for 183 analyses of groundwater samples from the Cambrian and Ordovician bedrock in Illinois is shown In figure 8. The concentration of dissolved minerals is a measure of the mineraliza- tion of the groundwater and ranges from 250 to 3220 mg/L. The concentra- tion of 226 Ra ranges from 0.1 to 25.0 pCi/L. All of the 226 Ra concentrations of less than 2.5 pCi/L occur in groundwater that has a total dissolved mineral concentration of less than 600 mg/L. Also, all of the groundwater samples with concentrations -46- 25- • • 20- • o • • • o oj*^# • • fj OJ t co C cc *-> C i_ ro co 03 -o t£ n 1 § o O o o o c IT J 8 CN o 0^ o o % ,> o CM in T" O -I- in (-|/!0d) ey 0) CD 131 II cn .2; > u c 5: CO ■£ "o ■£ _ •;: co O c c i> 5: ° O) UL £ 2 CD J3 8 .2 If . c O) co Ui g — CO u- O 91Z -54- would be efficiently released by these mechanisms to groundwater, with one exception. Precipitates of barite do not release 222 Rn generated by disintegration of coprecipitated 226 Ra (Hahn, 1936) . Some anomalously low and anomalously high 222 Rn concentrations measured in groundwater from wells in northeastern Illinois are attributed to secondary accumu- lations of barite on the rock matrix of the well bore. At these wells low concentrations of 222 Rn and 226 Ra in groundwater were measured during periods when extreme supersaturation of barium and sulfate ions in solu- tion resulted in secondary precipitation of barite. High concentrations of 226 Ra and 222 Rn in groundwater were measured at these wells during periods of dissolution or corrosion of the barite. URANIUM ISOTOPES IN GROUNDWATER The two isotopes of uranium that may occur naturally at signifi- cant concentrations in groundwater are 38 U and 3Lf U. Both isotopes are members of the 238 U decay series (fig. 5). The U.S. Environmental Pro- tection Agency has not promulgated a maximum contaminant level for uranium in water. A very preliminary study that applied the criteria used to determine the drinking water regulations for 2 6 Ra indicated that a maximum contaminant level for uranium in groundwater would be approximately 10 pCi/L (C. Richard Cothern, United States Environmental Protection Agency, personal communication). However, the element was excluded from the National Interim Primary Drinking Water Regulations because uranium is both chemically and radiologically toxic. Analyses for 34 U and 238 U were performed on groundwater samples collected from 130 wells in Illinois that produce groundwater from the Cambrian and Ordovician bedrock. Nineteen-liter (5-gallon) samples of groundwater were collected at the wells and acidified with nitric acid to a pH of 1. The laboratory analytical methods are described in Osmond and Cowart (1976). Both filtered (0.22 micron pore size membrane) and unfiltered samples were collected at many wells located widespread across the study area to evaluate the effect of particulate matter in groundwater on the measured concentration of the two nuclides. The results on the filtered and unfiltered samples were identical within the statistical limits of accuracy for the analytical method. -55- The mass concentration of uranium in the 130 groundwater samples ranged from less than 0.002 yg/L to 1.10 yg/L; the combined activity concentration of 23t+ U and 238 U in the samples ranged from less than 0.1 pCi/L to 7.9 pCi/L. For the majority of analyses (95 samples) the combined concentrations were less than 2.0 pCi/L. The combined con- centrations were greater than 3.0 pCi/L in 11 samples and the combined concentrations exceeded 5.0 pCi/L in only 3 samples. All of the com- bined activity concentrations greater than 1.0 pCi/L were present in groundwater that had marked enrichment in 23Lf U. Dissolved uranium species are very sensitive to the oxidation- reduction potential of groundwater. Uranium occurs as the mobile hexa- valent ion (usually as uranyl) in oxidizing environments and as the highly insoluble quadrivalent ion in reducing environments. Uranium (as uranyl complexes) may migrate away from the primary source and be emplaced in secondary accumulations. Precipitation of uranium at re- ducing barriers in groundwater flow systems is a mechanism responsible for the formation of some economic uranium deposits in sandstones (Hostetler and Garrels, 1962). Figure 10 presents the relation between Eh and uranium concentra- tion in groundwater for 77 wells in the study area that produce ground- water from the Cambrian and Ordovician bedrock and for 5 wells that produce groundwater from shallower aquifers. The data indicate that the oxidation-reduction potential is an important control on the uranium concentration in groundwater. Dissolved uranium concentrations were less than 0.08 yg/L in all groundwater samples in which Eh values were -50 millivolts or lower. The highest uranium concentrations are present where Eh values were +100 millivolts or greater. Many groundwater samples with Eh values between -50 to +50 millivolts have appreciable uranium concentrations. It is important to note that the wells have boreholes that are open to several stratigraphic units; groundwater from discrete units may differ markedly in dissolved sulfide concentrations, oxidation- reduction potential, and uranium concentration. The sensitivity of Eh values measured with the platinum electrode method to the presence of sulfide in groundwater is discussed in an earlier section of this report. -56- 400' 300- 200- 100- 0- 100- -200- Aquifer strata at well site Qunconfined confined Number value is 234 U/ 238 U activity ratio Pleist. = Pleistocene sand and gravel Sil. = Silurian dolomite Maq. = Maquoketa Shale Group ■*- Value shown is limit of detection; concentration is less A31.3 o 29.9, 25.6 1 23.8 »7.6 -o 2.6 Pleist.^ 1.3 O 5.5 o 8.8 o 2 i J4.3 25.1 |7.7 Sil. * 25 Q 23.5' k 20.8 k 24.5 ^18.9 i 2?4 f j i" |19 6 .9.0 &8.4 *14.3 17.4i 18.4. 025.9 W CB 152 256 #27.6 #23.5 22.6 16.6 -,, 25.6 A)A 16.4 11.9 21 o*t? o 8.1 8.5_ #!H dissolved sulfide absent D 21 6 dissolved sulfide present 105 26. 1^ ^P27.2 ^22.0 • # 20 -6 ,—.27.8 Maq. 2.3^ 17.3, |9.8 36. 1i k8.7 21.6| k 3.1 &21.4 |10.3 A6.1 40.7. ^^" ^16.3 ^18.7 m 22 - 6 n i.4 12.9 10.4^ '7.7 k5.4 t10.8 >6.8 7.0' 10.0 k 16.7 k 13.8 Pleist. k 22.2 »20.4 >7.2 k 10.8 ISGS 1983 I I I I I I I I I 0.005 0.01 1 I ' ' ' ' I 0.05 0.1 U Oig/L) 0.5 i i i i i 1.0 0.001 FIGURE 10. The relation between Eh (platinum electrode) and the concentration of uranium in groundwater from the Cambrian and Ordovician bedrock. The 234 U/ J3, U activity ratio is shown for each analysis. A few analyses are listed for groundwater from other aquifers. -57- Figure 10 presents the value for the 231+ U/ 238 U activity ratio with the analyses. The ratios range from 2.4 to 40.7 for the analyses on groundwater from the Cambrian and Ordovician bedrock and from 1.3 to 2.5 for the analyses on groundwater from shallower aquifers. For all of the analyses, there is no relation between the activity ratio and either the uranium concentration or the Eh of groundwater. Disequilibrium in the natural environment between 23l+ U and 238 U was first reported in 1955 (Cherdyntsev et al., 1955). From a review of the literature, Osmond and Cowart (1976) found that research has documented the disequilibrium in a wide range of materials and environ- ments. Most waters are found to have disequilibrium between the acti- vities of 23t+ U and 238 U. The dissolved uranium in surface waters is usually enriched in 23t+ U; ratios of the activity of 231+ U to 238 U as great as 6.4 have been reported for small streams. The ocean has about 15 percent more 23Lf U than 238 U. The uranium isotopic abundances and the disequilibrium between them may vary widely in groundwater; reported values for activity ratios range from as low as 0.5 to as high as 12.3. The widespread and extreme disequilibrium between 231+ U and 238 U that occurs in groundwater from the Cambrian and Ordovician bedrock of the study area is unique. The high activity ratios have not been found in other flow systems. The 23t+ U/ 238 U activity ratios in groundwater from the Cambrian and Ordovician bedrock do indicate regional patterns. The lowest activity ratios (values less than 3.0) were measured in the north-central and northwestern part of the study area in zones of primary recharge where the Pennsylvanian System and the Maquoketa Group are absent and ground- water has had a short residence time in the bedrock. With increasing distance along flow paths from the point of recharge there is an increase in the activity ratio, and values greater than 20 occur widely in the study area. The highest activity ratios are present in the northeastern, northwestern, and east-central parts of the study area where the Cambrian and Ordovician bedrock are confined. An interesting feature are the relatively low 23t+ U/ 238 U activity ratios in confined groundwater in the southwestern region of the study area. The activity ratios range from 5.4 to 14.8 and are commonly less than 10 over the central and southern -58- part of the region. The genesis of the extreme enrichment of 234 U in groundwater of the Cambrian and Ordovician is not well understood at the present time. The high activity ratios may be due to interaction of geochemical and radiological mechanisms. It is also important to consider geologic processes that may have supplied uranium to the groundwater flow system. Rosholt et al. (1963) advanced the concept that the recoil from the 238 U->- 23l+ Th alpha particle decay event may place the subsequently formed 231+ U nuclide ( 234 Th and 234 Pa have short half -lives) in an en- vironment that would favor a +6 valence; in fact, self -oxidation may occur as a result of electron stripping during the 23t+ Th-> 23t+ Pa-> 231+ U decay events. The higher valence state would enhance dissolution of 23Lf U into groundwater. Kigoshi (1971) demonstrated through experimentation that direct recoil of the Th nucleus (from 238 U decay) into solution could occur. The short-lived 231+ Th decays to 23t+ U, thereby enriching the groundwater in 23 ^U while depleting the surface of the aquifer matrix. A large number of the recoiling nuclei pass through the groundwater in the open pore spaces to become embedded in adjacent grains of the aquifer matrix, leaving fission tracks open to the surface (Huang and Walker, 1967). Fleischer (1980) experimented with natural minerals (mica, feldspar, quartz, and obsidian) to investigate the release of embedded nuclei as a result of the etching of fission tracks by solutions. An important finding of the study was that all of the solutions (including deionized water) would leach a significant portion of the embedded nuclei from all of the minerals. Kronfeld (1974) advanced 23t+ Th alpha-recoil as the mechanism re- sponsible for the generation of 23Lf U/ 238 U activity ratios as high as 12 in groundwater from a confined sandstone aquifer in Texas. Kronfeld proposed that the effectiveness of the alpha-recoil mechanism was en- hanced by the presence of a diffuse precipitate of secondary uranium on the sandstone matrix. Disintegration events that weaken chemical bonds or propel nu- clides into solution and solution etching of fission tracks may all be -59- mechanisms that are important in the enrichment of ^ 0H U in groundwater from the Cambrian and Ordovician bedrock in the study area. The trans- port of 231+ U and 238 U in regional groundwater flow systems with secondary accumulation on the sandstone bedrock is thought to be important to the presence of high concentrations of 6 Ra in groundwater. The wide- spread occurrence of high 226 Ra/ 228 Ra activity ratios in groundwater from the Cambrian and Ordovician bedrock (fig. 7) is evidence of the presence of secondary uranium on the sandstone bedrock. In the present report the authors are suggesting a new concept: that secondary uranium and possibly the extreme enrichment in 234 U in groundwater may be significantly related to glacial processes. Research by Perry, Grundl, and Gilkeson (1982) established that the isotopic composition of the water molecule is evidence of the importance of gla- cial processes to the recharge of groundwater to the Cambrian and Ordo- vician bedrock where these units are confined by the Maquoketa Shale Group and/or the Pennsylvanian System. Gilkeson, Perry, and Cartwright (1981) advanced a glacial recharge mechanism to explain the anomalous isotopic composition of both the water molecule and the dissolved sul- fate present in groundwater of the Cambrian and Ordovician bedrock in the northeastern, southeastern, and southwestern regions of the study area. The mechanisms of recharge related to glacial processes is not clearly understood. As a result of the geothermal gradient, as well as the heat generated by friction, meltwater was present at the base of the continental glaciers (Weertman, 1972; Clayton and Moran, 1974; Freeze and Cherry, 1979). It is postulated that pressure from the glacial ice overburden caused basal meltwater to recharge into the Cambrian and Ordovician bedrock. It is important to consider that these glacial recharge events were repeated with the numerous advances of continental ice sheets into Illinois during the Pleistocene. The pulverizing of uranium-bearing rocks to very fine particles by glaciers would increase the availability of uranium for dissolution. The con- cept advanced in the present report is that the dissolved uranium ex- tracted from the finely ground sediments into the basal meltwater was enriched in ' 34 U that was more mobile because of its radiogenic origin. -60- However, the extent of preferential dissolution of 231+ U in glacial en- vironments has not been established. Analyses on uranium isotopes in glacial materials are limited. A few analyses by Rosholt, Doe and Tatsumuto (1966) determined that glacial tills were greatly depleted in 238 U as compared with 232 Th and also depleted in 234 U as compared with 238 U. Additional analyses on glacial sediments and glacial meltwaters are necessary to confirm the extraction of uranium with preferential ex- traction of 34 U. The regional distribution of 234 U/ 238 U activity ratios in ground- water of the Cambrian and Ordovician bedrock in the study area is com- patible with the idea that glacial processes are involved in generation of the disequilibrium. The highest activity ratios (values from 25 to 40) occur in the eastern part of the study area where groundwater would have been recharged during the Wisconsinan by the most recent glacial advances. Wisconsinan glaciers did not advance over the southwestern region of the study area. The low activity ratios (values from 5 to 11) in ground- water of the Ordovician bedrock in the southwestern region may reflect recharge by an earlier glacial advance. The activity ratio diminishes over time as the excess 234 U disintegrates. It is possible that ground- water in the Cambrian and Ordovician bedrock in the southwestern region is quite old, and that recharge was due to a glacial advance during the Illinoian. THORIUM ISOTOPES IN GROUNDWATER The immediate parent nuclides of all four natural radium isotopes are isotopes of the element thorium. However, the concentration of thorium ions in solution is insignificant; the source for the radium isotopes in groundwater is the disintegration of thorium isotopes present either in accessory minerals in the aquifer matrix or as sec- ondary accumulations on the aquifer matrix. In the present research, alpha pulse height analyses for 232 Th (the parent nuclide for 228 Ra) and 230 Th (the parent nuclide for 226 Ra) were performed on groundwater samples from seven wells that produce groundwater from the Cambrian and Ordovician bedrock. Groundwater from the seven wells exhibits a large variation in radium concentrations; -61- 226 Ra concentrations range from 0.3 to 15.9 pCi/L and 228 Ra concentra- tions range from 2 to 13 pCi/L. The thorium species were concentrated from separate 19-liter (5-gallon) samples of unfiltered and filtered (0.22 micron pore size membrane) groundwater collected from each well. The analysis did not detect 232 Th or 230 Th in any of the samples. The limit of the analytical method is 0.01 ug/L expressed as 232 Th. Thorium ions, which have a +4 valence state, have extremely slight solubilities. Thorium present in groundwater is probably re- stricted to ultratrace concentrations transported as complexed species (Herman and Langmuir, 1978). Organic materials may act as an agent for complexation. The transport of thorium species in groundwater flow systems in the Cambrian and Ordovician bedrock is unknown at the present time. Of primary interest is the role of transport of 32 Th as complexed species on the high concentrations of 2 Ra that occur in some localities in the northeastern and east-central regions of the study area. Because of the long half-life of 232 Th, it is possible that transport of ultra- trace concentrations over long periods of geologic time may result in significant concentrations in zones of secondary accumulation. The sec- ondary accumulation of 32 Th may have occurred earlier in geologic time when thorium was transported in the Cambrian and Ordovician bedrock by brines that contained organic material. The transport of 230 Th to the concentration of 225 Ra in groundwater is thought to be insignificant be- cause of the short half-life of the nuclide. -62- BARIUM IN GROUNDWATER BACKGROUND Barium concentration in untreated groundwater from a large number of wells in Illinois exceeds the drinking water standard of 1.0 mg/L (U.S. EPA, 1975). The largest area where the groundwater ex- ceeds the standard is in northeastern Illinois. In this area of north- eastern Illinois and in more restricted settings of northern Illinois, high-barium groundwater is produced from bedrock aquifers of the Cam- brian and Ordovician Systems. Outside northern Illinois, high barium concentrations occur in groundwater from less productive aquifers in isolated locations. Gilkeson et al. (1978) established that: (1) the concentration of dissolved sulfate in groundwater is an important control on the concentration of dissolved barium because of the low solubility of the mineral barite (BaSO^) ; (2) the high barium concentrations occur in groundwater that is anomalously depleted in dissolved sulfate; (3) the high barium concentrations occur in confined aquifers; and (4) the high dissolved barium concentrations occur within anaerobic environments or downgradient these environments. Gilkeson, Perry, and Cartwright (1981) interpret the depletion of dissolved sulfate is due to bacterially mediated reactions occurring in anaerobic environments. Characteristics of the groundwater that support this interpretation are the presence of hydrogen sulfide, increased alkalinity, and enrichment of the heavy sulfur isotope ( 3t+ S) in the re- sidual dissolved sulfate. The source for dissolved sulfate in groundwater from the Cambrian and Ordovician bedrock of northeastern Illinois where the bedrock is con- fined by the Maquoketa Shale Group is problematic. Stable isotope anal- yses have established that the sulfate molecule is of a different geo- logic age than the Cambrian or Ordovician (Perry, Grundl , and Gilkeson, 1982). Gilkeson, Perry, and Cartwright (1981) have suggested that the sulfates are derived from evaporites of a younger geologic age and were recharged to the Cambrian and Ordovician bedrock as a result of glacial -63- processes. Prior to the glacial recharge events, groundwater in the Cambrian and Ordovician sandstone bedrock may have had low sulfate concentrations and high concentrations of barium. Feldspars present as accessory minerals in the sandstone bedrock may have been the source of barium. The interaction between an invaded groundwater with high concentrations of sulfate and a native groundwater with high concen- trations of barium resulted in secondary precipitation of barite on the aquifer matrix. At the present time, within localities where strong anaerobic conditions exist, the sulfates have been consumed by microbial reactions. With depletion of the sulfates, barium concen- trations are not controlled by the solubility of barite and may be present at significantly higher concentrations. BARIUM IN GROUNDWATER FROM BEDROCK OF THE CAMBRIAN AND ORDOVICIAN SYSTEMS IN NORTHEASTERN ILLINOIS Figure 11 shows the distribution of sulfate, and figure 12 the distribution of barium, in groundwater from municipal wells in a region in northeastern Illinois that produce groundwater from the Cambrian and Ordovician bedrock. Analyses providing data for figures 11 and 12 were performed on groundwater samples from wells open to the Ironton- Galesville Sandstone; a few of the wells were open only to the Ironton- Galesville but most of the wells were also open to other stratigraphic units in the Cambrian and Ordovician Systems. The samples analyzed were collected during periods of large withdrawals of groundwater from each well. Figure 11 indicates that over a large area of northeastern Illi- nois the dissolved sulfate concentration is less than 10 mg/L. East and southeastward from this area the dissolved sulfate concentration in- creases rapidly and is present at concentrations greater than 800 mg/L in southeastern Cook County. Figure 12 shows that over a large part of McHenry and Kane Counties, the northeastern and central part of De Kalb County, the western part of Lake County, the northwestern part of Cook County, and the northwestern part of Du Page County, groundwater pro- duced from wells open to the Cambrian and Ordovician bedrock exceeds the EPA drinking water limits of 1.0 mg/L for barium. At many wells the dissolved barium concentration exceeds 10 mg/L. At a few wells the -64- BOONE WISCONSIN MCHENRY Michigan 30 km /yO Sulfate isoline (mg/L) \j. Boundary of Maquoketa Shale (O^) isgs 1983 • Well site (the four named sites have experienced pump failure due to barite precipitation). (|>> Regional zone with anaerobic environment FIGURE 1 1. The distribution of dissolved sulfate in groundwater from the Cambrian and Ordovician bedrock for a region of northeastern Illinois (the Ironton-Galesville Sandstone is the principal aquifer). -65- BOONE WISCONSIN MCHENRY Maquoketa Shale present LAKE / i A I I \ l v • \K \ r Lake Michigan • • COOK \ i / •* • • • / ii • • t\ '! •• • (• # /) 9 DUPAGE •- J— . Stf _* . > U —-«-^ . /• • I \S/ • • • • (•/ .__-- ° 06 i • • 10 15 20 mi 10 20 =1 30 km N 9> Barium isoline (mg/L) £ Boundary of Maquoketa Shale (O^) • Well site ISGS 1983 FIGURE 12. The distribution of dissolved barium in groundwater from the Cambrian and Ordovician bedrock for a region of northeastern Illinois (the Ironton-Galesville Sandstone is the principal aquifer). -66- dissolved barium concentration is greater than 20 mg/L. Barium concentrations were determined for unfiltered and fil- tered (0.22 micron pore size membrane) untreated groundwater samples collected at many wells. The effect of the filtration on barium con- centrations was insignificant, even for groundwater having the highest barium concentrations. The barium in groundwater pumped from the wells was in solution; very little was present as particulate matter. A comparison of figures 11 and 12 shows the important relation be- tween the concentration of sulfate and the concentration of barium in groundwater. Where the rapid increase in sulfate concentration in groundwater occurs (in the eastern and southern part of the region) , there is a rapid decline in barium concentration. The two figures can be used as a basis for making decisions concerning the siting of high- capacity wells to produce groundwater from the Cambrian and Ordovician bedrock. In northeastern Illinois, the highest dissolved barium concen- trations are present in a locality in which there is a strong anaerobic environment in the Cambrian and Ordovician bedrock due to confinement by the Maquoketa Shale Group. Over a large part of Boone and De Kalb Counties, the Maquoketa is absent and low barium concentrations occur in groundwater from the Cambrian and Ordovician bedrock that has a high Eh. However, at a location in central De Kalb County where the Maquoketa is absent, dense, unfractured dolomites of the Galena and Platteville Groups confine the underlying bedrock, and concentrations of barium greater than 1.0 mg/L are present in groundwater, having a low Eh, that is produced from the St. Peter Sandstone and from the Ironton-Galesville Sandstone. In addition to the regional variation in barium and sulfate in groundwater from wells finished In the Cambrian and Ordovician bedrock (figs. 11 and 12), a marked variation in barium and sulfate may also occur in groundwater produced from different stratigraphic units open to the borehole at a specific well site. An objective of the present research was to identify the stratigraphic units of the Cambrian and Ordovician Systems that produce groundwater with high barium concentrations, -67- If the source of the barium is groundwater from discrete intervals of relatively nonproductive bedrock that are open to the borehole, then the cemented casing used in well construction to prevent production of groundwater from the discrete units would lower the concentration of barium in groundwater produced from the well. Data gathered in this study have established that the high dissolved barium concentrations occur in groundwater from the more productive aquifers in the Cambrian and Ordovician bedrock — the St. Peter Sandstone and the Ironton-Galesville Sandstone. In fact, groundwater from the Ironton-Galesville, regionally the most productive aquifer, has the highest dissolved barium concen- trations. Because of the high barium concentrations that occur in groundwater in the more productive bedrock, design features (cemented casing) do not have a significant effect on lowering barium concentra- tions in groundwater produced from high capacity wells. Specific examples of this general finding are presented in the following discussion, A test well for the community of Sleepy Hollow in Kane County was initially constructed to produce groundwater only from the St. Peter. Analysis of a sample collected during a production test determined a barium concentration of 13.2 mg/L for groundwater from the St. Peter. The test well was then deepened and constructed to produce groundwater only from the Ironton-Galesville. Analysis of a groundwater sample col- lected from the Ironton-Galesville during a production test determined a barium concentration of 18.2 mg/L. Additional examples of wells constructed to produce groundwater only from the Ironton-Galesville that produce groundwater with a barium concentration greater than the drinking water regulation of 1.0 mg/L are Carol Stream-5 (Du Page County) , with a barium concentration of 4.4 to 4.9 mg/L; Bloomingdale-8 (Du Page County), with a barium con- centration of 3.5 mg/L; and Cary-6 (McHenry County), with a barium con- centration of 8.0 rag/L. Examples of wells that produce groundwater with high barium concentrations and are constructed with cemented casing from the surface downward through the St. Peter to produce groundwater mainly from the Ironton-Galesville (with a small contribution from the Eminence Formation, the Potosi Dolomite, and the Franconia Formation) -68- are : Crystal Lake-6 (McHenry County) , with a barium concentration of 10.2 mg/L; Crystal Lake-8 (McHenry County), with a barium concentration of 16.0 mg/L; and Cary-4 (McHenry County), with a barium concentration of 4.2 mg/L. Analyses of groundwater from municipal wells located in two well fields at Elgin in Kane County indicate that the highest dissolved barium concentrations are present in groundwater from the Ironton- Galesville. Municipal wells at Elgin are constructed with uncased boreholes to receive groundwater from a large section of bedrock. All of the wells are open to stratigraphic units from the Galena downward through the Ironton-Galesville . In addition, several wells at Elgin are also open to the Eau Claire Formation and the upper part of the Mt . Si- mon Sandstone. The wells at Elgin that produce groundwater with the highest dissolved barium concentrations were developed with explosives to increase production of groundwater from the Ironton-Galesville. Examples are Elgin-6, with a barium concentration ranging from 16.7 to 22.4 mg/L, and Elgin-5A, with a barium concentration of 18.2 mg/L. Lower barium concentrations occur in groundwater from municipal wells at Elgin that were not developed with explosives and therefore produce a smaller component of groundwater from the Ironton-Galesville. Examples are groundwater samples collected from Elgin-4A, with a barium concen- tration of 8.4 mg/L, and Elgin-5 , with a barium concentration of 7.6 mg/L, Note that groundwater from Elgin-5 and Elgin-6 have significantly dif- ferent barium concentrations. The wells are only 198 meters (650 ft) apart and have open boreholes to the same stratigraphic units. The significant difference between the two wells is that explosives were used in Elgin-6 to increase the production of groundwater from the Ironton-Galesville . Three well sites located in central De Kalb County illustrate the variation in barium present in groundwater from different strati- graphic units of the Cambrian and Ordovician Systems. A groundwater sample from a well located at Donny Brooks Estates constructed with an open borehole to stratigraphic units from the Galena through the St. Peter has a barium concentration of 0.13 mg/L. A groundwater sample from a well at Kishwaukee College constructed with the borehole open -69- only to the Glenwood-St. Peter has a barium concentration of 1.4 mg/L. In central De Kalb County, groundwater with the highest barium concen- tration (1.0 to 2.3 mg/L) was collected from a well at the village of Malta that has an open borehole to stratigraphic units from the Glenwood- St. Peter through the Ironton-Galesville. Groundwater samples with the highest barium concentrations were collected from this well during periods of extensive pumpage when a larger component of groundwater was produced from the Ironton-Galesville. Municipal wells at De Kalb, Illinois in De Kalb County are an additional example of the occurrence of higher barium concentrations in groundwater from the Ironton-Galesville. The municipal wells are con- structed with an open borehole to stratigraphic units from the Glenwood- St. Peter through the Ironton-Galesville. Two municipal wells developed with explosives in the Ironton-Galesville have slightly higher barium concentrations than wells not developed with explosives. The two wells developed with explosives are De Kalb-6, with a barium concentra- tion of 1.8 mg/L, and De Kalb-4, with a barium concentration of 1.4 mg/L. The concentration of barium in groundwater from four municipal wells at De Kalb not developed with explosives ranges from 0.3 to 1.0 mg/L. A laboratory experiment (Sillen and Martell, 1964) has determined the solubility product constant for barite to be 10 -10,0 . The solu- bility product for barite can also be calculated from thermodynamic data. The value was determined to be 10 -8 - 9 (Gilkeson et al., 1978) from calculations with thermodynamic values from Garrels and Christ (1965).. Refined values for the thermodynamic properties of elements and minerals were presented by Robie, Hemingway, and Fisher (1978). The solubility product for barite, determined with the refined thermo- dynamic data, is 10 -9 ' 97 — a value identical to the solubility product determined by laboratory experiment. The relation of the activities of barium ion (Ba ) and sulfate ion (SOlj ~) in samples of untreated groundwater collected from selected wells finished in bedrock of the Cambrian and Ordovician Systems in northern Illinois is shown in figure 13. The activities (thermodynami- cally effective concentrations) were calculated from the analytical concentrations of barium and sulfate in groundwater samples by using -7n- 4.0 4.2 4.4 4.6 4.8 5.0- 5.2 5.4- 5.6- 5.8 £6.0- Q. 6.2- 6.4- 6.6 6.8- 7.0 7.2- 7.4- 7.6- 7.8- 110 >9 2c 8 A A 2bAA3c #5 A 1c 1aA A A2a O O BaS0 4 saturation state at well sites A Pump failure due to BaS0 4 precipitation Analyses for numbered well sites, county/owner number 1a-d, Cook/Hanover Park-4; 2a-c, Kane/Elgin-3A; 3a-c, Lake/Lake Barrington Shores-2; 4a-d, Lake/Lake Zurich-5 9 High supersaturation 5, DuPage/Carol Stream-5; 6, Kane/Elgin-6; 7, Kane/Elgin -5A; 8, Kane/Elgin -4A; 9, McHenry /Crystal Lake-7; 10, McHenry/ Crystal Lake -8 A Slight -to -moderate supersaturation # Equilibrium O Undersaturated 1 1 1 1 1 1 1 1 1 1 1 1 1 5.0 4.8 4.6 4.4 4.2 4.0 3.8 3.6 3.4 3.2 3.0 2.8 2.6 2.4 2.2 p (S0 4 2_ ) FIGURE 13. The relation of the activities of (Ba 2t ) and (S0 4 J ~) in groundwater from the Cambrian and Ordovician bedrock to the solubility of BaS0 4 . ■71- the extended form of the Debye-Hlickel equation. The procedure for per- forming calculations with the Debye-Huckel equation is given in Hem (1970) The analyses presented in figure 13 are divided into five sets. One set of 40 samples (displayed as small dots) collected from well sites widely distributed in northern Illinois shows a strong linear trend on the log-log plot. A linear regression line was fit to the 40 analyses with a correlation coefficient, r, of -0.98. The means of the solubility products of the 40 samples is 10 -9,95 with a standard deviation of ± 0.08. This value is essentially identical to the solubility product of barite determined by laboratory experiment or by calculation with the refined thermodynamic data (Robie et al, 1978). The activities of (Ba 2+ ) and (SO^ 2- ) in the 40 analyses located along the linear regression line in figure 13 are interpreted to be in equilibrium with the mineral phase barite. Six analyses (displayed as open circles) in figure 13 that plot a significant distance below the solubility product line represent a set of groundwater samples that are undersaturated with respect to barite. This interpretation is supported by the fact that the groundwater samples in the set are from wells located in north-central Illinois that have short flow paths for groundwater from the point of recharge. The low (Ba 2+ ) activities are due to the short residence time of the groundwater or to a limited source of barium in the aquifer matrix and thus have not reached saturation. The groundwater samples on figure 13 that plot a significant dis- tance above the solubility product line are supersaturated with respect to barite. All of the groundwater samples that have barium concentra- tions greater than 1.0 mg/L are part of this set. The supersaturation is a result of the mixing of groundwater with different chemical com- positions in the open well bore. Figure 13 presents multiple analyses (displayed as solid triangles) from four well sites where expensive rehabilitation was necessary because of precipitation of barite in the pump. The analyses indicate that there is a variation in the degree of supersaturation depending upon the pumping schedule of the individual wells. The figure also presents analyses from several wells (displayed as large solid dots) where there is a high degree of supersaturation -72- but well failure has not occurred to the present time. There is poten- tial for barite precipitation to occur at these wells; the precipitation may result in failure of the wells. CASE HISTORIES— PUMP FAILURES DUE TO SECONDARY PRECIPITATION OF BARIUM SULFATE At four wells in northeastern Illinois expensive maintenance has been necessary because of barite precipitation in the well bore, pump and discharge pipe. The four wells are Hanover Park-4 in Cook County, Elgin-3A in Kane County, and Lake Zurich-5 and Lake Barrington Shores-2 in Lake County. All four wells are constructed with the borehole open to stratigraphic units from the Galena Dolomite Group through the Ironton-Galesville Sandstone. The boreholes extend a short distance into the Eau Claire Formation. The purpose of drilling a short distance into the shale and dolomite of the Eau Claire is not to produce groundwater from the formation but to provide a reservoir for loose bedrock that sloughs off of the borehole over time. At the four wells discrete rock units are contributing ground- water of markedly different chemical compositions to. the well, and therefore, the chemical composition of groundwater discharged from the wells is related to the pumpage schedule. During periods of extensive withdrawals a larger component of groundwater is produced from the Ironton-Galesville Sandstone, the most productive unit, and this is reflected in an increase in barium concentration in the groundwater. Large withdrawals of groundwater from the Cambrian and Ordovician bed- rock in northeastern Illinois have resulted in the stratigraphic unit with the highest transmissivity, the Ironton-Galesville, having a lower hydraulic head than the overlying and underlying bedrock. During periods when wells are not pumped, there is a cross-formational flow of groundwater in the open borehole in response to differences in hydraulic head between the formations. The lower head in the Ironton- Galesville results in recharge of groundwater to the Ironton-Galesville from overlying bedrock in the open borehole. If the well is also open to the underlying Eau Claire or to the Mt . Simon, then groundwater will flow up the open borehole from these stratigraphic units and into -73- the Ironton-Galesville during periods of nonpumpage. As a result of this cross-formational flow the native groundwater of the Ironton- Galesville may be a small component of the groundwater produced from a well pumped only intermittently for short periods of time. The chemical analyses presented in figure 13 for each of the four wells show a wide range in composition. A comparison of the analyses with pumpage data indicates that the groundwater samples with high barium concentrations represent periods when a greater component of groundwater was from the Ironton-Galesville. Construction of municipal well Hanover Park-4 (fig. 13, la-Id) was completed in 1968. The well was drilled to a total depth of 437 meters (1434 ft) and the borehole was open to stratigraphic units from the Galena through the Ironton-Galesville. The bottom 12 meters (39 ft) of the borehole are in shale of the Eau Claire. Mechanical failure of the well due to the precipitation of secondary barite in the pump occurred in March 1975. The highest barium concentrations occurred in groundwater samples collected during periods when the well was extensively pumped. The highest barium concentration in the available analyses is 7.0 mg/L for a sample collected in August 1976 (fig. 13, lb) . The lowest barite concentration, 0.8 mg/L, was for a sample collected in August 1980 during a period when the well was seldom used. (This analysis is not shown on figure 13.) The higher barium concentrations during periods of large groundwater withdrawals from the well are due to the production of a larger component of the groundwater from the Ironton-Galesville. A significant lateral variation also occurs in addition to the vertical variation in dissolved constituents in the Cambrian and Ordovician bedrock at the location of Hanover Park-4; the well is sited where a regionally significant change occurs in barium and sulfate concentra- tions in groundwater from the Ironton-Galesville. The effect of lateral mixing of groundwater from the Ironton-Galesville increases the super- saturation in the mixed groundwater in the well bore during periods of extensive pumpage. Construction of municipal well Lake Zurich-5 (fig. 13, 4a-4d) was completed in 1963. The borehole of the well is open to stratigraphic -74- units from the Galena through the Ironton-Galesville . The well was drilled 9.8 meters (32 ft) into the Eau Claire, a total depth of 408. A meters (1340 ft). Lake Zurich-5 is the first well on record to require maintenance because of secondary barite precipitation; failure of the well pump occurred in May 1972. Analyses of seven groundwater samples collected from Lake Zurich-5 over the period 1963-1980 show barium concentrations ranging from 0.06 to 2.7 mg/L and sulfate concentrations ranging from 19.3 to 178 mg/L. The highest barium concentrations are in groundwater samples collected in 1971, 1972, and 1978 (fig. 13, 4a-4c) during periods of extensive pumpage. Except for a period of 1978, the well has not been used since 1975 as a source of groundwater supply to the community. The analyses with low barium and high sulfate concentrations are for groundwater samples collected in December 1980 (fig. 13, 4d) and August 1976, during periods when the well was not used as a community water supply. The great variation in barium and sulfate concentrations in the analyses indicates significant differences in the chemical composition of ground- water produced from different stratigraphic units open to the borehole. Native groundwater in the Ironton-Galesville has a high barium concen- tration and a low sulfate concentration. Groundwater with high sulfate concentrations and low barium concentrations is produced from less pro- ductive bedrock that has a higher hydraulic head than the Ironton- Galesville. Unfortunately, there is no information on the chemical composition of groundwater in the region of Lake Zurich in the St. Peter. The groundwater samples with low barium concentrations collected from this well in 1963, 1976, and 1980 may represent groundwater from the St. Peter. Construction of the community well Lake Barrington Shores-2 (fig. 13, 3a-3c) was completed in 1976. The borehole of the well is open to stratigraphic units from the Galena through the Ironton- Galesville. The borehole extends 4.6 meters (15 ft) into the Eau Claire for a total depth of 398 meters (1305 ft). The well failed in March 1980 because of secondary precipitation of barite on the pump and the pump intake screen. Barium concentration in four groundwater samples collected from the well over the period 1976 to 1980 ranged -75- from 0.5 to 8.4 mg/L. The sulfate concentration in the four samples ranged from 10.6 to 314 mg/L. The high sulfate concentration of 314 mg/L (fig. 13, 3b) was de- termined for a groundwater sample collected in May 1980 during a pro- duction test performed after the well was rehabilitated. Lake Barring- ton Shores groundwater is supplied not only from the deep well open to the Cambrian and Ordovician bedrock, but also from a shallow well finished in sand and gravel in the glacial drift with a total depth of 39.3 meters (129 ft). Both wells are connected to a common distribution system; check valves prevent the flow of groundwater from one well down the borehole of the other. The chemical composition of the ground- water sample collected from the deep well in May 1980 closely resembles that of groundwater produced by the shallow well. This analysis sug- gests that the check valve in the distribution system failed and that groundwater from the shallow well was pumped down the borehole of the deep well; mixing of the shallow groundwater containing high sulfate concentrations with groundwater containing high barium concentrations resulted in secondary precipitation of barite. Analyses 3a and 3c (fig. 13) show that groundwater produced from the well in September 1978 and December 1980 was extremely supersaturated with regard to barite. A component of groundwater from the glacial drift cannot be recognized in these chemical analyses. The analyses indicate that there is potential for precipitation of secondary barite within the well because of the mixing of dissimilar groundwaters pro- duced from the Cambrian and Ordovician bedrock. Construction of municipal well Elgin-3A (fig. 13, 2a-2c) was completed in August 1967; this is one of five wells in a well field located on the west side of Elgin. The well was drilled to a depth of 420 meters (1378 ft) and the borehole is open to stratigraphic units from the Galena through the Ironton-Galesville . The bottom 16.1 meters (53 ft) of the borehole is in interlayered shale and dolomite of the Proviso Member of the Eau Claire Formation. A stratigraphic record of Elgin 3-A follows: -76- Strata Quaternary System Pleistocene Series glacial drift Silurian System Alexandrian Series Ordovician System Maquoketa Group shale, some dolomite Galena Dolomite Group Platteville Dolomite Group Ancell Group Glenwood-St. Peter Sandstone Cambrian System Eminence Formation and Potosi Dolomite Franconia Formation dolomite, shale, sandstone Ironton-Galesville Sandstone Eau Claire Formation Proviso Siltstone Member dolomite and shale Thickness (meters) (feet) Depth 48.2 3.7 82.9 30.5 21.3 56.4 16.1 158 12 272 100 70 185 53 (meters) (feet) 48.2 158 51.8 170 63.1 207 114.9 377 80.2 263 195.1 640 17.7 58 212.8 698 295.7 326.2 347.5 403.9 970 1070 1140 1325 420.0 1378 The well has failed several times because of secondary barite precipitation. The first failure occurred in October, 1977, the most recent in January and November, 1982. Analyses of four samples collected from 1972 to 1980 determined that the concentration of barium in groundwater pumped from Elgin-3A ranged from 3.5 to 10.9 mg/L. The sulfate concentration in the samples ranged from 10 to 13 mg/L. The degree of supersaturation in the 1972 samples (fig. 13, 2a) is much less than in samples collected in 1979 and 1980 (fig. 13, 2b and 2c). Downhole geophysical measurements (temperature, conductance, and natural gamma radiation) were made in the open borehole as part of the present research. Downhole groundwater samples were also collected. During these tests, the bottom of the open borehole was measured at 403 meters (1322 ft); the lower 16.8 meters (55 ft) of the well had filled in with rock debris from the wall of the borehole. The 403-meter (1322-ft) depth is 0.9 meters (3 ft) above the contact of the Galesville Sandstone with the Eau Claire Formation. The measurements of temperature -77- and conductance of groundwater in the open borehole determined that groundwater from the upper aquifer units was flowing down the open borehole and into the Ironton-Galesville . The record of the downhole natural gamma radiation measurements determined the extensive presence of anomalously high natural gamma radiation in the borehole. The high radiation values are interpreted as due to the coprecipitation of radium with barium during the secondary precipitation of barite on the rock matrix in the wall of the open borehole. 226 Ra and 228 Ra were present in a sample of barium sulfate precipitate collected from the pump intake check valve. The anomalously high natural gamma radiation was present in the depth interval from 266.7 meters (875 ft) to the bottom of the borehole. Extremely high natural gamma radiation was measured in the open borehole over the interval from 315 meters (1034 ft) to the bottom of the open borehole. High natural gamma radiation was also measured in the depth interval of 213 meters (700 ft) to 218 meters (715 ft) and 227 meters (745 ft) to 235 meters (770 ft) . These zones reflect precipitation of secondary barite at two different setting depths for the well pump intake. In 1982 chemical analyses were performed on groundwater samples collected with a downhole sampler at five discrete depths in the open borehole. A filtered sample (0.45 micron membrane) collected at 198 me- ters (650 ft) where the borehole is open to the Galena and Platteville had a dissolved barium concentration of 0.71 mg/L and a dissolved sul- fate concentration of 30 mg/L. The downhole sample collected at a depth of 259 meters (850 ft) where the borehole is open to the St. Peter Sandstone is a mixture of groundwater from the Galena and Platteville and the St. Peter. A sample filtered at the well site (0.45 micron pore size membrane) con- tained a dissolved barium concentration of 19.0 mg/L and a dissolved sulfate concentration of 12 mg/L. Note the significant difference in barium concentration between the 198-meter (650-ft) and the 259-meter (850-ft) samples. The downhole groundwater sample collected at a depth of 320 me- ters (1050 ft) where the borehole is open to the Eminence and Potosi -78- is a mixture of groundwater produced from the Galena and Platteville, the Glenwood-St. Peter, the Eminence and Potosi. At the time of collec- tion, fine particles of secondary barite were observed to be suspended in the sample. An analysis of an unfiltered sample determined that the barium concentration was 20.4 mg/L and the sulfate concentration was 28.3 mg/L. The downhole groundwater sample collected at a depth of 390 meters (1280 ft) where the borehole is open to the Galesville Sandstone is a mixture of groundwater produced from the Galena and Platteville through the upper section of the Eau Claire. Analysis determined a dissolved barium concentration of 17.0 mg/L and a dissolved sulfate concentration of 22.0 mg/L for a sample filtered at the well site (0.45 micron membrane) The deepest downhole sample was collected at the bottom of the open borehole at a depth of 403 meters (1322 ft). At this depth, a sandpump was used to collect groundwater and sediments from the bottom of the well. The sediments were a heterogeneous mixture of materials that sloughed off the wall of the borehole. Analyses of a sample fil- tered at the well site (0.45 micron membrane) determined a barium con- centration of 8.3 mg/L, a sulfate concentration of 12 mg/L, and a chloride concentration of 80 mg/L. The concentrations of barium and sulfate in the sample are lower than the concentrations in the sample collected at 390 meters. The lower concentrations are thought to be due to the longer residence time of the groundwater in the sloughed material. The high chloride concentration is interpreted to indicate that a portion of this groundwater sample was from the Proviso Siltstone Member of the Eau Claire. The available chemical analyses indicate that a signifi- cant increase in the chloride concentration in groundwater pumped from Elgin-3A occurred between the years 1972 and 1976. Between 1972 and 1976, the chloride concentration increased from 7 to 40 mg/L; an analy- sis in 1980 determined a concentration of 59 mg/L. The concentration of chloride is significantly higher in this well than in other wells in the well field. Elgin-3A is the only well of five in the well field that is ex- periencing pump failure due to precipitation of secondary barite. The pump failures began after the increase in the chloride concentration -79- in groundwater pumped from the well. The increase in chloride concen- tration in groundwater pumped from the well is interpreted as due to the initiation of production of a significant component of groundwater from the Proviso. The production of groundwater from the Proviso is believed to result from a marked decrease in hydraulic head in the Ironton-Galesville due to the large withdrawals of groundwater from the well field. The concentration of sulfate in groundwater produced from the Proviso created extreme disequilibrium with regard to barite solu- bility in the mixed groundwater and resulted in massive barite precipi- tation in the well pump. The cross-f ormational flow of groundwater from the Proviso to the Ironton-Galesville may not be restricted to the open borehole. When the well was constructed in 1967, explosives were used to increase the production of groundwater from the Ironton- Galesville Sandstone. It is possible that the use of explosives caused fractures in the upper bedrock in the Proviso Member. The fractures could be pathways for the flow of groundwater. Municipal well Elgin-3A is a large-capacity well with a sustained yield greater than 4160 liters per minute (1100 gallons per minute). It is located in a well field that is a vital source of water to the community of Elgin. The other four municipal wells in this field have not experienced maintenance problems due to barite precipitation. The other wells are constructed similarly to Elgin-3A; they produce ground- water with high concentrations of dissolved barium. However, the high chloride concentrations are present only in groundwater from well 3A. The chloride concentration in the other wells in the well field on the west side of Elgin should be monitored closely. At the present time, groundwater produced from these wells is greatly supersaturated with respect to barite. Analyses numbers 7 and 8 in figure 13 illustrate the high degree of supersaturation present in groundwater produced from two of the wells. The entry of groundwater from the Proviso Siltstone Member of the Eau Claire Formation into these wells may lead to the secondary precipitation of barite. An increase in chloride concentra- tion would be a signal that the Proviso was contributing groundwater to the wells. -80- Three remedial measures may be effective in controlling the secondary barite precipitation at municipal well Elgin-3A. It is pos- sible that application of any of the three measures may eliminate the problem of well failure due to barite precipitation. 1. Constructing the well with cemented casing to a depth of 232 meters (760 ft) will prevent entry of groundwater with a high concentration of sulfate from the Galena and Platteville. The in- stallation of cemented casing to the 232 meter (760 ft) depth will also prevent entry of groundwater from the Glenwood Formation, a highly varied unit of sandstone, dolomite, and shale overlying the St. Peter Sandstone. The chemical composition of groundwater in the Glenwood is not known. However, there is a possibility that the ground- water from this bedrock unit may have a high sulfate concentration. 2. The installation of 30.5 meter (100 ft) of cemented casing in the depth interval of 293.7 meters (970 ft) to 326.2 meters (1070 ft) will prevent the production of groundwater with a high sulfate concen- tration from the Eminence and Potosi. 3. Deepening the borehole to produce groundwater from the Elm- hurst Sandstone Member of the Eau Claire Formation and the upper sec-' tion of the Mt . Simon Sandstone will reduce the production of ground- water with high concentrations of chloride and sulfate from the Proviso of the Eau Claire. An additional advantage of this measure is that long-term productivity of the well (and the well field) will be in- creased. Lowered pumping costs may also result from a higher pumping water level in the well. It is important not to drill deep into the Mt . Simon where groundwater is very mineralized with high chloride and sulfate concentrations. Construction of the well to a total depth of greater than 580 meters (1902 ft) is not recommended. The downhole methods (geophysical and groundwater sampling) used at Elgin can be used in determining remedial measures to prevent secondary barite precipitation in the wells Hanover Park-4, Lake Zurich-5, and Lake Barrington Shores-2. Analyses of groundwater samples collected from discrete stratigraphic units would permit identification of the bedrock that produces groundwater with high sulfate concentrations. The em- placement of cemented casing from the land surface downward through the -81- St. Peter may be necessary to prevent high maintenance costs at these wells due to secondary barite precipitation; however, the barium con- centrations in groundwater produced from the wells will be high, and the yield of groundwater from the wells may be reduced. An alternative measure for the Lake Barrington Shores well would be to backfill the borehole and seal off production of ground- water from the Ironton-Galesville . This measure would lower the con- centration of barium in groundwater, perhaps to a concentration within compliance of the drinking water regulations. Sealing off the Ironton- Galesville will greatly reduce the yield of the well. However, the St. Peter may be capable of producing a sufficient supply of groundwater for the community. BARIUM IN GROUNDWATER AT LOCATIONS OUTSIDE NORTHEASTERN ILLINOIS An inventory of chemical analyses on public groundwater supplies in Illinois (outside the region of northeastern Illinois shown in figure 11) determined that the barium concentrations in untreated groundwater were generally insignificant. However, at many restricted locations throughout Illinois, analyses of groundwater from public and domestic wells determined barium concentrations greater than 1.0 mg/L. The ana- lytical data available on barium in groundwater of Illinois primarily relates to public water supplies; analysis for barium is rarely performed on groundwater samples collected from domestic or industrial supplies. Because of the limited data base, it is not possible to evaluate ac- curately the occurrence of barium at significant concentrations in aquifer strata that provide domestic supplies of groundwater but do not produce sufficient groundwater for public supplies. The data base on sulfate concentration in groundwater from do- mestic wells is much larger than the data base on barium concentration in groundwater from domestic wells. The sulfate concentration may be used to evaluate the presence of a significant barium concentration. Groundwater that contains a sulfate concentration of 10 mg/L or greater will generally have a barium concentration much less than 1.0 mg/L. A groundwater analysis that could indicate the possibility of a barium concentration of 1.0 mg/L or greater would be a sulfate concentration -82- of less than 10 mg/L, the presence of hydrogen sulfide, and an alka- linity greater than the hardness when both are expressed as equivalents of calcium carbonate. These characteristics indicate an anaerobic en- vironment in which sulfate concentrations may be low because of micro- bially mediated reactions. Barium may occur in the sulf ate-depleted groundwater at high concentrations. However, the presence of barium in groundwater requires a source of barium in the aquifer matrix. Quite often groundwater with a very low sulfate concentration also has a low barium concentration because there is no significant source of barium from the matrix. Analyses of groundwater from public and domestic wells in north- central and northwestern Illinois that produce groundwater from strati- graphic units of the Ordovician System have identified a region of southern Stephenson County, northwestern Ogle County and northern Car- roll County where groundwater from wells finished in the Cambrian and Ordovician bedrock may have barium concentrations greater than the drinking water regulation of 1.0 mg/L. Groundwater with the highest barium concentration (21.0 mg/L) was from a domestic, well in Stephenson County that has the borehole open to the Glenwood-St. Peter Sandstone. A barium concentration of 3.0 mg/L was determined for a groundwater sample from municipal well no. 2 at Shannon in Carroll County that has an open borehole in the interval Galena through St. Peter. Analysis of a groundwater sample from municipal well no. 1 at German Valley in Stephenson County determined a barium concentration of 4.9 mg/L. The borehole of this well is open only to the St. Peter. In the vicinity of German Valley, the Maquoketa is absent and the anaerobic environment in the St. Peter is due to confinement by the Galena and Platteville. Municipal well no. 2 at German Valley is near municipal well no. 1, but is constructed with the borehole open only to the Galena and Platteville. Analysis of a groundwater sample collected from this well determined a barium concentration of 0.3 mg/L. Analysis of groundwater samples from two domestic wells in the vicinity of German Valley that are con- structed with boreholes open only to the Galena and Platteville deter- mined low barium concentrations of 0.1 mg/L. -83- An inventory of the available data has determined the widespread occurrence of groundwater in Illinois with a significant barium concen- tration from aquifers other than rocks of the Cambrian and Ordovician Systems. Aquifer strata include glacial drift, fractured shale and sandstone of the Pennsylvanian System, and dolomite of the Devonian and Silurian Systems. Barium concentrations in untreated groundwater samples range from 0.4 mg/L to 3.9 mg/L. Within Illinois there is a large, widely distributed set of chemical analyses on public supplies of groundwater from sand and gravel in the glacial drift. The analyses indicate that although there is a great range in sulfate concentrations in groundwater from the glacial drift, the barium concentrations are generally very low. Often the sulfate concentrations are low and the analyses indicate than an anaerobic environment is present. However, barium concentrations in groundwater from these wells are also very low, with one known exception. The low barium concentrations probably reflect the absence of a source for barium in the aquifer matrix. The exception is a community well located at Farmer City in De Witt County. The sand and gravel deposit that produces groundwater to the well is overlain and confined by a thick sequence of fine- grained glacial tills. The sand and gravel is located at the base of the glacial materials and overlies shale bedrock of the Pennsylvanian System. The groundwater from these deposits is anaerobic and contains methane gas. The Eh measured with platinum electrodes was -165 milli- volts. The concentration of barium in three groundwater samples col- lected from the well range from 0.7 to 1.2 mg/L. The shale bedrock may be the source for the barium present in the groundwater. -84- SUMMARY NATURAL RADIOISOTOPES IN GROUNDWATER The concentrations of 222 Rn, 226 Ra, 228 Ra, 234 U and 238 U in groundwater were investigated as part of a study of mechanisms re- sponsible for high concentrations of 226 Ra and 228 Ra in groundwater from the Cambrian and Ordovician bedrock in northern Illinois. A lim- ited number of analyses were performed for '- 32 Th and 230 Th in groundwater, The concentration of dissolved 226 Ra in potable groundwater ranged from 0.2 to 36.9 pCi/L. A concentration of 40.2 pCi/L was mea- sured in a brine sample from the Mt . Simon Sandstone. The lowest concentrations are present in primary recharge zones in north-central Illinois where groundwater has a short residence time in the Cambrian and Ordovician bedrock. Along flow paths away from the zone of recharge there is an increase in dissolved 6 Ra concentrations. The highest concentrations occur in mineralized groundwater in locations where the Cambrian and Ordovician bedrock is confined. An important mechanism related to the concentration of 226 Ra in groundwater is the transport of 238 U and 234 U in groundwater, and subsequent secondary accumulation of 238 U, 234 U, and 230 Th on the matrix of the sandstone bedrock. The ionic strength of groundwater is an important control on the desorption of 6 Ra from the matrix. In some localities the dissolved Ra con- centrations are affected by precipitation-dissolution of secondary barite. The concentration of dissolved 228 Ra in potable groundwater from the Cambrian and Ordovician bedrock ranged from 0.4 to 18.9 pCi/L. A concentration of 32.7 pCi/L was measured in mineralized groundwater from the Mt. Simon Sandstone. Over a large part of northern Illinois the dissolved 228 Ra concentrations are less than 3.0 pCi/L. The high concentrations of 228 Ra are present in localities of northeastern Illi- nois. Unlike the concentration of 226 Ra, the concentration of dissolved 228 Ra is not directly related to the ionic strength of groundwater. In some localities the Ra concentrations are affected by precipi- tation-dissolution of secondary barite. A primary control on the -85- concentration of 228 Ra in groundwater may be the abundance of accessory minerals in the bedrock that contain the 232 Th parent. The combined concentration of 226 Ra and 228 Ra in groundwater of the Cambrian and Ordovician bedrock ranged from 2.3 to 50.2 pCi/L. Most analyses exceed the 5.0 pCi/L maximum contaminant level for the combined concentration of the two nuclides. The 6 Ra/ Ra activity ratio ranged from 0.2 to 41.0. The concentrations of the two nuclides in groundwater are not related and a 6 Ra analysis has no validity for prediction of the 8 Ra concentration. The comparison of historical and recent analyses determined that the concentration of 226 Ra and 2 Ra produced from most public supply wells has not changed over time. The gross alpha radiation analyses performed on public supply wells over a 6-year period created a specter that the groundwater was increasing in radioactivity; the variation in the gross alpha concentration (with a general trend of increase) was actually due to an uncontrolled variable in the analytical method. Analyses determined that the high dissolved concentrations of 6 Ra and 2 8 Ra are present in groundwater from sandstones — the productive aquifers in the Cambrian and Ordovician bedrock. Therefore, remedial well construction measures will not lower the concentrations. In north- eastern Illinois the highest dissolved concentrations in potable ground- water are from the Ironton-Galesville Sandstone. Therefore, radium concentrations may increase in the future in groundwater from wells in localities where the aquifers above the Ironton-Galesville are being dewatered by large withdrawals. The mass concentration of uranium in groundwater from the Cam- brian and Ordovician bedrock ranged from <0.002 to 1.10 ug/L. The lowest concentrations were present in groundwater with a low Eh. The combined activity concentrations of 234 U and 238 U ranged from <0.1 to 7.9 pCi/L and were less than the 10.0 pCi/L maximum contaminant level determined for drinking water in a very preliminary study. The 234 U/ 238 U activity ratios for groundwater from the Cambrian and Ordovician bed- rock ranged from 2.1 to 40.7. The lowest activity ratios were present in groundwater in primary recharge zones. Downgradient, in the direction of flow, the activity ratios increase and values greater than 20 occur -86- widely in northern Illinois. The extreme enrichment in 234 U in ground- water was unexpected and the mechanisms responsible are not well under- stood. The high activity ratios are thought to reflect a diffuse secondary accumulation of uranium on the sandstone matrix. The presence of the uranium may be significantly related to recharge events related to glacial processes. The concentration of !22 Rn in groundwater from the Cambrian and Ordovician bedrock ranged from 40 to 670 pCi/L. The concentrations are much less than the proposed maximum contaminant level of 10,000 pCi/L for drinking water. 222 Rn is an inert gas with a short half-life that is generated by the disintegration of 22b Ra. However, the dissolved concentrations of 222 Rn and 226 Ra are not directly related. Some of the highest 222 Rn concentrations are in groundwater with the lowest concentrations of 226 Ra. For all analyses the 222 Rn concentrations ware at least an order of magnitude greater than the 226 Ra concentrations. The 222 Rn concentrations indicate that 226 Ra is ubiquitous on the aquifer matrix and that geochemical mechanisms control the concentration of 226 Ra in solution. Thorium ions have ultratrace solubilities in aqueous systems. , In the present research, seven groundwater samples were analyzed for 3 Th and 3 °Th. The isotopes were not detected in any of the analyses. The limit of detection for the analytical method is 0.01 ug/L expressed as 232 Th. BARIUM IN GROUNDWATER High dissolved barium concentrations in the Cambrian and Ordo- vician bedrock of northern Illinois occur in groundwater anomalously depleted in dissolved sulfate. The depletion of dissolved sulfate is due to bacterially mediated reactions in anaerobic environments of the bedrock. In northeastern Illinois the anaerobic environments exist in aquifer units overlain and confined by the Maquoketa Shale Group. In some localities of north-central and northwestern Illinois where the Maquoketa is absent, anaerobic environments are present where dense, unfractured dolomites of the Galena and Platteville Groups overlie the aquifer units. -87- Th e most extensive area in which high dissolved barium concen- trations occur in the Cambrian and Ordovician bedrock is in northeastern Illinois. Within this area, barium concentrations greater than 5 mg/L are widespread, and concentrations greater than 20 mg/L are measured in groundwater at a few wells. Significantly, the high dissolved barium concentrations are present in the stratigraphic units that are the most productive aquifers — the St. Peter Sandstone and the Ironton-Galesville Sandstone. The available analyses indicate that in northeastern Illi- nois the highest dissolved barium concentrations occur in the most pro- ductive aquifer, the Ironton-Galesville Sandstone. Therefore, remedial well construction measures will not lower barium concentrations in groundwater from high-capacity wells. After depletion of sulfates within the aquifer units, the dis- solved barium concentrations are not controlled by the solubility of barite and may be present at high concentrations. Feldspars present as accessory minerals in the sandstone matrix are a possible source of barium. Maps were constructed showing the distribution of dissolved barium and sulfate in groundwater from the Cambrian and Ordovician bed- rock in northeastern Illinois. Regional planners or municipal water superintendents involved in decisions concerning the location of new wells drilled into the Cambrian and Ordovician bedrock should find these maps helpful. Several wells in northeastern Illinois produce groundwater that is extremely supersaturated with respect to barite. Four wells have required expensive maintenance because of the massive precipitation of barite in the well pump and discharge pipe. A remedial measure that may prevent the barite precipitation is to construct these wells with cemented casing to prevent entry of groundwater from the less productive bedrock units that produce groundwater having high dissolved sulfate concentrations. At many locations in Illinois, barium concentrations greater than 1.0 mg/L are present in anaerobic groundwater from aquifers other than bedrock of the Cambrian and Ordovician Systems. Aquifer strata -88- include fractured shale and sandstone of the Pennsylvanian System and dolomite of the Devonian and Silurian System. Because of the limited data, it is not possible to evaluate accurately the regional occurrence of dissolved barium in groundwater from these aquifers that provide important domestic supplies of groundwater but do not produce the quan- tity of groundwater necessary for public supplies. -89- REFERENCES Argonne National Laboratory, 1981, Appendix B. Radium-induced malig- nancies: Radiological and Environmental Research Division Annual Report, July 1980-June 1981, p. 237-243. Asikainen, Matti, 1981, Radium content and the 226 Ra/ 228 Ra activity ratio in groundwater from bedrock: Geochimica et Cosmochimica Acta, v. 45, p. 1375-1381. Back, William, and Ivan Barnes, 1965, Relation of electrochemical potentials and iron content to ground-water flow patterns: U.S. Geological Survey Professional Paper 498-C, 16 p. Bennett, Dorothy L., Charles R. Bell, and Ira M. Markwood, 1976, Determination of radium removal efficiencies in Illinois wa- ter supply treatment processes: United States Environmental Protection Agency, Office of Radiation Programs, Technical Note ORP/TAD-76-1, 108 p. Berner, R. A., 1963, Electrode studies of hydrogen sulfide in marine sediments: Geochimica et Cosmochimica Acta, v. 27, p. 563-575. Cherdyntsev, V. V., 1971, Uranium-234: Israel Program for Scientific Translations, 234 p. Cherdyntsev, V. V., I. P. Chalov, and G. Z. Khaidarov, 1955, in Transactions: Third Session of the Committee for Determina- tion of Absolute Ages of Geologic Formations, Izv. Akad, Nauk SSSR., p. 175. Clayton, L., and S. R. Moran, 1974, A glacial process-form model, in D. D. Coates [ed.], Glacial Geomorphology : State Univer- sity of New York, Binghampton, NY, p. 89-119. Emrich, Grover H., and H. F. Lucas, Jr., 1963, Geologic occurrence of natural radium-226 in ground water in Illinois: Bull. I.A.S.H., v. 8, n. 3, p. 5-19. Fleischer, Richard L. , 1980, Isotopic disequilibrium of uranium: Alpha-recoil damage and preferential solution effects: Science, v. 207, p. 979-981. Freeze, R. Allan, and John A. Cherry, 1979, Groundwater: Prentice- Hall, Inc., Englewood Cliffs, N J , 604 p. Garrels, R. M. , and C. L. Christ, 1965, Solutions, Minerals, and Equil- ibria: New York, Harper and Row, 459 p. Gilkeson, R. H., and J. B. Cowart, 1982, A preliminary report on uranium-238 series disequilibrium in groundwater of the Cambrian- Ordovician aquifer system of northeastern Illinois, in E. C. Perry, Jr., and Carla Montgomery [eds.], Isotope Studies of Hydrologic Processes: Northern Illinois University Press, De Kalb, 118 p. -90- Gilkeson, Robert H., Eugene C. Perry, Jr., and Keros Cartwright, 1981, Isotopic and geologic studies to identify the sources of sulfate in groundwater containing high barium concentra- tions: Water Resources Center Report no. 81-0165, University of Illinois, Urbana-Champaign, Illinois, 39 p. State Geological Survey Contract/Grant Report 1981-4, 39 p. Gilkeson, Robert H., S. A. Specht, Keros Cartwright, R. A. Griffin, and T. E. Larson, 1978, Geologic studies to identify the source for high levels of radium and barium in Illinois ground- water supplies: A preliminary report: University of Illi- nois at Urbana-Champaign Water Resources Center Research Re- port No. 135, 27 p. Gordon, L. , and K. Rowley, 1957, Coprecipitation of radium with barium sulfate: Analytical Chemistry, v. 29, p. 34-37. Hahn, 0., 1936, Applied Radiochemistry : Cornell University Press, Ithaca, NY, 278 p. Hem, John D. , 1970, Study and interpretation of the chemical char- acteristics of natural water — second edition: U.S. Geological Survey Water-Supply Paper 1473, 363 p. Herman, Janet S., and Donald Langmuir, 1978, Thorium complexes in natural waters (abstract) : Geological Society of America Ab- stracts with Program, 1978 Joint Annual Meeting, v. 10, n. 7, p. 419. Hess, C. T., S. A. Norton, W. F. Brutseart, R. E. Casparium, E. G. Coombs, and A. L. Hess, 1979, Radon-222 in potable water sup- plies in Maine: the geology, hydrology, physics and health effects: Report of the Land and Water Resources Center, University of Maine at Orono, 119 p. Holtzman, Richard B., 1964, Lead-210 (RaD) and polonium-210 (RaF) in potable waters in Illinois: in John A.S. Adams and Wayne M. Lowder [eds.] , The Natural Radiation Environment: University of Chicago Press, p. 227-237. Horick, Paul J., and Walter L. Steinhilber, 1978, Jordan Aquifer of Iowa: Iowa State Geological Survey Miscellaneous Map Series 6, 3 sheets. Hostetler, P. B., and R. M. Garrels, 1962, Transportation and precipi- tation of uranium and vanadium at low temperatures, with special reference to sandstone-type uranium deposits: Economic Geology, v. 57, p. 137-167. Huang, W. H. , and R. M. Walker, 1967, Fossil alpha particle tracks: A new method of age determination: Science, v. 155, p. 1103- 1106. -91- Kempton, John P., and David L. Gross, 1971, Rate of advance of the Woodfordian (late Wisconsinan) glacial margin in Illinois: Geological Society of America Bulletin 82, p. 3197-3200. Kigoshi, K., 1971, Alpha-recoil ' ^Th: Dissolution into water and the 238 U/ 231+ U disequilibrium in nature: Science, v. 43, p. 47-48. Kirby, H. W., and Murrell L. Salutsky, 1964, The radiochemistry of radium: Nuclear Science Series, National Academy of Science-<- National Research Council Series No. 3057, 205 p. Krause, Douglas P., 1960, Radium-228 (MESOTHORIUM I) in Midwest well waters: Argonne National Laboratory Radiological Physics Divi- sion Semiannual Report no. 6199, p. 85-87. Krause, Douglas P., 1959, Radium-228 (MESOTHORIUM I) in Illinois well waters: Argonne National Laboratory Radiological Physics Division Semiannual Report no. ANL-6049, p. 52-54. Krieger, H. L., 1976, Interim radiochemical methodology for drinking water: U.S. Environmental Protection Agency Environmental Monitoring Series EPA-60014-75-008 (Revised), 49 p. Krieger, Herman L. , and Earl L. Whittaker, 1980, Prescribed proce- dures for measurement of radioactivity in drinking water: U.S. Environmental Protection Agency Report no. EPA-600/4-80- 032, 57 p. Kronfeld, Joel, 1974, Uranium deposition and Th-234 alpha-recoil: An explanation for extreme U-234/U-238 fractionation within the Trinity Aquifer: Earth and Planetary Science Letters n. 21, p. 327-330. Lucas, Henry, 1982, 226 Ra and 228 Ra in drinking water: (abstract) Health Physics, v. 43, no. 1, p. 90-91. Lucas, Henry F. , Jr., 1977, Alpha-scintillation radon counting, in E. D. Harwad [ed.], Workshop on Methods for Measuring Radiation in and around Uranium Mills: Atomic Industrial Forum, Washing- ton, DC, p. 69-96. Lucas, Henry F., Jr., and F. H. Ilcewicz, 1958, Natural radium-226 content of Illinois water supplies: Journal of the American Water Works Association, v. 50, p. 1523-1532. Michel, Jacqueline and Willard S. Moore, 1980, 228 Ra and 226 Ra content of groundwater in fall line aquifers: Health Physics, v. 38 (April) , p. 663-671. Miholic, A., 1958, Radioactive waters from sediments: Geochimica et Cosmochimica Acta, v. 14, p. 223-233. -92- Osmond, J. K. , and Cowart, J. B., 1976, The theory and uses of natural uranium isotopic variations in hydrology: Atomic Energy Review, v. 14, n. 4, p. 621-679. Perry, E. C. , Jr., T. Grundl, and R. H. Gilkeson, 1982, H, 0, and S isotopic study of groundwater in the Cambrian-Ordovician aquifer system of northern Illinois, in E. C. Perry, Jr. and Carla Montgomery [eds.J, Isotope Studies of Hydrologic Pro- cesses: Northern Illinois University Press, De Kalb, 118 p. Prichard, H. M. , and T. F. Gesell, 1977, Rapid measurements of 222 Rn concentration in water with a commercial liquid scintillation counter, Health Physics, v. 33, p. 577-581. Robie, R. A., B. S. Hemingway, and J. R. Fisher, 1978, Thermodynamic properties of minerals and related substances at 298.15 K and 1 bar (105 Pascals) pressure and at higher temperatures: United States Geological Survey Bulletin 1452, 456 p. Rosholt, J. N., B. R. Doe, and M. Tatsumoto, 1966, Evolution of the isotopic composition of uranium and thorium in soil profiles: Geological Society of America Bulletin, v. 77, p. 987-1004. Rosholt, J. N., W. R. Shields, and E. L. Garner, 1963, Isotopic frac- tionation of uranium in sandstone: Science, v. 139, p. 224-226. Sasman, Robert T. , Curtis R. Benson, R. Scott Ludwigs , and Tamara L. Williams, 1982, Water-level trends, pumpage and chemical quality in the Cambrian-Ordovician Aquifer in Illinois: Illinois State Water Survey Circular 154, 64 p. Sillen, L. G. , and A. E. Martell, 1964, Stability constants of metal- ion complexes: Chemical Society of London Special Publication 17, 754 p. Standard Methods for the Examination of Water and Wastewater, 13th Edition, 1971: American Public Health Association, Washington, DC, 874 p. Starik, I. J., 1963, Grundlagen der Radiochemie: Akademie Verlag. Stehney, A. F. , 1955, Radium and thorium X in some potable waters: Acta Radiologica, v. 43, p. 43-51. Stumm, Werner, and James J. Morgan, 1970, Aquatic Chemistry: Wiley Interscience, New York, NY, 583 p. Suter, Max, R. E. Bergstrom, H. F. Smith, G. H. Emrich, W. C. Walton, and T. E. Larson, 1959, Preliminary report on ground-water re- sources of the Chicago region, Illinois: Illinois State Geo- logical Survey and Illinois State Water Survey Cooperative Ground-Water Report 1, Champaign, IL, 89 p. -93- Tanner, Allan B., 1980, Radon migration in the ground: a supplementary review, in Thomas F. Gesell and Wayne M. Lowder [eds.], Natural Radiation Environment III: National Technical Information Ser- vice/U.S. Department of Commerce, p. 5-56. Tanner, Allan B., 1964A, Radon migration in the ground: a review, in John A.S. Adams and Wayne M. Lowder [eds.], The Natural Radiation Environment: The University of Chicago Press, p. 161-190. Tanner, Allan B., 1964B, Physical and chemical controls on distribu- tion of radium-226 and radon-222 in groundwater near Great Salt Lake, Utah, in John A.S. Adams and Wayne M. Lowder [eds.], The Natural Radiation Environment: The University of Chicago Press, p. 253-276. U.S. Environmental Protection Agency, 1975, National Primary Drinking Water Regulations: Federal Register, Dec. 24, 1975, v. 40, p. 248. Visocky, Adrian P., 1982, Impact of Lake Michigan allocations on the Cambrian-Ordovician aquifer system: Illinois State Water Survey Contract Report 292, 36 p. Weertman, J., 1972, General theory of water flow at the base of a glacier or ice sheet: Review of Geophysics and Space Physics, n. 10, p. 287-333. Whitfield, M., 1974, Thermodynamics limitations on the use of the platinum electrode in Eh measurements: Limnology and Oceano- graphy, v. 19, p. 857-865. Wickham, S. S., and W. H. Johnson, 1981, The Tiskilwa Till, a regional review of its origin and depositional processes: Annals of Glaciology 2, p. 176-182. Willman, H. B., et al., 1967, Geologic Map of Illinois: Illinois State Geological Survey, Champaign, IL. 1 plate. Willman, H. B., Elwood Atherton, T. C. Buschbach, Charles Collinson, J. C. Frye, M. E. Hopkins, J. A. Lineback, and J. A. Simon, 1975, Handbook of Illinois Stratigraphy: Illinois State Geo- logical Survey Bulletin 75, 261 p.