Se Minnesota Karst

MINNESOTA GEOLOGICAL SURVEY V.W. Chandler, Interim Director

HYDROGEOLOGY OF THE PALEOZOIC BEDROCK IN SOUTHEASTERN MINNESOTA

Anthony C. Runkel

Robert G. Tipping

Minnesota Geological Survey

Minnesota Geological Survey

E. Calvin Alexander, Jr.

Jeffrey A. Green

Department of Geology and Geophysics, University of Minnesota

Minnesota Department of Natural Resources, Rochester

John H. Mossler

Scott C. Alexander

Minnesota Geological Survey

Department of Geology and Geophysics, University of Minnesota

Report of Investigations 61 ISSN 0076-9177

Saint Paul — 2003 iii

HYDROGEOLOGY OF THE PALEOZOIC BEDROCK IN SOUTHEASTERN MINNESOTA

iv

This publication is accessible from the home page of the Minnesota Geological Survey (http://www.geo.umn.edu/mgs) as a PDF file readable with Acrobat Reader 4.0. Date of release: February, 2003

Recommended citation Runkel, A.C., Tipping, R.G., Alexander, E.C., Jr., Green, J.A., Mossler, J.H., and Alexander, S.C., 2003, Hydrogeology of the Paleozoic bedrock in southeastern Minnesota: Minnesota Geological Survey Report of Investigations 61, 105 p., 2 pls.

Minnesota Geological Survey 2642 University Avenue West Saint Paul, Minnesota 55114-1057 Telephone: 612-627-4780 Fax: 612-627-4778 E-mail address: [email protected] Web site: http://www.geo.umn.edu/mgs ©2003 by the Regents of the University of Minnesota All rights reserved. ISSN 0076-9177 The University of Minnesota is committed to the policy that all persons shall have equal access to its programs, facilities, and employment without regard to race, color, creed, religion, national origin, sex, age, marital status, disability, public assistance status, veteran status, or sexual orientation.

v

CONTENTS page

ABSTRACT ............................................................................................................................................ 1 INTRODUCTION ................................................................................................................................ 3 STRATIGRAPHY AND STUDY AREA ........................................................................................... 3 DATA AND METHODS ..................................................................................................................... 3 Hydrostratigraphic analyses .................................................................................................. 3 Hydraulic analyses .................................................................................................................. 10 OVERVIEW OF HYDROSTRATIGRAPHIC, HYDRAULIC, AND HYDROGEOLOGIC ATTRIBUTES ............................................................................................. 14 Hydrostratigraphy ................................................................................................................... 14 Matrix porosity and permeability ............................................................................ 14 Secondary porosity: fractures and dissolution features ........................................ 16 Hydraulic character ................................................................................................................. 22 Hydrogeologic framework ..................................................................................................... 26 HYDROGEOLOGIC ATTRIBUTES OF INDIVIDUAL LITHOSTRATIGRAPHIC UNITS ................................................................................................ 28 MT. SIMON SANDSTONE ................................................................................................................ 28 Hydrostratigraphic attributes ................................................................................................ 28 Matrix porosity ........................................................................................................... 28 Secondary porosity ..................................................................................................... 28 Hydraulic attributes ................................................................................................................ 29 Hydrogeologic synthesis ........................................................................................................ 37 EAU CLAIRE FORMATION .............................................................................................................. 38 Hydrostratigraphic attributes ................................................................................................ 38 Matrix porosity ........................................................................................................... 38 Secondary porosity ..................................................................................................... 38 Hydraulic attributes ................................................................................................................ 40 Hydrogeologic synthesis ........................................................................................................ 41 IRONTON AND GALESVILLE SANDSTONES .......................................................................... 41 Hydrostratigraphic attributes ................................................................................................ 41 Matrix porosity ........................................................................................................... 41 Secondary porosity ..................................................................................................... 41 Hydraulic attributes ................................................................................................................ 41 Hydrogeologic synthesis ........................................................................................................ 44 FRANCONIA FORMATION ............................................................................................................. 45 Hydrostratigraphic attributes ................................................................................................ 45 Matrix porosity ........................................................................................................... 45 Secondary porosity ..................................................................................................... 46 Hydraulic attributes ................................................................................................................ 46 Hydrogeologic synthesis ........................................................................................................ 51

vi

ST. LAWRENCE FORMATION ......................................................................................................... 55 Hydrostratigraphic attributes ................................................................................................ 55 Matrix porosity ........................................................................................................... 55 Secondary porosity ..................................................................................................... 55 Hydraulic attributes ................................................................................................................ 56 Hydrogeologic synthesis ........................................................................................................ 57 JORDAN SANDSTONE ..................................................................................................................... 59 Hydrostratigraphic attributes ................................................................................................ 59 Matrix porosity ........................................................................................................... 59 Secondary porosity ..................................................................................................... 60 Hydraulic attributes ................................................................................................................ 60 Hydrogeologic synthesis ........................................................................................................ 63 PRAIRIE DU CHIEN GROUP ........................................................................................................... 65 Hydrostratigraphic attributes ................................................................................................ 65 Matrix porosity ........................................................................................................... 65 Secondary porosity ..................................................................................................... 65 Hydraulic attributes ................................................................................................................ 67 Hydrogeologic synthesis ........................................................................................................ 71 ST. PETER SANDSTONE ................................................................................................................... 77 Hydrostratigraphic attributes ................................................................................................ 77 Matrix porosity ........................................................................................................... 77 Secondary porosity ..................................................................................................... 79 Hydraulic attributes ................................................................................................................ 79 Hydrogeologic synthesis ........................................................................................................ 79 GLENWOOD FORMATION .............................................................................................................. 81 Hydrostratigraphic attributes ................................................................................................ 81 Matrix porosity ........................................................................................................... 81 Secondary porosity ..................................................................................................... 81 Hydraulic attributes ................................................................................................................ 81 Hydrogeologic synthesis ........................................................................................................ 81 PLATTEVILLE FORMATION ............................................................................................................ 81 Hydrostratigraphic attributes ................................................................................................ 81 Matrix porosity ........................................................................................................... 81 Secondary porosity ..................................................................................................... 81 Hydraulic attributes ................................................................................................................ 82 Hydrogeologic synthesis ........................................................................................................ 83 DECORAH SHALE .............................................................................................................................. 84 Hydrostratigraphic attributes ................................................................................................ 84 Matrix porosity ........................................................................................................... 84 Secondary porosity ..................................................................................................... 84 Hydraulic attributes ................................................................................................................ 86 Hydrogeologic synthesis ........................................................................................................ 86

vii

GALENA THROUGH CEDAR VALLEY GROUPS ...................................................................... 87 Hydrostratigraphic attributes ................................................................................................ 87 Matrix porosity ........................................................................................................... 87 Secondary porosity ..................................................................................................... 87 Hydraulic attributes ................................................................................................................ 89 Hydrogeologic synthesis ........................................................................................................ 92 DISCUSSION: CLASSIFICATION OF AQUIFERS AND CONFINING UNITS ................... 95 SUMMARY ............................................................................................................................................ 96 RECOMMENDATIONS ..................................................................................................................... 97 ACKNOWLEDGMENTS .................................................................................................................... 98 REFERENCES ........................................................................................................................................ 98

Plates 1 and 2 are located in the back pocket of this report.

viii

NOTE ON MEASUREMENTS USED IN THIS REPORT Although the metric system is preferred in scientific writing, certain measurements are still routinely made in English customary units; for example, distances on land are measured in miles and depths in drill holes are measured in feet. Preference was given in this report to retaining the units in which measurements were made. To assist readers, conversion factors for some of the common units of measure are provided below. English units to metric units: To convert from inch inch foot mile

to

multiply by

millimeter centimeter meter kilometer

25.40 2.450 0.3048 1.6093

Metric units to English units: To convert from millimeter centimeter meter kilometer

to inch inch foot mile

multiply by 0.03937 0.3937 3.2808 0.6214

ix

HYDROGEOLOGY OF THE PALEOZOIC BEDROCK IN SOUTHEASTERN MINNESOTA Anthony C. Runkel, Robert G. Tipping, E. Calvin Alexander, Jr., Jeffrey A. Green, John H. Mossler, and Scott C. Alexander

ABSTRACT The Paleozoic bedrock of southeastern Minnesota contains some of the most heavily used aquifers in Minnesota. In this report we characterize the hydrogeologic attributes of these strata by compiling and interpreting a large volume of hydrostratigraphic and hydraulic data. The result is a hydrogeologic framework for southeastern Minnesota that can be used to formulate more effective ground-water management strategies, and in particular it improves our ability to predict aquifer productivity and contaminant transport paths. This report describes the hydrostratigraphic heterogeneity within individual Paleozoic lithostratigraphic units in detail for the first time. Our hydrostratigraphic analysis is based chiefly on plug tests of rock samples, outcrop and core observations of secondary pores, and a number of borehole geophysical techniques. Collectively, this information allows us to define "hydrostratigraphic units"—bodies of rock defined on the basis of their characteristic porosity and permeability—without regard for traditional lithostratigraphic boundaries (Seaber, 1988). Our hydrostratigraphic characterization provides a depiction of the spatial distribution of matrix and secondary porosity in a spectrum of geologic settings across southeastern Minnesota. Of particular importance is our effort to fully integrate the distribution and abundance of fractures and dissolution cavities into the hydrostratigraphic characterization. The Paleozoic bedrock of southeastern Minnesota can be divided into three principal matrix hydrostratigraphic components: coarse clastic rock of high porosity and permeability; fine clastic rock of low porosity and permeability; and carbonate rock, also of low porosity and permeability. All three of these matrix components contain secondary pores such as systematic fractures, dissolution features, and nonsystematic fractures, but they are most abundant in "shallow" bedrock conditions—areas where Paleozoic strata are within about 200 feet of the bedrock surface. In deeper bedrock conditions, secondary pores such as systematic and bedding-plane fractures are known to occur, but their distribution and abundance is poorly understood. They appear to be concentrated along a few discrete stratigraphic intervals, separated from one another by strata with few secondary pores. Hydraulic analyses of Paleozoic strata provide information on the manner in which ground water travels through matrix and secondary pores, and is evaluated in this report based chiefly on interpretation of pump tests, dye-trace studies, borehole flowmeter logs, water chemistry, and potentiometric data within the context of our hydrostratigraphic framework. The groundwater system appears to be relatively simple and predictable in conditions of deep burial by younger bedrock. Under these conditions, coarse clastic strata are of relatively high hydraulic conductivity, typically ranging from a few feet per day to a few tens of feet per day, presumably reflecting flow through large, well-connected intergranular pore spaces. In contrast, the matrix conductivity of the fine clastic and carbonate rock components is low enough in a vertical direction (10-7 to 10-3 foot per day) that intervals dominated by these components can provide hydraulic confinement. Intervals of carbonate rock containing abundant dissolution features have hydraulic conductivity values commonly as high as hundreds of feet per day, and in locally deep bedrock settings, have flow speeds so rapid that they are measured in miles per day along discrete intervals where well-developed conduit systems are present. The enhanced development of secondary pores in shallow bedrock conditions corresponds to a measurable increase in hydraulic conductivity for the Paleozoic bedrock of southeastern Minnesota. Individual layers composed of coarse clastic, fine clastic, or carbonate components in relatively shallow bedrock conditions are very different hydrogeologically from the same layers in relatively deep bedrock conditions because secondary porosity is vastly different.

1

In shallow settings they have higher bulk conductivity, greater range in conductivity, and are likely to transmit the greatest volumes of ground water through conduit networks. Our new hydrogeologic framework for southeastern Minnesota is based on hydraulic data interpreted within the context of the hydrostratigraphic attributes. It differs considerably from previously published frameworks in its classification of regionally extensive aquifers and confining units, and because it places greater emphasis on the importance of flow through secondary pores. Eleven regional aquifers separated by ten confining units are recognized in the bedrock of southeastern Minnesota. The "major" confining units are regionally extensive, relatively thick intervals of fine clastic and carbonate rock that have been demonstrated to be of sufficiently low bulk vertical conductivity to provide confinement under particular conditions of hydraulic stress, and where they are not breached by vertical fractures. The aquifers we define are the bodies of rock dominated by coarse clastic strata or relatively thick intervals of carbonate rock with abundant secondary pores that are known to yield moderate to large volumes of water in deep bedrock settings. The coarse clastic aquifers typically have a bulk horizontal conductivity between 5 and 60 feet per day in deep bedrock conditions. The carbonate rock aquifers are much more variable in hydraulic conductivity, and typically consist internally of relatively narrow intervals of high to very high conductivity (tens to thousands of feet per day) separated by thick intervals of tight carbonate rock that is orders of magnitude lower in conductivity. Our hydrogeologic framework also delineates three major "karst systems," based largely on the work of Alexander and Lively (1995), Alexander and others (1996), and Green and others (1997). A karst system is an integrated mass-transfer system in soluble rocks with a permeability structure dominated by conduits dissolved from the rock and organized to facilitate the circulation of fluid (Klimchouk and Ford, 2000). Southeastern Minnesota karst systems are composed of carbonate-dominated strata where they lie in shallow bedrock conditions. Each karst system is characterized by relatively abundant secondary pores that include large cavities and dissolution-enlarged systematic and nonsystematic fractures, and rapid, direct connections between surface and ground water. The karst aquifers are of particular importance to ground-water management because the ground-water movement through conduits can be rapid and difficult to predict. Our synthesis of the hydrogeologic attributes of Paleozoic bedrock in southeastern Minnesota highlights the need for a better understanding of ground-water flow through secondary pores. Most models of ground-water flow in southeastern Minnesota do not adequately account for the importance of flow through secondary pores in both the aquifers and confining units. In shallow bedrock conditions, the ground-water system may be dominated by relatively rapid movement of water through interconnected networks of secondary pores. The ability of confining units to protect underlying aquifers in such settings has not been carefully evaluated. Furthermore, flow paths and travel times in such conditions are less predictable than commonly depicted in models formulated under the assumption that intergranular flow is dominant. Limited hydrogeologic data for deep bedrock conditions are also not entirely compatible with simple, intergranular flow interpretation. Regional-scale connectivity of secondary pores in deep bedrock settings may provide an enhanced large-scale conductivity to the aquifers and confining beds in southeastern Minnesota that has not been measured by the standard hydraulic tests performed thus far. Researchers are encouraged to analyze both new and existing data in the context of our new hydrogeologic framework to further its development by addressing these and other problems.

INTRODUCTION The Paleozoic bedrock of southeastern Minnesota (Figs. 1, 2) contains some of the most heavily used aquifers in Minnesota. Over one-half of the wells in this part of the state draw water from Paleozoic bedrock, and most municipalities rely entirely on these strata for their potable water supply (County Well Index database maintained by the Minnesota Geological Survey and Minnesota Department of Health). Despite their importance as a source of ground water, the hydrogeologic attributes of these strata have not been comprehensively characterized in a scientifically consistent manner that considers substantial variations in porosity and

2

permeability. In this report we provide such a characterization based on the compilation of the results of a number of studies conducted largely over the past twenty years. The result is a hydrogeologic framework for southeastern Minnesota that is important to environmental managers and scientific investigations because it increases the accuracy and usefulness of ground-water protection plans, and improves our ability to predict aquifer productivity and contaminant transport. In this report, hydrostratigraphic heterogeneity within individual Paleozoic lithostratigraphic units is described in detail for the first time. We define "hydrostratigraphic units"—bodies of rock defined on the basis of their characteristic porosity and permeability—without regard for traditional lithostratigraphic boundaries (Seaber, 1988). Our hydrostratigraphic characterization provides a depiction of the spatial distribution of matrix and secondary porosity in a spectrum of geologic settings across southeastern Minnesota. Of particular importance is our effort to fully integrate the distribution and abundance of fractures and dissolution cavities into the hydrostratigraphic characterization. Our new hydrogeologic framework for southeastern Minnesota (Plates 1, 2; back pocket) is based on hydraulic data such as potentiometric levels, water chemistry, and pump tests, interpreted within the context of the hydrostratigraphic attributes. It differs considerably from previous frameworks in its classification of regionally extensive aquifers and confining units, and in the relatively great importance of flow through secondary pores.

STRATIGRAPHY AND STUDY AREA This report synthesizes the results of a large number of studies that collectively provide a depiction of hydrogeologic attributes of the entire Paleozoic stratigraphic section in a variety of geologic settings across southeastern Minnesota (Figs. 1, 2). The thickness and distribution of individual Paleozoic lithostratigraphic units in southeastern Minnesota (Fig. 2; Mossler, 1987, 1998) are shown on bedrock geologic maps constructed by the Minnesota Geological Survey at scales ranging from 1:24,000 to 1:250,000 (Sloan and Austin, 1966; Olsen, 1982, 1988a; Mossler and Book, 1984; Olsen and Bloomgren, 1989; Mossler, 1990, 1995a, b, 1998, 2001; Mossler and Bloomgren, 1990, 1992; Runkel, 1996a, b, 1998; Mossler and Tipping, 2000). In addition, lithostratigraphic units are delineated for individual water-well sites in the County Well Index database available at the Minnesota Geological Survey.

3

DATA AND METHODS The investigative methods and data synthesized in this report are grouped into one of two major categories: hydrostratigraphic analyses and hydraulic analyses. Hydrostratigraphic analyses provide information about the distribution of porosity and permeability, chiefly through plug tests of rock samples, outcrop and core observations of secondary pores, and a number of borehole geophysical techniques. Hydraulic analyses of the Paleozoic strata provide information on the manner in which ground water travels through pores, and is evaluated in this report based chiefly on interpretation of pump tests, dye-trace studies, borehole flowmeter logs, water chemistry, and potentiometric data within the context of our hydrostratigraphic framework. These methods are described in greater detail below.

Hydrostratigraphic analyses Core analysis—Paleozoic bedrock cores provide information on the character and distribution of hydrostratigraphic units at varying depths of burial beneath the bedrock surface. Porosity and permeability in these cores are described in four manners (Fig. 3, for example): 1. Plug porosity is the measurement of pore space that can be filled with air or water in a rock. Porosity is measured in a laboratory using a small sample of core, typically a one-inch diameter cylinder called a "plug." The porosity value is the percentage of the plug volume that is pore space. Values of porosity typically range from 5 to 30 percent. 2. Plug permeability is a measurement of the ability of a rock to transmit fluid. It is measured in a laboratory using a small sample of core, such as a plug sample. Permeability values provide quantitative calculation of the ability to transmit water through intergranular pore spaces. Vertical permeability measures the ability to transmit fluid in a direction perpendicular to bedding, whereas horizontal permeability reflects the ability to transmit fluid in a direction parallel to bedding. Reservoir geologists consider values less than 5 millidarcies (md) to be very low, representative of "tight" strata. Values greater than 100 md are considered relatively high (Levorsen, 1967). The bulk of our plug-scale porosity and permeability data were synthesized from unpublished reports by the Minnesota Gas Company (Minnegasco), which in the 1970s conducted a subsurface study to assess the feasibility of underground natural gas storage in

MINNESOTA

Enlarged area WISCONSIN

IOWA

MILLE LACS

ILLINOIS

BENTON Mi

Lithograph City Formation, Coralville Formation, and Hinkle and Eagle Center Members of the Little Cedar Formation

S t . Cr

i v er

CHISAGO ANOKA Anoka County Site (unique number 165573)

Bassett Member of Little Cedar Formation and Pinicon Ridge and Spillville Formations Maquoketa and Dubuque Formations and Galena Group Decorah Shale, St. Peter Sandstone, and Platteville and Glenwood Formations

Washington County Site (unique number 227031) RAMSEY

New Brighton HENNEPIN

45°

ATES

Minneapolis & St. Louis Park MCLEOD

Chickasaw Member of Little Cedar Formation ver

A

i

R

ISANTI

Ri

pp

SHERBURNE

WRIGHT

MEEKER

Cedar Valley and Wapsipinicon Groups

WASHINGTON

ss

i

45°30'

PINE

oi x

ss i

STEARNS

KANABEC

CARVER

Lakeland

Prairie du Chien Group Cambrian—Mt. Simon, Ironton–Galesville, and Jordan Sandstones, and St. Lawrence, Franconia, and Eau Claire Formations

Cottage Grove Hastings

Savage SCOTT

Decorah Shale, mapped where possible

DAKOTA Prior Lake

Bedrock of Precambrian and Cretaceous age

Redwing Comprehensive investigations

SIBLEY

Minne

WATONWAN BROWN

GOODHUE

Faribault

Cores i pp

i

WABASHA

Waseca– Waterville

Borehole geophysical studies, including flowmeter logs

BLUE EARTH

Oronoco

WASECA

STEELE

DODGE

H-1 FREEBORN MARTIN

iss

ve r

44°

Miss

Northfield RICE

LESEUR

Ri

NICOLLET

s o ta R i ve

r

44°30'

OLMSTED Rochester

MOWER

FARIBAULT

WINONA Winona

Spring Valley FILLMORE

HOUSTON

Austin Bricelyn 43°30'

B

B'

A'

94°

92° MITCHELL, IOWA

FLOYD, IOWA

93°

4

Paleozoic bedrock of southeastern Minnesota. The raw data collected as part of that work are stored at the Minnesota Geological Survey and cited as "Minnegasco Underground Gas Storage Project," or MUGSP (1980). 3. Visual porosity logs of core establish the stratigraphic position and relative degree of development of cavities and open fractures, which can be the principal ground-water conduits in bedrock. Such logs account for the pore spaces that are larger than those measured in most plug samples (plug samples are typically collected from intervals without visible fractures and cavities). Values of nearly 100 percent correspond to well developed bedding-plane fractures/dissolution features in core. Lesser values of porosity represent the percent of core that consists of open cavities based on a visual estimate (Fig. 3, for example). Although permeability is a feature that cannot be visually estimated, ground-water conduits with relatively high hydraulic conductivity should be expected to have a high visual porosity if they are intersected by core. Visual porosity in cores was estimated in such a manner that the logs probably underrepresent the abundance of secondary pores, especially large cavities. Intervals of core loss and breaks between cores, which may correspond to such features, were recorded as secondary

pores only if the core ends showed clear evidence of dissolution or mineralization, indicating the presence of a cavity. Furthermore, the abundance of vertical, systematic fractures in the deep subsurface (described later in this report), which may be hydraulically important features at some scale, is probably underestimated because individual vertical cores and boreholes have a small probability of intersecting such features. 4. Vertical fracture abundance is a visual estimate of the number of subvertical fractures per foot of core. These are mostly "mesoscopic" fractures (Price and Cosgrove, 1990)—irregular, sinuous fractures that typically cannot be traced more than a few inches. The apertures are largely healed or open only a fraction of a millimeter, and their hydraulic significance is not known. However, such fractures can be preferential pathways for relatively slow-moving ground water in strata that are otherwise of negligible permeability (Watts, 1983; Lorenz and Finley, 1991). Additionally, these narrow fractures may be preferentially opened compared to nonfractured rock when subjected to stress-release and weathering in near-surface conditions. Therefore the stratigraphic position of mesoscopic fractures in deep core may correspond to intervals where secondary pores are preferentially developed in near-surface bedrock conditions.

Figure 1. Map of southeastern Minnesota showing the distribution of Paleozoic lithostratigraphic units where they occur as the uppermost bedrock and locations of cores, borehole flowmeter studies, and sites of comprehensive hydrogeologic studies referenced in this report. The comprehensive hydrogeologic studies utilize a number of techniques and include formally published aquifer characterization of the Eau Claire through lower St. Lawrence Formations as part of the Aquifer Thermal Energy Storage Project (ATES Project; Miller and Delin, 1993), and of the Galena through Cedar Valley Groups in Floyd and Mitchell Counties, northern Iowa (Libra and Hallberg, 1985; Witzke and Bunker, 1985). Informally published site remediation investigations include work focused on the Ironton–Galesville Sandstone and Franconia Formation near Lakeland, Minnesota (Braun Intertec, 1992; Delta Environmental Consultants, Inc., 1992); the Jordan Sandstone and Prairie du Chien Group at an abandoned landfill near Oronoco (Alexander, 1990; Donahue and Associates, Inc., 1991; RMT, Inc., 1992), and in the Arden Hills–New Brighton area (Camp, Dresser and McKee, 1991); the Platteville Formation in St. Louis Park (for example ENSR International, 1991) and northeast Minneapolis (for example Barr Engineering, 1991); and the Galena Group and Dubuque and Maquoketa Formations at the Spring Valley Amoco terminal (Delta Environmental Consultants, Inc., 1995). The Ironton–Galesville Sandstone and Franconia Formation were also studied at a proposed expansion of an ash disposal site near Red Wing (Wenck and Associates, Inc., 1997). The heavy lines outline the seven-county Twin Cities Metropolitan area, and the locations of crosssections on Plates 1 and 2 also appear.

5

0

100

Lower Ordovician (505–478 m.y.)

57-61

Coralville Formation Dcum

15-43

25-35

Hinkle & Eagle Center Mbrs

Chickasaw Member Dclc

40-70

Little Cedar Formation

Cedar Valley Group

Dcuu

Bassett Member

Shakopee Formation

Opsh

Oneota Dolomite

Opod

20-47

Pinicon Ridge Formation

Omaq

Dubuque Formation

Odub

St. Lawrence Formation

110-120

Maquoketa Formation

Jordan Sandstone

36-85

Dspl

65-70

21-84

Coon Valley Member

Spillville Formation

23-40

Wapsipinicon Group

Dclp

PALEOZOIC

Middle Devonian (387–374 m.y.)

Lithograph City Formation

Up to 50

API-G units

320-340

Increasing count

Prairie du Chien Group

Lithology

Thickness (in feet)

Series

Group, Formation, Member

Label

Natural gamma log

Cstl G G

45-50

Opvl Ph

Ph Ph Fe

Ostp

Fe

G

155-160

G G

Cfrn

G G G G

Birkmose Member

G G

G

Ironton and Galesville Sandstones

40-45

Franconia Formation

G

Reno and Tomah Members

Reduced scale 50%

Cigl G

G

G

G G

Eau Claire Formation

Cecr

Mt. Simon Sandstone

Cmts

G

G

G G G

200

St. Peter Sandstone

Fe

Odcr

Ogwd

G

Upper Cambrian (523–505 m.y.)

70-75 Fe

Platteville Formation Glenwood Formation

PALEOZOIC

75-85

Ogpr

Cummingsville Ogcm Formation

Decorah Shale

G

40-50

Prosser Limestone

Ogsv

5-6 20-30

Stewartville Formation Galena Group

Upper Ordovician (458–444 m.y.)

G

70-80

Era

Figure 2. Standard bedrock stratigraphic column showing Paleozoic lithostratigraphic units of southeastern Minnesota and typical gamma log. Modified from Mossler (1987, 1998). Hydrostratigraphic components are depicted in Plates 1 and 2. Figure explanation is on the following page.

6

200 or less

Continued above right

EXPLANATION Cavities (commonly filled with coarse calcite)

Limestone Dolostone

Chert K-bentonite bed (altered volcanic ash bed

Sandy

Sandstone

Oolites

Very fine- to fine-grained

Worm bored Pebbles (gravel in unconsolidated units) Flat-pebble conglomerate Cross-bedded (festoon)

G

Glauconite

Cross-bedded (planar to tangential)

Fe

Iron stain

Ripple cross-laminations

Ph

Phosphate pellets

Dolomitic

Algal mats

Calcareous

Siltstone

Algal domes; stromatolites

Shale

Fossiliferous; fossils (symbols not used in limestone or dolostone units)

Contact marks a major erosional surface

Fine- to medium-grained Medium- to coarse-grained Shaly

Field observations—A number of outcrop-based stratigraphic and sedimentologic investigations conducted over the past 50 years delineate facies that are now known to differ considerably from one another in intergranular porosity and permeability (for example Berg, 1954; Nelson, 1956; Setterholm and others, 1991; Runkel, 1996a, b, 1999, 2000; Runkel and Tipping, 1998; Runkel and others, 1999). Those investigations, supplemented with field work conducted as part of recent Minnesota Geological Survey mapping in southeastern Minnesota (for example Mossler and Book, 1984; Mossler, 1990, 1995a, b, 1998, 2001; Mossler and Bloomgren, 1990, 1992; Runkel, 1996a, b, 1998; Mossler and Tipping, 2000), allow us to delineate individual hydrostratigraphic units within mapped lithostratigraphic units across the outcrop belt of Paleozoic bedrock where cores are generally scarce. Additionally, outcrops provide an opportunity to examine secondary pores in Paleozoic bedrock, and their interaction with surface waters. The abundance, size, and stratigraphic distribution of fractures and dissolution features were described for much of the Paleozoic section in representative large outcrops, quarries, and road cuts along the Mississippi River and its tributaries. An interval of strata in which secondary porosity is preferentially developed in outcrop (Fig. 4) can be an important ground-water conduit in saturated subsurface conditions (for example Gianniny and others, 1996). The distribution of springs and sinkholes (for example Alexander and others, 1996; Witthuhn and Alexander, 1996) also provides insight into stratigraphic control of hydraulically

7

Facies change

important fractures and dissolution cavities in nearsurface bedrock settings. Borehole logs—Natural gamma logs have been used extensively by Minnesota Geological Survey scientists to distinguish hydrostratigraphic components that differ from one another in intergranular porosity and permeability (for example Setterholm and others, 1991; Runkel, 1996b). A slimline probe measures gamma rays naturally emitted by rocks as it is slowly raised in a borehole. In the Paleozoic strata of southeastern Minnesota, fine-grained siliciclastic rocks with low intergranular permeability contain potassium in sufficient abundance to emit relatively high levels of gamma rays, and therefore cause strong positive deflection on gamma logs (Fig. 2). Coarse-grained siliciclastic rocks with higher permeability have low potassium content and therefore correspond to low readings on the gamma logs. Carbonate strata most commonly have readings between those of fine and coarse siliciclastic rocks. Borehole video, borehole televiewer (BHTV), and caliper logs provide information similar to that of rock cores in that they are used to document the size, shape, and stratigraphic position of fractures and dissolution features. Such logs are available at several state agencies, including the Minnesota Department of Health, Minnesota Pollution Control Agency, and the Minnesota Geological Survey. Borehole cuttings—High-quality sets of cuttings were used in conjunction with outcrop study to demonstrate a correspondence between gamma log

100%

0

5

0

40% 10-6

Position of sampled plug

Plug vertical permeability

Plug porosity

Fractures per foot

Visual porosity 0

100 md

Mesoscopic fracture

1 foot

Figure 3. Example of a presentation of plug porosity and permeability values, and logs of visible porosity for an 8-foot core of fine clastic and carbonate rock in the St. Lawrence Formation collected from the Waseca–Waterville area. Lines are drawn from dissolution features in the core to corresponding tick marks on the visual porosity log.

8

Fracture

Figure 4. Large, interconnected dissolution cavities parallel to bedding in the carbonate strata along the Shakopee Formation–Oneota Dolomite contact at a quarry near Red Wing in Goodhue County. Note the vertical fracture with a large aperture in the quarry wall on the left side of the photo. Water commonly travels rapidly downward through such fractures and subsequently travels laterally along bedding-plane parallel conduits such as the interconnected dissolution cavities shown here (marked by arrows). The short vertical line below the cavities is approximately 5 feet tall.

signatures and hydrostratigraphic units (for example Setterholm and others, 1991). This correspondence was successfully used to delineate subsurface hydro-stratigraphic units in the Rochester area (Runkel, 1996b) and in Houston and Goodhue Counties (Runkel, 1996a, 1998). Cuttings alone typically cannot be used to determine the precise thickness of hydrostratigraphic units because the sample stream from the drilling process has inherent inaccuracies related to poor collection methods and recirculation problems.

Hydraulic analyses Pump and slug tests—A large database of hydraulic conductivity values is based on a compilation of pump and slug tests conducted on Paleozoic bedrock in southeastern Minnesota and adjacent states. They can be grouped into three principal categories based on the quality of the test and amount of associated supplementary information on borehole construction, testing procedures, and geologic setting. 1. Discrete interval tests—Hydraulic data available from comprehensive hydrogeologic reports that describe controlled pump tests as well as detailed

9

stratigraphic and well-construction information are considered to be of the highest quality used in this report. They can be used to calculate the hydraulic conductivity of individual hydrostratigraphic units with confidence. Frequently cited examples of these kinds of studies include discrete-interval packer testing of Cambrian siliciclastic strata by Nicholas and others (1987) and Miller and Delin (1993), and of Ordovician and Devonian age, carbonate dominated strata by Libra and Hallberg (1985), Nicholas and others (1987), Graese and others (1988), Donahue and Associates, Inc. (1991), and Delta Environmental Consultants, Inc. (1995). 2. Specific capacity tests—Specific capacity data obtained from water well construction reports in the County Well Index database were used to calculate hydraulic conductivity following an approach described by Bradbury and Rothschild (1985). Specific capacity values are corrected for the effects of partial penetration, well loss, and borehole diameter. The hydraulic conductivity values calculated in this manner are

believed to be a more accurate measure of aquifer performance than specific capacity values alone. Hydraulic conductivity has been calculated for 8,626 wells that draw water from the Paleozoic strata of southeastern Minnesota. Runkel (2000) demonstrated that a large database of such values can be used to recognize relative differences in aquifer performance that are consistent with the results of higher quality, controlled pump tests and therefore can provide information about geologic controls on aquifer performance. Except where otherwise noted, our database excludes wells constructed to draw water from more than one of the eleven aquifers defined in this report. These data are summarized in this report chiefly as scatter plots, and as box plots that show median values and statistically acceptable ranges. Conductivity values calculated from specific capacity tests may be less indicative of hydraulic performance than high quality, discrete-interval pump and slug test data because pumping rates and drawdown measurements are typically collected in a less rigorous fashion, and because the tests are usually of short duration. Additionally, the database from which these hydraulic conductivity values have been calculated consists of tests of water wells constructed expressly for the purpose of extracting economic quantities of water. The values of conductivity are therefore chiefly representative of the most productive intervals of Paleozoic strata in a given geologic setting, and do not include a large sample of values representative of strata with relatively low conductivity. 3. Standard aquifer tests—A large number of hydraulic conductivity values for individual hydrostratigraphic units are based on aquifer tests conducted by private consultants and by staff from state and federal agencies that include the U.S. Geological Survey, the Division of Waters of the Minnesota Department of Natural Resources, the Minnesota Department of Health, and the Minnesota Pollution Control Agency. These are cited as "standard aquifer tests" in this report to distinguish them from specific capacity tests and discrete interval tests accompanied by higher quality ancillary information described above. Standard aquifer pump test results are typically not accompanied by reports in which the raw pump test data nor pumping procedures are provided, and detailed hydrostratigraphic context is not available for the wells in this database. However, we have used drilling

10

records and natural gamma logs (where available) to roughly determine the hydrostratigraphic components exposed in the open-hole interval for each of the wells in the database. Our evaluation of the results of these aquifer tests indicates that they commonly yield hydraulic conductivity values that are higher than those calculated on the basis of discrete interval and specific capacity tests of the same hydrostratigraphic material. A possible explanation is that standard aquifer tests in southeastern Minnesota have most commonly been performed on large diameter industrial and municipal wells that are developed to increase productivity through methods such as blasting. Large, high capacity wells such as these may be better connected to secondary pore networks compared to narrower diameter, undeveloped boreholes subjected to packer tests, and to small diameter domestic wells that compose the majority of our specific capacity database. Borehole geophysical and video logs—Vertical groundwater flow within a borehole in saturated stratigraphic intervals can be detected by electromagnetic and heat pulse flowmeters. Flowmeter logs collected under ambient conditions are used to recognize the hydraulically dominant intervals of matrix and secondary pores in an individual borehole, and the confining unit(s) that separates them (Fig. 5). Some boreholes are also flowmeter logged during stressed conditions created by pumping from, or injecting water into, a borehole. Flowmeter logs collected under stressed conditions allow the hydraulic properties of discrete intervals to be quantified when compared to ambient flowmeter measurements and accompanied by ancillary information such as the change in potentiometric level of the borehole. These techniques are explained in greater detail in Paillet and others (2000). A borehole video camera and a multi-parameter probe that measures temperature, pH, and chloride provide information similar to that of flowmeters, but in a more qualitative and inconsistent fashion. Video logs can be used to identify seeps and cascading water along discrete intervals in open boreholes above the static water level. Water entering or exiting a borehole along discrete conduits in saturated conditions can be recognized on video logs by the movement of well sediment held in suspension and by shifts in temperature, pH, and chloride content measured by the multiparameter probe.

Dye tracing—Dye-trace investigations have been successful in providing quantitative measures of ground-water flow speeds, and the degree of vertical connectivity across adjacent hydrostratigraphic units in near-surface bedrock conditions where flow along secondary pores is of particular importance (for example Wheeler, 1993; Alexander and others, 1996). Flow speeds are typically expressed as nominal flow rates, in feet per day or miles per day, and are lower limits on the true flow velocities. Water chemistry—Chemical constituents such as tritium, nitrates, and chlorides have commonly been used to determine flow paths and hydraulic connection between water-bearing bodies of rock. A large volume of ground-water chemistry data for southeastern Minnesota are scattered among several state agencies and private consultants, and in a number of publications. In this report we focus on ground-water chemistry data collected and interpreted as part of site-specific studies in which the geologic setting, well construction, and hydrostratigraphic attributes are well understood. Potentiometric data—This report incorporates potentiometric data compiled from the results of site-specific studies that include the information necessary to interpret the data within our hydrostratigraphic framework. We also cite published county and larger-scale potentiometric maps that provide water level information that can be used in the context of our hydrostratigraphic framework. We use a difference in static water levels (heads) above and below a low permeability hydro-stratigraphic unit as one line of evidence that the unit provides confinement. Previous hydrogeologic investigations in southeastern Minnesota have struggled with the question of what constitutes a significant difference in head for the purpose of recognizing discrete hydrogeologic units. For example, is a five-foot difference significant at the scale of a site-specific study? Is it significant at a county or regional scale? Historically, hydrogeologic units have been defined regionally, using potentiometric data from water wells, with elevations determined using 7.5-minute topographic maps. Under these conditions, a fivefoot head difference is smaller than the error associated with a well elevation. Using these methods, individual hydrogeologic units cannot be distinguished. Unfortunately, regionally defined hydro-geologic units have been applied to sitespecific studies, where a five-foot head difference is important in distinguishing ground-water flow

11

paths. Head differences across a confining unit occur under conditions of stress: either natural due to ground-water flow patterns, or induced by pumping. Certain conditions of stress can cause aquifers separated by a confining unit to have similar heads even though they are not hydraulically wellconnected. For example, aquifers that are recharged and discharge near the same elevation may not show large head differences between zones of recharge and discharge. In this way, confining characteristics of individual hydrostratigraphic units are not revealed using potentiometric data alone. We infer that a hydrostratigraphic unit that provides confinement at an individual site has the ability to provide confinement elsewhere, because by definition it has more or less consistent properties of porosity and permeability across its extent. Whether or not its confining properties are breached by fractures is a question that potentiometric data can help answer on a site-by-site basis.

OVERVIEW OF HYDROSTRATIGRAPHIC, HYDRAULIC, AND HYDROGEOLOGIC ATTRIBUTES Hydrostratigraphy Matrix porosity and permeability The Paleozoic strata of southeastern Minnesota can be generally divided into three distinct hydrostratigraphic components based entirely on matrix characteristics (Runkel and Tipping, 1998; Runkel, 1999). The components are: coarse clastic, fine clastic, and carbonate rock (Fig. 6; Plates 1, 2). The values for matrix porosity and permeability of these three components where they occur in settings with relatively minor development of secondary porosity (fractures and dissolution features) have been determined at the smallest scale through laboratory testing of plug samples (Norvitch and others, 1973; MUGSP, 1980; Setterholm and others, 1991; Walton and others, 1991; Wenck and Associates, Inc., 1997). The coarse clastic component is a poorly cemented, moderately to well-sorted, fine- to coarse-grained sandstone composed of about 98 percent quartz. Plugsample tests indicate it has a high porosity and vertical permeability, commonly more than 20 percent and 1,000 md, respectively, due to relatively large, well-connected intergranular pore spaces. Horizontal permeability typically is equal to, or as much as an order of magnitude greater than vertical permeability. The fine clastic component consists of very fine-

12

D.

Logged borehole

Logged borehole

Open hole

Cased hole

Static water level

water enters hole, travels up

water exits hole

A.

0

Aquifer

Confining unit

_

+

Flowmeter

Interpretation line

E. 400

500

450

400

350

300

250

200

1

Casing bottom

0

2

Trolling flow, ambient conditions (gallons per minute) 0

Interpretation line

1

Stationary flow, ambient conditions (gallons per minute)

0

+

_

Flowmeter

0

+

_

Logged borehole

Flowmeter

B.

2

Logged borehole

No measurable flow

Water enters through intergranular pores and travels uphole

Consistent upflow past confining unit

Water enters abruptly at three thin intervals of secondary pores. Entrances are separated by confining units.

Consistent, relatively strong upflow past confining unit

Water exits abruptly at two thin intervals of secondary pores separated by confining unit

Consistent upflow along upper part of open hole and into casing (water exits higher in casing)

Ambient conditions: explanation of flow

water exits hole

water enters hole, travels down

C.

_

0

+

Flowmeter

F.

Stationary flow, ambient conditions (gallons per minute)

Trolling flow, ambient conditions (gallons per minute) -6

-2

-4

0

-4

-6

-2

Ambient conditions: explanation of flow 0

400 No measurable flow in casing, nor in upper part of open hole

Casing bottom Interpretation line

450

Water enters abruptly at five thin intervals of secondary pores and travels downhole. Entrances are separated by confining units.

500

Consistent, relatively strong downflow past confining unit

550

Water exits gradually through intergranular pores

600 Consistent weak downflow past confining unit Water exits abruptly No measurable flow

G.

Trolling flow, during injection (gallons per minute) -10

-5

500 Casing bottom

0

-15

-10

-5

0

5

Interpretation lines Ambient flow

550

600

650

Ambient conditions: explanation of flow

Stationary flow (gallons per minute)

Stressed conditions: location and percent of borehole transmissivity of dominant permeable intervals during injection

No measurable flow in casing, nor in upper part of open hole

Injected water travels downhole along casing and upper part of open borehole with no loss

Water exits at fracture Consistent upflow past confining unit

Partial loss of injected flow: exits through fractures—18% of transmissivity Consistent downflow of remaining injected water

Water enters through intergranular pores and travels uphole

Loss of remaining injected water: exits through intergranular pores—82% of transmissivity

No measurable flow

No measurable flow

Injection flow

Figure 5. Flowmeter logs are a depiction of vertical water movement in a borehole: positive values on the logs correspond to flow up a borehole, negative values correspond to flow down a borehole, and zero represents no measurable flow. Ambient borehole flow in a vertical direction is driven by vertical hydraulic gradient. Trolling flowmeter logs are a continuous record of flow measured by a slimline electromagnetic probe as it is raised at 10 feet per minute up the borehole. Stationary logs show a series of flow measurements taken at various depths in the borehole with the probe stopped, or "stationary." These two kinds of flowmeter logs are used in conjunction with geophysical logs that measure physical rock properties (such as gamma, caliper, video, and BHTV logs) to interpret flow conditions in the borehole, shown graphically as an "interpretation line" on the stationary logs. A. Schematic hydrogeologic setting and corresponding stationary flowmeter log. Flowmeter logging records no vertical borehole flow because the open borehole exposes only a single aquifer with no vertical gradient. Figure 5 explanation continued on page 14 Figure 5 explanation continued from page 13

13

B. Schematic hydrogeologic setting and corresponding stationary flowmeter log. Flowmeter logging records no vertical flow in the borehole, even though the hole fully penetrates a confining unit, because the aquifers above and below the confining unit have similar heads. C. and D. Schematic hydrogeologic settings and corresponding stationary flowmeter logs. Flowmeter logs show vertical flow that occurs in boreholes that intersect two (intergranular) aquifers with heads that differ from one another. E., F., and G. Flowmeter logs collected in southeastern Minnesota and used in this report. Changes in magnitude of vertical flow along the interpretation line mark permeable intervals through which water enters (inflow) or exits (outflow) the borehole. Abrupt changes in magnitude of vertical flow correspond to relatively thin intervals of hydraulically active secondary pores, most commonly bedding-plane fractures; gradual changes correspond to intervals where intergranular flow is dominant. The beds that separate these hydraulically active intervals are of relatively low permeability and can be considered confining units at the scale of the immediate vicinity of the borehole. The confining units that directly separate an entrance from an exit maintain differential heads above and below them, which drives ambient borehole flow. Confining units that separate successive entrances or exits along a borehole may or may not separate heads that differ from one another. G. also provides an example of flowmeter logging under stressed conditions. The borehole was injected with water at a rate of 9 gallons per minute. The relative transmissivity of the two permeable intervals that accommodate the injected water is quantified to the right of the column following the procedure described by Paillet and others (2000).

grained sandstone, siltstone, and shale in thin to medium beds that are strongly to moderately cemented. This component has very low to low relative permeability, several orders of magnitude less than that of the coarse clastic component described above. Plug tests indicate a vertical permeability that typically ranges from 10-6 to 10-2 md. Horizontal permeability is commonly about two orders of magnitude greater than vertical. The carbonate rock component consists of very fine- to fine-grained dolostone and limestone with variable amounts of silt, sand, and shale as interbeds or admixed in the carbonate matrix. Matrix porosity and vertical permeability values are typically less than 15 percent and 10-1 md, respectively. Limited tests of horizontal permeability indicate that it is commonly about two orders of magnitude greater than vertical permeability in laminated carbonate rocks. Horizontal permeability is probably roughly equal to vertical permeability in plug samples of structureless carbonate rocks such as those common in the Oneota Dolomite.

Secondary porosity: fractures and dissolution features The hydrostratigraphic character of the three components described above is affected by lateral and vertical variability in the abundance and interconnectivity of fractures and dissolution features (Plates 1, 2). Calculated values for porosity and permeability within each of these components can vary

14

substantially depending on the scale of the tested rock sample, and the degree of development of fractures and dissolution features. Permeability is very high where such features are well developed and interconnected, and very low, even on a large scale, where minimally developed (for example Liesch, 1973; Libra and Hallberg, 1985; Graese and others, 1988; Gianniny and others, 1996; Eaton and others, 2000). Core logging, borehole videos, geophysical logs, dye-trace investigations, and field observations of exposed bedrock in southeastern Minnesota presented in this report, and studies of generally similar sedimentary bedrock in other parts of North America (for example Ferguson, 1967; Nichols, 1980; Wyrick and Borchers, 1981; Graese and others, 1988) suggest that bedrock conditions can be separated into two general categories based on the nature of secondary porosity: "shallow" bedrock conditions, and "deep" bedrock conditions (Fig. 7; Plates 1, 2). Shallow bedrock conditions differ from deep conditions because they have a relatively high density of large, wellconnected fractures and dissolution cavities. Shallow bedrock conditions are characterized by relatively strong development of three kinds of secondary pores (Fig. 8). Systematic fractures are flatsided openings oriented perpendicular to bedding. They are also referred to as "joints," and are typically the most prominent fractures in large outcrops—commonly

A.

D.

C.

B.

Figure 6. Examples of the three principle matrix hydrostratigraphic components in core. A. Coarse clastic component from the Mt. Simon Sandstone consisting of medium- to coarse-grained, friable sandstone. B. Fine clastic component in the Eau Claire Formation. Consists of very fine-grained sandstone and siltstone with thin shale laminations. C. Fine clastic component in the Franconia Formation. Consists largely of shale (dark beds) with interbedded very fine-grained sandstone and siltstone. D. Carbonate rock in the Platteville Formation. The core on the right has thin, irregular interbeds of shale. Cores are from Ramsey County (ATES Project, cores AC-1 and BC-1).

15

evident at distances of hundreds of feet as straight, vertical openings with a more or less consistent spacing. The walls of systematic fractures typically have strike orientations that fall within one or two tightly clustered sets (Olsen, 1988b; Ruhl, 1995; Runkel, 1996a). Nonsystematic fractures are more randomly distributed and more variable in their orientation and shape. They include openings that parallel bedding planes as well as irregular, curved, or conchoidal fractures that intersect bedding obliquely. Both systematic and nonsystematic fractures are common in all three matrix hydrostratigraphic components where they occur in shallow bedrock conditions. Dissolution features are a secondary pore developed through the dissolution of carbonate rock. Dissolution can enlarge the apertures of nonsystematic and systematic fractures, and can also create cavities that have no apparent relationship to fractures. In shallow bedrock conditions, the permeability of the coarse clastic, fine clastic, and carbonate rock components may be several orders of magnitude higher than that of deep conditions at scales greater than that of plugs because of the greater development of these three kinds of secondary pores (for example Donahue and Associates, Inc., 1991; Gianniny and others, 1996; Wenck and Associates, Inc., 1997). Our understanding of secondary porosity in deep bedrock conditions relies mostly on examination of cores (Fig. 9, for example) and borehole video and caliper logs collected from southeastern Minnesota (Donahue and Associates, Inc., 1991; Walton and others, 1991; Delta Environmental Consultants, Inc., 1995; Runkel, 1999; Runkel and others 1999; this study) and on recent studies of analogous Paleozoic bedrock settings in Iowa, Wisconsin, Illinois, and Michigan (Witzke and Bunker, 1984; Graese and others, 1988; Hurley and Swager, 1991; Gianniny and others, 1996; Eaton and others, 2000). Collectively, this information suggests that deep bedrock conditions differ fundamentally from shallow conditions in that secondary pores are diminished in abundance, size, and degree of interconnectivity, principally because dissolution features and nonsystematic fractures are less common. Our limited borehole data in southeastern Minnesota indicate that open nonsystematic fractures and macroscopic dissolution cavities are apparently uncommon to absent in the fine clastic and coarse clastic components. Where present, discrete bedding-plane fractures are separated by tens of feet of strata with no evident secondary pores, and some individual cores and boreholes have no recognizable bedding-plane fractures across hundreds of feet of strata. Carbonate rock in deep settings varies in its development of secondary porosity (Fig. 9). Some carbonate intervals,

16

such as the lower two-thirds of the Oneota Dolomite, the Platteville Formation, and Galena Group, have relatively few open fractures and macroscopic cavities compared to their character in shallow bedrock conditions (Graese and others, 1988; Delta Environmental Consultants, Inc., 1995; Tipping and Runkel, 2001; this report). In contrast, core and borehole video logs analyzed in this report (and by Tipping and Runkel, 2001) demonstrate that other carbonate intervals in deep bedrock settings, such as much of the Shakopee Formation, parts of the St. Lawrence Formation, the uppermost Oneota Dolomite, and a thin, carbonate-rich interval in the lower part of the Franconia Formation, have a relatively high density of dissolution features, including large cavities (greater than 4 inches), and dissolution-enlarged, mesoscopic fractures oriented in directions both perpendicular and parallel to bedding. It is not known whether interconnected networks of open systematic fractures are common or rare in deep bedrock conditions of southeastern Minnesota, in part because subsurface information is almost entirely limited to vertical boreholes that have a small probability of intersecting such features. Because systematic fractures are probably the result of regionalscale stresses (Price and Cosgrove, 1990), their presence in outcrop indicates that they are likely present in individual layers of strata even at depths hundreds of feet below the bedrock surface. They are most likely to occur locally in well-indurated layers such as those dominated by carbonate rock and cemented siliciclastics (Price and Cosgrove, 1990; Helgeson and Aydin, 1991; Hurley and Swager, 1991; Narr and Suppe, 1991). They are theoretically less likely to occur in friable sandstones and poorly indurated shales, although exceptions are well-documented (for example Ryder, 1996). Borehole video logs of a few wells open to deep bedrock conditions in the Twin Cities Metropolitan area reveal the presence of vertical systematic fractures with apertures of several inches in coarse clastic strata of the Jordan Sandstone and the fine clastic strata of the Eau Claire Formation (Minnesota Department of Health borehole video library; for example unique well numbers 200519, 206169, and 205821). These water wells were developed to increase productivity and it is possible that the apertures of these fractures were widened when the borehole was blasted with dynamite and bailed. Nevertheless, their presence demonstrates that Paleozoic strata do contain systematic planes of weakness in deep bedrock settings. The abundance, dimension, aperture size, and interconnectivity of these fractures are entirely unknown, but are presumably diminished compared to systematic fractures in shallow bedrock settings.

A.

~100 feet

EXPLANATION Non-systematic fractures (some dissolution enlarged)

Coarse clastic component Fine clastic component

Surficial deposits

Carbonate component

B.

Figure 7. Typical development of stress-relief fractures in layered Paleozoic bedrock. Note that nonsystematic stress-relief fractures decrease in abundance at greater distances from the bedrock surface. A. Diagrammatic sketch based on studies of Paleozoic bedrock in eastern North America (Ferguson, 1967), modified with observations from southeastern Minnesota discussed in this report. B. Quarry exposing carbonate rock of the Shakopee Formation and Oneota Dolomite near Mankato in Blue Earth County. Nonsystematic fractures are abundant in the upper part of the bedrock exposed in the quarry. Only widely spaced, systematic fractures are evident in the lower part of the quarry. The depth to which nonsystematic and systematic fractures extend continuously beneath the bedrock surface will vary from place to place in southeastern Minnesota.

17

Figure 8. Characteristic secondary pores in shallow bedrock conditions. A. Systematic fracture in interbedded fine clastic and carbonate rock component of the St. Lawrence Formation at Barn Bluff in Red Wing, Goodhue County. Note the vertical systematic fracture with large aperture (hammer for scale is circled) and flat surfaces of the outcrop characteristic of systematic fractures. B. Systematic fractures in the coarse clastic component of the Jordan Sandstone near Whitewater State Park, northeastern Winona County. C. Carbonate rock of the Oneota Dolomite in Stillwater, Washington County. Nonsystematic fractures occur parallel to bedding and as irregular, subvertical fractures typically confined to individual beds. Systematic fractures are relatively straight, and have wide apertures that cut vertically across the entire outcrop. Many of the fractures have some evidence of enlargement by dissolution. Staff is 5 feet tall. D. Nonsystematic, stress-relief fractures in interbedded fine clastic and carbonate rock of the St. Lawrence Formation at Barn Bluff in Red Wing, Goodhue County (hammer for scale). E. Similar fractures in fine clastic rock (chiefly shale) of the Decorah Shale at Lilydale Regional Park in Ramsey County. F. Large dissolution cavities (marked by arrows) developed in carbonate rock in the upper part of the Oneota Dolomite in eastern Wabasha County. These cavities typically are preferentially developed along discrete beds. The large cavity in the center of the photograph is about 2 feet in height. G. Small dissolution cavities in a carbonate bed within the upper Oneota Dolomite near Mankato, Blue Earth County.

A.

B.

C.

18

D.

E.

F.

G.

19

20

A.

C.

Figure 9. Variable development of secondary pores in carbonate rock from deep bedrock conditions. A. Interval with relatively few, small cavities in the Oneota Dolomite from Ramsey County (ATES Project, core AC-1). B. Larger, more abundant cavities in the Shakopee Formation from the same site (core BC-1). C. Irregular cavities and dissolution-enlarged fractures in the Shakopee Formation from core H-1 in Freeborn County. D. and E. Cavities in the St. Lawrence Formation from the Waseca–Waterville area (core Prehn 3).

B.

D.

E.

The greater development of secondary porosity in shallow bedrock conditions compared to deep conditions is the result of several processes. Uplift, unloading of younger bedrock, and weathering in shallow conditions opens the apertures of systematic planes of weakness in addition to producing the ubiquitous nonsystematic bedding-plane and curvilinear fractures (Ferguson, 1967; Wyrick and Borchers, 1981; Price and Cosgrove, 1990) characteristic of all bedrock outcrops in southeastern Minnesota. These latter features are commonly referred to as "stress-relief fractures" in reference to their common origin during the removal of overlying material. Vertical and horizontal stresses that accompany glacial advances and retreats across Paleozoic bedrock in southeastern Minnesota can contribute to the production of these features (for example Moerner, 1978; Liszkowski, 1993). In addition, dissolution of carbonate rock is typically more pronounced in relatively near-surface settings compared to conditions of relatively deep burial. The depth to which these processes collectively produce hydrogeologically significant secondary porosity will vary from place to place depending on several factors. For example, a system of interconnected systematic fractures in layered bedrock typically terminates downward at or near the uppermost friable sandstone, relatively ductile shale, or along a discontinuity such as a bedding-plane fracture (Price and Cosgrove, 1990; Helgeson and Aydin, 1991; Narr and Suppe, 1991). Dissolution of carbonate rock typically diminishes with depth (Goldstrand and Shevenell, 1997; Shevenell and Goldstrand, 1997), especially below the uppermost impermeable layer of siliciclastic bedrock that can serve as a confining bed. There is no precise or consistent depth at which the boundary between "shallow" and "deep" bedrock conditions occurs. A study of the Prairie du Chien and Jordan aquifers by Runkel and others (1999) placed the lower boundary of shallow bedrock conditions at 100 feet below the bedrock surface everywhere in the Minneapolis–St. Paul metropolitan area because examination of borehole videos, core, and outcrops indicated that open fractures and dissolution features are relatively uncommon below that depth. Runkel (1999, 2000) conducted a larger-scale investigation, which included most of the Paleozoic stratigraphy across a nine-county area of southeastern Minnesota, and proposed a 200-foot-deep boundary between deep and shallow bedrock conditions as a regional-scale generalization (Plates 1, 2). Such an interpretation is consistent with studies outside of Minnesota that similarly depict a relatively well-connected, high-density system of secondary pores in the uppermost 100 to 200 feet of layered sedimentary bedrock (Ferguson, 1967;

21

Wyrick and Borchers, 1981; Williams and others, 1984; Graese and others, 1988; Hatcher and others, 1992; Soloman and others, 1992; Sasowsky and White, 1994; Michalski and Britton, 1997; Morin and others, 1997). A synthesis of engineering data from five dam sites in North America by Snow (1968) suggested that fracture porosity decreases an order of magnitude from the land surface to a depth of 200 feet regardless of the dominant lithology at the individual sites. Investigations of the Cambrian Maynardville Limestone in Tennessee demonstrated that secondary pores were more prevalent in the uppermost approximately 110 feet of bedrock and extremely rare below a depth of about 240 feet, except in specific facies with a high susceptibility to dissolution, such as evaporites (Goldstrand and Shevenell, 1997; Shevenell and Goldstrand, 1997). The term "shallow bedrock conditions" in this report refers to the upper 200 feet of Paleozoic bedrock regardless of the thickness and composition of overlying unconsolidated materials. This 200-foot boundary is chosen with the understanding that the change from what we have characterized as shallow bedrock conditions to deep bedrock conditions is in reality transitional, and will vary in depth from place to place.

Hydraulic character The geologic controls on hydraulic conductivity of Paleozoic bedrock is evaluated in this report by comparing values of conductivity calculated for individual wells and scientific boreholes to the hydrostratigraphic setting of their open-hole interval (Fig. 10, for example). Additionally, for selected parts of southeastern Minnesota, plots showing the spatial distribution of hydraulic conductivity for individual hydrostratigraphic units were compared to maps of bedrock topography, bedrock geology, structure, and isopachs to assess the effects of features such as bedrock valleys and the influence of faults and folds to aquifer performance. Our evaluation indicates that calculated hydraulic conductivity at an individual borehole largely reflects the hydrostratigraphic character of its open-hole interval. For example, borehole measurements of hydraulic conductivity within individual lithostratigraphic units in deep bedrock settings largely reflect the permeability and thickness of its matrix hydrostratigraphic component(s), and the degree of development of secondary pores (Fig. 10, for example). Rocks dominated by a fine clastic or carbonate rock component with few secondary pores have relatively low hydraulic conductivity and serve as confining units in deep bedrock settings (for example Libra and Hallberg, 1985; Nicholas and others, 1987; Graese and others, 1988; Miller and Delin, 1993). Horizontal hydraulic

conductivity of the fine clastic component based on discrete interval packer tests commonly ranges from as low as 10 -7 foot per day for units composed almost entirely of shale (Freeze and Cherry, 1979; Graese and others, 1988; Eaton and others, 2000), to 10-2 to 10-1 foot per day for interbedded, very fine-grained sandstone and shale (for example Miller and Delin, 1993). Vertical conductivity in the fine clastic component is estimated to be about two orders of magnitude less than horizontal (MUGSP, 1980; Miller, 1984; Kanivetsky, 1989; Setterholm and others, 1991; Miller and Delin, 1993). Packer tests of discrete intervals of unfractured, dense Paleozoic carbonate rock in Wisconsin, Illinois, and Iowa have typically indicated horizontal conductivities of about 10-4 foot per day or less (Graese and others, 1988; Gianniny and others, 1996). In contrast, the coarse clastic component, and intervals of carbonate rock containing abundant dissolution features, have relatively high hydraulic conductivity values in deep bedrock settings. Discrete interval packer tests, specific capacity data, and standard aquifer tests of wells open only to the coarse clastic component typically range from a few feet per day to as much as 60 feet per day (for example Nicholas and others, 1987; Young, 1992; Miller and Delin, 1993; Runkel and others, 1999). The carbonate rock component can have conductivity values commonly as high as hundreds of feet per day, and dye traces through locally deep bedrock settings demonstrate flow speeds as rapid as miles per day along discrete intervals where well-developed conduit systems are present (Libra and Hallberg, 1985; Wheeler, 1993; Alexander and Lively, 1995; Paillet and others, 2000; Tipping and Runkel, 2001). The inverse relationship between the degree in the development of secondary porosity and depth of burial beneath bedrock (Plates 1, 2) is reflected by hydraulic performance. The enhanced development of secondary pores in shallow bedrock conditions corresponds to a measurable increase in hydraulic conductivity for the Paleozoic bedrock of southeastern Minnesota. The scatter and box plots (Fig. 11) of 8,626 conductivity values calculated from specific capacity tests compared against the depth of the open-hole interval below the bedrock surface show increased conductivity corresponding to decreased burial beneath younger bedrock. Similar comparisons of hydraulic properties to depth of burial beneath the bedrock surface are made for each individual Paleozoic lithostratigraphic unit in subsequent sections of this report, and they demonstrate that individual matrix hydrostratigraphic units in southeastern Minnesota, including those dominated by coarse clastic and fine clastic strata, have a much higher average conductivity, and a greater range in conductivity where they occur in shallow bedrock conditions compared to where the same units occur in deep bedrock

22

conditions. Additionally, ongoing borehole flowmeter investigations (for example Paillet and others, 2000; Tipping and Runkel, 2001) indicate that ambient flow rates in boreholes exposed only to deep bedrock conditions are typically subdued compared to the much higher and more variable flow rates in boreholes open to shallow bedrock conditions. This relationship likely reflects the higher permeability due to enhanced secondary porosity, and the greater stresses of nearsurface recharge and discharge in shallow bedrock conditions. A number of studies of sedimentary bedrock in North America likewise show a positive correlation between hydraulic performance and proximity to the bedrock surface. Bedding-plane fractures that provide preferential flow paths in siliciclastic strata have been demonstrated to decrease in frequency and in hydraulic conductivity with depth (Michalski and Britton, 1997; Morin and others, 1997). Packer tests of Ordovician and Silurian carbonate rock and shale in northern Illinois demonstrated that the highest conductivities were consistently within the uppermost 100 feet of the bedrock surface (Kempton and others, 1987). The same study indicated that bedrock strata within the uppermost 40 feet of the bedrock surface are on average 100 times more permeable than the rocks below (Kempton and others, 1987; Curry and others, 1988). A regionally extensive "confining unit" composed of Paleozoic shale and limestone in Illinois and adjacent states has a hydraulic conductivity of 10-7 to 10-5 foot per day where deeply buried by younger bedrock, and 10-3 to 12 feet per day closer to the bedrock surface (Eberts and George, 2000). A study of the karstic Maynardville Limestone in Tennessee demonstrated that "quick flow" conduits of the highest conductivity were rare below 110 feet and absent 240 feet below the bedrock surface, where only slow flow through matrix pores was recorded (Goldstrand and Shevenell, 1997; Shevenell and Goldstrand, 1997). The Galena Group in northeastern Iowa has similar attributes: where relatively deeply buried by younger bedrock that includes a shaly confining unit, the Galena Group is characterized by diffuse flow along relatively unmodified, narrow fractures. In conditions of shallow burial, particularly where it occurs as the uppermost bedrock, it is characterized by cavernous pores and more rapid conduit flow (Rowden and Libra, 1990; Keeler, 1997). The greater abundance, interconnectivity, and aperture of fractures, and increased susceptibility of dissolution accounts for the relatively high magnitude and variability in conductivity in shallow bedrock conditions at each of the sites described above. The results of our study revealed our limited understanding of the relative hydraulic importance of

100%

100

10-6

100

0

5

0

Packer intervals

(feet per day)

Packer test hydraulic conductivity

Fractures per foot

Plug permeability horizontal (md)

Plug permeability vertical (in md) 40% 10-6

0

8

450

500

550

Shaly sandstone

650

Eau Claire Formation

Ironton–Galesville Sandstone

Franconia Formation

Mazomanie Member

St. Lawrence Formation

0

Plug porosity

Visual porosity

Depth in feet below bedrock surface

Matrix hydrostratigraphic component

Gamma log Increasing counts

700

Coarse clastic component

Carbonate component

Fine clastic component

Plug sample

Figure 10. Hydrostratigraphic and hydraulic attributes for the Eau Claire through St. Lawrence Formations in a deep bedrock setting at the ATES Project site in Ramsey County. Plug tests of porosity and permeability characterize the small-scale matrix hydrostratigraphic attributes. Visual examination of core provides information on the distribution and character of macroscopic secondary pores. Discrete interval packer tests measure the hydraulic performance of the hydrostratigraphic components. Note that plug-scale permeability values positively correlate with hydraulic conductivity where secondary pores are absent or rare. Fine clastic and carbonate rock components that test at relatively low permeability have a correspondingly low conductivity except in the lower part of the St. Lawrence Formation where dissolution cavities are present. Based on data in Walton and others (1991), Miller and Delin (1993), and core logging as part of this investigation.

23

secondary pores at scales larger than that measured by an individual borehole pump test, especially in deep bedrock conditions of southeastern Minnesota. A number of studies summarized in this report have demonstrated that stratigraphically controlled networks of dissolution cavities can be important hydraulic conduits along discrete intervals of carbonate rock in deep bedrock settings. The abundance and hydraulic importance in deep bedrock settings of secondary pores such as networks of systematic and bedding-plane fractures, and other enhanced permeability features in clastic rocks, is less defined. Borehole flowmeter tests of clastic strata in deep bedrock conditions (described later in this report) document the dominance of intergranular flow in some individual boreholes and, conversely, of flow along bedding-plane fractures in other boreholes. Our limited subsurface data demonstrate that hydraulically active secondary pores therefore clearly exist in deep bedrock conditions, and such features may be significant in the transport of water through the aquifer at larger horizontal scales. Indirect evidence such as water chemistry and potentiometric levels indicates that any such fracture networks are poorly connected vertically across units dominated by fine clastic and carbonate strata. Those low-permeability components have been demonstrated to provide vertical hydraulic separation in deep bedrock settings. Furthermore, discrete interval packer tests of individual boreholes (Fig. 10, for example) open to siliciclastic strata in deep bedrock conditions generally show a positive correlation between hydraulic conductivity and intergranular porosity and permeability. The results of less rigorous, "standard" aquifer tests and of thousands of specific capacity tests in southeastern Minnesota largely show the same positive correlation between conductivity and intergranular porosity and permeability. Such results are compatible with either single or multiple porosity interpretations. On the other hand, the pump test results from a small percentage of boreholes open to deep bedrock conditions do not reflect intergranular permeability alone, which suggests that a larger-scale conductivity accommodated by secondary pores exists. For example, the specific capacity database for some units dominated by the coarse clastic component such as the Mt. Simon, Ironton–Galesville, and Jordan Sandstones includes a small percentage of statistically outlying conductivity values that are higher than expected if intergranular permeability was the only control on well yield. Aquifer tests of large diameter wells, particularly those that have been developed by blasting, include the greatest percentage of relatively high conductivity values. These values of hydraulic conductivity possibly correspond to the small percentage of wells in deep bedrock

24

conditions that intersect networks of hydraulically significant secondary pores such as bedding-plane and systematic fractures. The majority of wells do not intersect such fracture networks, and pump tests of those wells do not clearly reflect the enhanced permeability provided by this larger-scale pore system. These observations are not compatible with simple, single porosity, intergranular flow interpretations but are compatible with multi-porosity interpretations.

Hydrogeologic framework Our hydrogeologic framework (Plates 1, 2) is based on hydraulic data interpreted within the context of the hydrostratigraphic attributes of the Paleozoic stratigraphic section, described in detail in the subsequent sections of this report. The framework we present is more complex than those depicted in previous publications because many of the individual lithostratigraphic units that were formerly considered single hydrogeologic units across their entire extent are in this report subdivided at a regional scale into two or more aquifers and confining units. In deep bedrock conditions, the Paleozoic strata of southeastern Minnesota include at least eleven aquifers, chiefly hydrostratigraphic units dominated by coarse clastic rock or intervals of carbonate rock with relatively abundant secondary pores. These aquifers are separated by ten regional confining units composed of fine clastic strata or carbonate rock with few interconnected secondary pores. In shallow bedrock conditions, each of the individual hydrostratigraphically defined units, including those that provide confinement in deep bedrock settings, are of much greater bulk conductivity, and could be considered aquifers because each has been demonstrated to yield economic quantities of water. The ability of the confining units to provide hydraulic separation in such conditions is diminished and greatly variable on a local scale because of the relatively enhanced development of secondary porosity. Our hydrogeologic framework also delineates three major "karst systems" (Plates 1, 2), based largely on the work of Alexander and Lively (1995), Alexander and others (1996), and Green and others (1997). A karst system is an integrated mass-transfer system in soluble rocks with a permeability structure dominated by conduits dissolved from the rock and organized to facilitate the circulation of fluid (Klimchouk and Ford, 2000). Southeastern Minnesota karst systems composed of carbonate-dominated strata where they lie in shallow bedrock conditions in ascending stratigraphic order include the Prairie du Chien, Galena–Spillville, and Cedar Valley karst systems. Each karst system is characterized by relatively abundant secondary pores including large cavities, and dissolution enlarged

B. 400

1200

350

Conductivity in feet per day

1400

1000 800 600 400 200

300 250 200 150 100 50

0

0

0

100

200

300

400

500

600

700

800

0

Distance in feet between the bedrock surface and the open-hole top

C.

100

200

300

400

500

600

700

800

Distance in feet between the bedrock surface and the open-hole top

D.

DEEP BEDROCK CONDITIONS

SHALLOW BEDROCK CONDITIONS 100

80

80

60

60

Range

100

Range

Conductivity in feet per day

A.

40

40

20

20

0

0 1,434 samples

7,192 samples

Figure 11. Hydraulic conductivity data for 8,626 wells open only to Paleozoic bedrock in southeastern Minnesota. Calculated from specific capacity tests in the County Well Index database. Box plots (C and D) provide an easily observable manner to view the distribution of hydraulic conductivity values. Each box in this and subsequent conductivity figures encloses 50 percent of the values with the median value of the variable displayed as a line. The top and bottom of the box mark the limits of 25 percent of the variable population. The lines extending from the top and bottom of each box mark the minimum and maximum values that fall within an acceptable range. Any value outside of this range, called an outlier, is displayed as an individual point. Note, outliers are used in the calculations for the box plot, as well as in calculations of mean values in subsequent figures. A. and B. Scatter plots showing the relationship between the depth of the open-hole interval below the bedrock surface and hydraulic conductivity. Shallower wells tend to have higher conductivity. The two plots show the same set of data at different scales. C. Box plot of hydraulic conductivity values for deep bedrock conditions. Plot does not show 43 outlying values greater than 100 feet per day. D. Box plot of hydraulic conductivity values for shallow bedrock conditions. Plot does not show 682 outlying values greater than 100 feet per day.

25

systematic and nonsystematic fractures, and rapid, direct connections between the surface and ground water (Fig. 12). These features may be expressed at the land surface by caves, numerous springs, and many sinkholes in areas with only a thin cover of unconsolidated material, but the same aquifer properties often exist in areas with few if any obvious surface karst features. In the subsurface the hydraulic properties of karst systems are very heterogeneous, with large conduits that allow water to travel as rapidly as miles per day draining matrix blocks of very low conductivity that store water moving much slower. The karst aquifers, also called "triple porosity aquifers" (Worthington, 1999), are important to ground-water management because the ground-water movement through conduits can be extremely rapid and difficult to predict. Catastrophic introduction and rapid travel of contaminants are well-documented in southeastern Minnesota (Alexander and Book, 1984; Alexander and others, 1993; Wheeler, 1993).

HYDROGEOLOGIC ATTRIBUTES OF INDIVIDUAL LITHOSTRATIGRAPHIC UNITS The hydrostratigraphic and hydraulic attributes of each of the major lithostratigraphic units of southeastern Minnesota are described in detail in the remainder of this report. We present our results organized in a manner in accordance with the standard lithostratigraphic nomenclature of southeastern Minnesota because these lithostratigraphic units are entrenched in our literature, databases, maps, and vocabulary. In addition, only lithostratigraphic units are delineated for individual water-well sites in the County Well Index database. Presented in this manner, our characterization can be used in combination with the County Well Index database and bedrock geologic maps as a guide to predict internal hydrostratigraphic and hydraulic variability within individual lithostratigraphic units at regional as well as site-specific scales. The characterization of the hydrogeologic attributes of individual lithostratigraphic units begins with a description of the character and distribution of the three principal matrix hydrostratigraphic components and a discussion of the development of secondary porosity in deep and shallow bedrock conditions, features that are shown on Plates 1 and 2. This is followed by a synopsis of hydraulic properties, compiled on Figures 13 and 14. Lastly, a discussion combines the hydrostratigraphic and hydraulic information in an integrated synthesis of the hydrogeologic character of each lithostratigraphic unit, including delineation of individual aquifers and confining units.

26

Hydraulic conductivity and flow speeds measured for individual lithostratigraphic units are divided into subsets of values where measurable differences in hydraulic conductivity and productivity correspond to internal variations in hydrostratigraphic character. For example, hydraulic properties of each lithostratigraphic unit are described separately for deep and shallow bedrock conditions. Conductivity values for shallow bedrock conditions are based on pump and flowmeter tests of wells that are open at least in part to within 200 feet of the bedrock surface, and attributes of deep bedrock conditions are based on tests for wells where the uppermost 200 feet of bedrock is cased off. The degree to which individual units are characterized varies greatly. The Prairie du Chien Group, Jordan Sandstone, and Franconia Formation have been the subject of a large number of detailed investigations in a wide range of geologic settings in southeastern Minnesota, and as a result we summarize significant recent advances in our understanding of the hydrogeologic attributes of those units. Other intervals, such as the Mt. Simon Sandstone and Galena Group, are known to vary internally in hydraulic properties and therefore to consist of multiple hydrostratigraphic units, but they have not been subjected to rigorous hydrostratigraphic and hydraulic study in Minnesota. These units are characterized in a more cursory fashion, and we draw heavily upon investigations of these strata conducted outside of Minnesota.

MT. SIMON SANDSTONE Hydrostratigraphic attributes Matrix porosity The Mt. Simon Sandstone is broadly divisible into two parts based on intergranular attributes (Figs. 15, 16). The lower Mt. Simon Sandstone across most of southeastern Minnesota consists chiefly of a coarse clastic component that is moderately to poorly cemented. Plug test permeability is commonly greater than 1,000 md in both vertical and horizontal directions (MUGSP, 1980). Fine clastic interbeds are a subordinate component in the lower Mt. Simon Sandstone; they are most abundant in Fillmore, Houston, and Winona Counties in southeastern Minnesota The upper part of the Mt. Simon Sandstone across most of southeastern Minnesota is hydrostratigraphically more complex (Figs. 15, 16). It consists of approximately equal parts coarse clastic and fine clastic strata intercalated in beds from a few feet to as much as 30 feet thick. Plug tests from individual boreholes demonstrate that vertical permeability ranges over ten orders of magnitude: coarse clastic intervals are

commonly over 1,000 md, whereas fine clastic intervals commonly have values from 10 -6 to 10 -3 md. Some individual plug samples are strongly anisotropic, with a horizontal permeability 1,000 times greater than vertical.

Secondary porosity Deep bedrock conditions—Our knowledge of secondary porosity in the Mt. Simon Sandstone in deep bedrock settings is extremely limited. Conventional wisdom maintains that as a friable, mostly high porosity unit covered by layers of younger bedrock of contrasting material properties, interconnected networks of open

systematic fractures may not be developed to an appreciable degree (Price and Cosgrove, 1990; Helgeson and Aydin, 1991; Narr and Suppe, 1991). An unproved assumption by previous investigators is that interconnected networks of fractures and dissolution features are rare to absent in the Mt. Simon Sandstone in deep bedrock settings and therefore porosity and permeability are determined chiefly by intergranular attributes. Shallow bedrock conditions—The Mt. Simon Sandstone occurs in a relatively shallow bedrock setting along the Mississippi River and the lower reaches of its tributaries in southeastern Minnesota (Mossler and

Stream sinks Blind valley

Sinkholes

Seepage

Surface stream

Fractures Caves Sump Spring

EXPLANATION Coarse clastic component

Non-systematic fractures (some dissolution enlarged)

Fine clastic component

Systematic fractures (some dissolution enlarged)

Carbonate component

Dissolution features—cavities and enlarged bedding-plane fractures

Surficial deposits

Water

Figure 12. Typical attributes of a karst system in southeastern Minnesota. Such systems are developed most commonly in carbonate rock in shallow bedrock conditions. Any fracture or cavity may contain ground water in this karst system.

27

Book, 1984; Mossler, 1990, 2001; Mossler and Bloomgren, 1990; Runkel, 1996a, 1998). However, it is deeply buried by unconsolidated Quaternary sediment in these areas and has not been studied using subsurface techniques that provide information on secondary pores. The Mt. Simon Sandstone has open, systematic and nonsystematic fractures where it is well exposed in large outcrops in west-central Wisconsin, such as in the cities of Eau Claire and Chippewa Falls. Although Minnesota lacks large outcrops that can be confidently assigned to the Mt. Simon Sandstone, the characteristics of the formation in Wisconsin suggest that open fractures are present in the Mt. Simon Sandstone where it occurs in shallow bedrock conditions in southeastern Minnesota.

Hydraulic attributes Deep bedrock conditions—A large number of discrete interval and standard aquifer tests of the Mt. Simon Sandstone where it occurs in deep bedrock conditions (Fig. 13) in Illinois, southern Wisconsin, and southeastern Minnesota have a range in hydraulic conductivity from 0.38 to 21 feet per day (Nicholas and others, 1987; Young, 1992; Carlson and Taylor, 1999). The lowest values are calculated from tests of wells where the Mt. Simon Sandstone is buried by several hundred to thousands of feet of younger bedrock. Studies in Illinois have demonstrated that in such settings the intergranular permeability in the sandstone is reduced compared to shallower conditions of burial as a result of the enhanced development of pore-filling cement and compaction (Hoholick and others, 1984). The highest values of hydraulic conductivity, those greater than 12 feet per day, are calculated from tests of wells within a few miles of the Mississippi River, where the Mt. Simon Sandstone occurs much closer to the bedrock surface. The hydrostratigraphically complex upper Mt. Simon Sandstone has not been individually tested to determine its conductivity in Minnesota. Nicholas and others (1987) calculated a bulk horizontal hydraulic conductivity of 1.3 feet per day for the upper part of the Mt. Simon Sandstone in Illinois, where it contains alternating beds of coarse clastic and fine clastic strata (Fig. 15), similar to Minnesota. This bulk conductivity value most likely is a measure of the individual, high permeability coarse clastic beds in the upper part of the formation. In contrast, individual beds of the fine clastic component in the upper Mt. Simon Sandstone will have a horizontal hydraulic conductivity of about 10 -3 to 10 -1 foot per day, and a vertical hydraulic conductivity between 10 -5 and 10 -3 foot per day based on tests of strata with similar matrix porosity and permeability in other Paleozoic formations in

28

southeastern Minnesota (for example Miller and Delin, 1993). Hydraulic conductivity values based on specific capacity tests for 25 wells open to the Mt. Simon Sandstone in deep conditions of burial are depicted in Figure 17. The values typically range from less than one to as much as 50 feet per day, and average conductivity is 39.5 feet per day. Exclusion of two outlying values greater than 240 feet per day from the database results in a calculated average conductivity of 21 feet per day, a value more consistent with the higher-quality pump tests described above, and with conductivity values calculated for other deeply buried Paleozoic aquifers dominated by intergranular flow through the coarse clastic component in southeastern Minnesota (described later in this report). The two outlying values of greater than 240 feet per day were calculated from tests of wells that were developed by blasting or decompression techniques, and if accurate they may reflect a significant contribution of yield from a network of hydraulic fractures. Shallow bedrock conditions—Hydraulic conductivity values calculated from specific capacity tests for 165 wells open to the upper and lower parts of the Mt. Simon Sandstone under shallow bedrock conditions have a range from less than 1 to 70 feet per day. Average hydraulic conductivity is 29.3 feet per day (Fig. 17). The greater range in conductivity compared to deep bedrock conditions probably reflects a higher percentage of wells in which significant contribution occurs through networks of fractures.

Hydrogeologic synthesis The Mt. Simon Sandstone is hydro-stratigraphically complex because it contains two intercalated components, the coarse clastic and fine clastic components, that differ markedly from one another in permeability (Figs. 15, 16). Similar hydrostratigraphic properties have been described for the Mt. Simon Sandstone in Illinois and Wisconsin, where hydrogeologic investigations have demonstrated that the intercalations of fine clastic material serve as a confining unit(s) in deep bedrock settings (Nicholas and others, 1987; Carlson and Taylor, 1999). Packer tests in the Illinois study demonstrated that even though the upper Mt. Simon Sandstone contains coarse clastic intervals that provide moderately high bulk horizontal conductivity, fine clastic interbeds have low enough vertical conductivity to function as a confining unit(s) that separates an upper from a lower Mt. Simon aquifer (Fig. 15). Potentiometric heads in the two aquifers differed from one another by over 50 feet, and water in the lower aquifer is different from the upper in

fundamental water chemistry, particularly in its order of magnitude higher concentration of chloride and sodium (Nicholas and others, 1987). Preliminary results of an ongoing investigation of the Mt. Simon Sandstone in southeastern Wisconsin using borehole, flowmeter, temperature, and water chemistry data have similarly indicated the presence of a "middle" confining unit that divides the formation into two distinct aquifers (K. Bradbury, unpub. data, 2001). Unpublished investigations of the Mt. Simon Sandstone in the three-county Waseca–Waterville area of southeastern Minnesota (Fig. 1) indicated that fine clastic interbeds in the middle to upper Mt. Simon Sandstone may serve as confining units (MUGSP, 1980), as they do in Illinois and Wisconsin (Fig. 15). There are at least two fine clastic intervals of Mt. Simon Sandstone in that area that are of sufficient thickness and low permeability to function as hydraulic caps that could potentially confine natural gas stored in the coarse clastic material beneath them. These intervals, designated Cap Rock A and Cap Rock B as part of the MUGSP, suggest that the Mt. Simon Sandstone in the Waseca–Waterville area may be divisible into a lower, middle, and upper aquifer. The lower and middle aquifers are dominated by high-conductivity coarse clastic strata, whereas the upper Mt. Simon aquifer consists of alternating thin to medium beds of coarse and fine clastic material. Ground-water chemistry from two boreholes indicated that water from the lower Mt. Simon aquifer has measurably greater dissolved solids, including chloride concentration that is as much as 6 times higher than that of water in the middle and upper Mt. Simon aquifers (MUGSP, 1980). The common practice of depicting the entire Mt. Simon Sandstone as a single aquifer in Minnesota should be reevaluated because the formation may be internally compartmentalized into two or more hydrogeologic units (Fig. 15, for example). Although coarse clastic beds across its entire stratigraphic extent can yield economic quantities of water, the fine clastic beds, which are most common in its upper one-half, are commonly as thick and of similar matrix hydrostratigraphic properties as younger Paleozoic layers demonstrated to act as confining beds in southeastern Minnesota, such as those intercalated with coarse clastic layers in the lower St. Peter Sandstone in the Twin Cities Metropolitan area (Kanivetsky and Cleland, 1992). Studies of ground-water chemistry in the Mt. Simon Sandstone should in particular be conducted with consideration of multiple aquifers, particularly because water contributed from the Precambrian "basement" may be confined largely to the lowermost part of the formation.

29

EAU CLAIRE FORMATION Hydrostratigraphic attributes Matrix porosity The Eau Claire Formation is composed chiefly of the fine clastic component, with vertical permeabilities that typically range from 10-5 to 10-3 md (Figs. 16, 18). It varies little in its matrix character across most of southeastern Minnesota, containing only relatively thin (typically less than 10 feet) coarse clastic interbeds in its upper part in Wabasha, Winona, and Houston Counties (Runkel, 1996a). One exception is in Faribault and adjacent counties where the Eau Claire Formation reaches its greatest thickness, and includes a hydrostratigraphically distinct "greensand" facies dominated by fine- to medium-grained glauconitic sandstone as thick as 100 feet (Fig. 18; Mossler, 1992). This coarse clastic component has plug-scale permeabilities that are orders of magnitude higher than the bulk of the Eau Claire Formation elsewhere in southeastern Minnesota.

Secondary porosity Deep bedrock conditions—Our knowledge of secondary porosity in deep bedrock conditions is limited. Intergranular permeability is widely believed to be the chief control on hydraulic behavior in deep bedrock settings. Logged cores and a few borehole videos corroborate that secondary pores are rare in the Eau Claire Formation (Figs. 16, 18), occurring mostly as small dissolution cavities in dolostone beds that compose a relatively minor part of the Eau Claire Formation across southeastern Minnesota. A borehole video of Brooklyn Park municipal well 4 (Minnesota Department of Health borehole video library, unique number 203265) revealed the presence of a systematic fracture with an inch-scale aperture over 300 feet below the bedrock surface in the Eau Claire Formation. The well had been blasted and bailed to increase productivity—procedures that may have increased the aperture of the fracture. Shallow bedrock conditions—In shallow bedrock conditions, the Eau Claire Formation has features typical of fine clastic rock that has been subjected to stress relief and weathering. Nonsystematic fractures are common in outcrops of the Eau Claire Formation in Minnesota and adjacent parts of Wisconsin, including irregular, sub-vertical fractures a few inches in width and bedding-plane fractures that can be traced tens of feet laterally. Large quarries of Eau Claire Formation rock in west-central Wisconsin commonly have systematic fractures with apertures as wide as several

30

Ostp

Ogwd

Kv 10-3 ft/day (12)

Kh 1.8 ft/day (7)

Kv 10-6 ft/day (9) Kh 8.4 ft/day (4) Kh 15.9 ft/day (2) Kh 1.3 to 3.7 ft/day (13)

Kh <10-2 ft/day (7) Kh 10-3 to 10-2 ft/day with thin intervals to 10-1 ft/day (4)

Kh mostly 10-3 to 10-2 ft/day except discrete thin intervals 1.4 to 14 ft/day (4)

Kh <10-2 ft/day (7)

Opvl

Kh 6.5 ft/day (2)

Kh >174 ft/day (1)

Kh >190 ft/day (1)

Kh <10-6 ft/day (9)

Maximum of 2 to 7 ft/day some substantially lower (pumped dry) (6)

Kh <10-4 to 10-3 ft/day (4)

Kh 5.3 ft/day (1) Kh 0.5 ft/day (1) Kh 39 ft/day (3)

Kh 0.8 ft/day (1)

Kh 25 ft/day (1)

Kh 26 ft/day (1)

Odcr

Ogal

Odub

Omaq

Dspl

Dclp

Dclc

Dcum

Dcuu

DEEP BEDROCK CONDTIONS

Vertical flow speed ~260 ft in less than 7 months (6)

Horizontal flow speeds 0.23 miles/yr to 1.8 miles/day (6)

Kh 67 ft/day (2)

Kh 64.6 ft/day (2)

Kh 38.7 ft/day (2) Kh 20 to 30 ft/day (14) Kh 24 to 30 ft/day (15) Kh ~20 ft/day (16)

Kh <10-1 to > hundreds of ft/day (10) Travel times of up to 1 mile/day (11)

Kh 60.1 ft/day (2)

Kh 3 to 11 ft/day (6)

Kh 8.9 x 10-4 to 3 x 10-2 ft/day (6)

Kh 10-3 to 300 ft/day (8)

Kh 72 ft/day (2)

Kh 5.3 ft/day (1)

Kh 92 ft/day (1)

Kh maximum of 1 to 2 ft/day (6)

Kh 13.4 ft/day (1)

Pumped at 17 gpm with no drawdown (1)

Kh 10-3 to 28 ft/day (4, 5)

Pumped at 30 gpm with no drawdown (1)

SHALLOW BEDROCK CONDITIONS

Ostp

Ogwd

Opvl

Odcr

Ogal

Odub

Omaq

Dspl

Dclp

Dclc

Dcum

Dcuu

31


mts


ecr


igl


frn


stl


jdn

Opod

Opsh

Kh 1.5 ft/day (7)

Coarse clastic Kh 1.3 ft/day (7) Fine clastic Kh 10-2 ft/day; kv 10-4 ft/day (25)

Kh 10-2 ft/day (27) Kv 10-4 ft/day (27) Kh 10-3 to 10-2 ft/day (35)

Figure 13

Kh 60.8 ft/day (2)

Kh 3 ft/day with discrete intervals avg. 220 ft/day (30)

Kh 31.7 ft/day where Mazomanie Member is thick (2)

Kh 163 ft/day (23) Kv 1.75 ft/day (23)

Kh 29.3 ft/day (2)

Kh 36.7 ft/day (2)

Kh 26.8 ft/day (2) Kh 10 to 100 ft/day (35) Kh avg. 20 ft/day (35)

Kh 10-1 ft/day (32)

Kh 46 ft/day (2)

Kh approx. 30 to >500 ft/day (22)

Kh 43.2 ft/day (2)

Kh 7.5 x 10-3 ft/day (23) Kv 1.5 x 10-4 ft/day (23) Kv 10-4 ft/day (24)

Kh 45 ft/day (22)

Kh 10-2 to 85 ft/day (31)

Kh 0.1 ft/day (21)

Kh 32.3 ft/day where Mazomanie Member is thin to absent (2)

Bulk Kh ranges from 5.3 to 18 ft/day (20)

Discrete <3 ft intervals Kh range from 2.2 to 1,023 ft/day (20)

Up to 800 ft/day travel (19)

10 ft intervals Kh range from 1.6 to 65 ft/day (19)

Explanation to Figure 13 is on pages 32 and 33.

Kh 39.5 ft/day (2) Kh 17 ft/day (36) Kh 1.5 to 5 ft/day (37) Kh 0.38 to 21 ft/day (38)

Kh 10.8 ft/day (2) Kh 3 ft/day (18) Kh 10 ft/day (18) Kh 11 ft/day (33) Kh 2.9 to 31 ft/day (34)

Kh 5 ft/day (27) Kv 0.5 ft/day (27)

Kh 1.6 to 7.9 ft/day (27) Kv 0.16 to 0.79 ft/day (27)

Kh 27.8 ft/day where Mazomanie Member is thick (2)

Kh 5.9 ft/day where Mazomanie Member is thin to absent (2)

Kh 9.3 and 20 ft/day (29)

Kh 8.0 to 24 ft/day (26)

Kh 0.1 to 100 ft/day (22)

Kh 17.4 ft/day (2)

(likely yielded by discrete intervals of high conductivity separated by intervals with low conductivity)

Kh 33.5 ft/day (2)

Kh ≤ 10-2 ft/day (27) Kv ≤ 10-4 ft/day (27)

Kh 1.4 to 7.5 ft/day (27) Kv 0.14 to 0.75 ft/day (27)

Kh 14.0 ft/day (2) Kh 10-2 to 6.7 ft/day (27) Kv 10-4 ft/day (28)

Kh 10-2 ft/day (25) Kv 10-4 ft/day (25)

Kh 837 ft/day (20a)

Flow speed 6.5 miles/day (17)

Kh 9 ft/day (18)


mts


ecr


igl


frn


stl


jdn

Opod

Opsh

EXPLANATION Coarse clastic component

Non-systematic fractures (some dissolution enlarged)

Fine clastic component

Systematic fractures (some dissolution enlarged)

Carbonate component

Dissolution features—cavities and enlarged bedding-plane fractures

Kh—Horizontal hydraulic conductivity Kv—Vertical hydraulic conductivity ft/day—Feet per day gpm—Gallons per minute

Figure 13. Figure appears on pages 30 and 31. Generalized stratigraphic column of Paleozoic strata in southeastern Minnesota showing matrix hydrostratigraphic components, typical development of secondary porosity, and hydraulic data compiled for this report. Figure is not to scale. Hydraulic data for these figures (numbers in parentheses) are from the following (see Fig. 1 for the location of listed counties): 1. Discrete interval packer testing in northern Iowa (Floyd and Mitchell Counties) by Libra and Hallberg (1985). 2. Average value of conductivity calculated based on specific capacity tests in the County Well Index database. 3. Standard aquifer pump test of the LeRoy municipal well (unique number 127280) by J. Green of the Minnesota Department of Natural Resources, Mower County. 4. Discrete interval packer tests in Illinois by Kempton and others (1987), Curry and others (1988), and Graese and others (1988). 5. Discrete interval packer tests in Wisconsin by Eaton and others (2000). 6. Discrete interval packer tests and dye-trace studies at the Spring Valley Amoco terminal by Delta Environmental Consultants, Inc. (1995, 1998), Fillmore County. 7. Discrete interval packer tests in Illinois by Nicholas and others (1987). 8. Discrete interval slug tests in Wisconsin by Stocks (1998). 9. Based on laboratory analysis of plug tests and typical results of pump tests of similar strata (Freeze and Cherry, 1979). 10. Standard aquifer pump tests at Reilly Tar and Chemical site in St. Louis Park, Hennepin County, reported by ERT (1987) and ENSR International (1991); at General Mills Solvent Disposal site in northeast Minneapolis, Hennepin County, by Barr Engineering (1991); and at Minnehaha Park tunnel in Hennepin County by Liesch (1973). These and other site investigations with similar results are summarized in Hoffman and Alexander (1998). 11. Dye-trace investigations in western Wisconsin (Hoffman and Alexander, 1998) and at Camp Coldwater Spring, Hennepin County (Alexander and others, 2001). 12. Standard aquifer pump tests and modeling in the seven-county Twin Cities Metropolitan area by Schoenberg (1991). 13. Seven boreholes packer tested in southern Wisconsin (Young, 1992). 14. Standard aquifer tests of several wells at the Reilly Tar and Chemical site in St. Louis Park, Hennepin County. Includes tests of St. Louis Park municipal well number 3, and monitor well W410 (Barr Engineering, 1976; and an anonymous report in the Minnesota Pollution Control Agency site files). 15. Standard aquifer tests of four wells in Ramsey and Hennepin Counties reported in Norvitch and Walton (1979). Conductivity was calculated using an aquifer thickness of 100 feet. 16. Standard aquifer test at the Nutting Site in Faribault, Rice County (Barr Engineering, 1986). 17. Fillmore County dye-trace study by Wheeler (1993). 18. Standard aquifer pump test of Spring Grove municipal well #4 (unique number 433257), Houston County (Eder and Associates, 1997). 19. Discrete interval tests and dye-trace studies at Oronoco Landfill, Olmsted County, by Donahue and Associates, Inc. (1991) and RMT, Inc. (1992). 20. Borehole flowmeter logging and pumping at wells in:

32

A. Faribault (unique number 625327), Rice County. B. Rochester (unique number 485610), Olmsted County. 21. Standard aquifer pump test of a well at Chatfield Fish and Game Club in Fillmore County (unique number 227394). 22. Twenty-six standard aquifer pump tests conducted in southeastern Minnesota. Tests of 12 boreholes located in the seven-county Twin Cities Metropolitan area are reported by Runkel and others (1999). Tests of 14 boreholes outside of the metropolitan area are from unpublished data compiled by the U.S. Geological Survey and include the following: Rochester Municipal wells 23 (unique number 220660), 27 (unique number 224212), 28 (unique number 180567), 29 (unique number 161425), 30 (unique number 239761), 31 (unique number 434041), 32 (unique number 506819), and 34 (unique number 463536), Rochester public schools wells for Ridgeway (unique number 235583), Burr Oak (unique number 220615), and Golden Hill (unique number 220679), all in Olmsted County, and Rice County wells at Carleton College (unique number 171005), Dundas (unique number 132294), and St. Olaf College (no number). 23. Standard aquifer pump tests at the New Brighton and Arden Hills Twin Cities Army Ammunition Plant site, Ramsey County, by Camp, Dresser and McKee (1991). 24. Standard aquifer pump test by the Minnesota Department of Natural Resources and Minnesota Department of Health at Plainview in Wabasha County. 25. Based on discrete interval packer tests of similar strata of other parts of the Paleozoic section in Ramsey County by Miller and Delin (1993). 26. Three boreholes packer tested in southwest Wisconsin (Young, 1992). 27. Discrete interval packer test in Ramsey County by Miller and Delin (1993). 28. Discrete interval packer tests and thermal profiling at the ATES Project site in Ramsey County by Kanivetsky (1989). 29. Two boreholes packer tested in southwest Wisconsin (Young, 1992). 30. Borehole flowmeter and discrete interval slug tests in southeastern Wisconsin (Swanson, 2001). Strata tested include interbeds of the Mazomanie Member. 31. Slug testing of monitor wells at a proposed ash disposal site at Red Wing in Goodhue County by Wenck and Associates, Inc. (1997). 32. Standard aquifer pump tests at Lakeland, Washington County by Braun Intertec (1992). 33. Standard aquifer pump test of Peterson Fish Hatchery (unique number 467232), Fillmore County. 34. Nine wells packer tested in Illinois and southern Wisconsin, reported by Young (1992). 35. Discrete interval slug tests in southeastern Wisconsin (Bradbury, 2001). 36. Standard aquifer pump test of Goodview municipal well 3 (unique number 449410), Winona County. 37. Packer tests in Wisconsin by Carlson and Taylor (1999). 38. Seventeen wells packer tested in southern Wisconsin and Illinois, reported by Young (1992). Key to lithostratigraphic units:
mts—Mt. Simon Sandstone
frn—Franconia Formation Opod—Prairie du Chien Group– Oneota Dolomite Ogwd—Glenwood Formation Ogal—Galena Group Dspl—Spillville Formation Dcum—Coralville Formation and Hinkle and Eagle Center Members of the Little Cedar Formation


ecr—Eau Claire Formation
stl—St. Lawrence Formation Opsh—Prairie du Chien Group– Shakopee Formation Opvl—Platteville Formation Odub—Dubuque Formation Dclp—Pinicon Ridge Formation and Bassett Member of the Little Cedar Formation Dcuu—Lithograph City Formation

33


igl—Ironton–Galesville Sandstone
jdn—Jordan Sandstone Ostp—St. Peter Sandstone Odcr—Decorah Shale Omaq—Maquoketa Formation Dclc—Chickasaw Member of the Little Cedar Formation

DEEP BEDROCK CONDITIONS

SHALLOW BEDROCK CONDITIONS

80

80

60

60

Range

Range

100

100

40

40

20

20

0

0
mts


igl


frn1
frn2
stl


jdn

Opdc Ostp


mts
ecr
igl
frn1
frn2
stl
jdn Opdc Ostp Opvl Odcr Ogal Dcom

Ogal Dspl

SHALLOW BEDROCK CONDITIONS

600

600

500

500

400

400

Range

Range

DEEP BEDROCK CONDITIONS

300

300

200

200

100

100

0

0
mts
igl


frn1
frn2
stl
jdn Opdc Ostp


mts
ecr
igl
frn1
frn2
stl
jdn Opdc Ostp Opvl Odcr Ogal Dcom

Ogal Dspl

Figure 14. Comparison of conductivity values (in feet per day) calculated from specific capacity data. The datasets for deep and shallow bedrock conditons are each plotted at two different scales. In deep bedrock conditions, the formations with the highest conductivity are those that contain coarse clastic strata, or are composed of carbonate rock with dissolution cavities. The former include the Mt. Simon, Ironton–Galesville, Jordan, and St. Peter Sandstones, and the Franconia Formation only where the Mazomanie Member is present. The formations that are composed of carbonate rock with dissolution cavities include the St. Lawrence Formation and the Prairie du Chien Group. In shallow bedrock conditions, conductivity is much more variable, and each of the lithostratigraphic units can have moderate to high conductivities, regardless of matrix characteristics. Therefore all are used as an economic source of ground water. The database from which these conductivity values have been calculated consists of tests of water wells constructed expressly for the purpose of extracting economic quantities of water. The values of conductivity are therefore chiefly representative of the most productive intervals of Paleozoic strata in a given geologic setting, and do not include a large sample of values representative of intervals of strata with relatively low conductivity. Lithostratigraphic units:
mts—Mt. Simon Sandstone
ecr—Eau Claire Formation
igl—Ironton–Galesville Sandstone
frn 1—Franconia Formation where the Mazomanie Member is thin to absent
frn 2—Franconia Formation where the Mazomanie Member forms a substantial component of the formation
stl—St. Lawrence Formation
jdn—Jordan Sandstone Opdc—Prairie du Chien Group Ostp—St. Peter Sandstone Opvl—Platteville Formation Odcr—Decorah Shale Ogal—Galena Group Dspl—Spillville Formation Dcom—All strata above the Galena Group

34

Lower Mt. Simon Sandstone

Upper Mt. Simon Sandstone

Eau Claire Formation

A.

A

Increasing counts

200190

St. Paul

500 feet

A

A'

brian "base cam Pr e nt "

213644 and 213646

235526

226617

35

e

cam

Pre

n bria

nt"

eme

"bas

Figure 15. A. Hydrostratigraphic attributes of the Mt. Simon Sandstone and representative natural gamma logs across part of southeastern Minnesota. Note that although the Mt. Simon Sandstone consists chiefly of coarse clastic strata, it also contains substantially thick intervals dominated by the fine clastic component, particularly in its upper part. Such interbeds were determined to be of low conductivity and have the ability to provide confinement in the Waseca–Waterville area. Unique numbers are listed above the boreholes. B. Natural gamma log and hydrogeologic units defined for the hydrostratigraphically similar Mt. Simon Sandstone in Illinois (Nicholas and others, 1987).

Fine clastic component

Coarse clastic component

Lower Mt. Simon aquifer

Cap Rock B

Middle Mt. Simon aquifer

Cap Rock A

Upper (transition) Mt.Simon aquifer

Waseca–Waterville

Lonsdale

Vermillion

Lower Mt. Simon aquifer

Mt. Simon confining unit

Upper Mt. Simon aquifer

Eau Claire Formation

Increasing counts

Illinois

223082

219022

B.

Hollandale

Owatonna

200 feet

m

A' 467232

Lower Mt. Simon Sandstone

Upper Mt. Simon Sandstone

Eau Claire Formation

Rushford

0

100% 0

5

0

Vertical permeability

Porosity

Fractures per foot

Visual porosity

Depth below bedrock surface

Lithostratigraphic unit Matrix hydrostratigraphic component

Gamma log

40% 10-6 100 md

200'

Prehn #3 Pratt #3

Oneota Dolomite

Increasing counts

St. Lawrence Formation

Jordan Sandstone

250'

300'

Coarse clastic component 350'

Fine clastic component Carbonate component

400'

Interval of no core recovery

Franconia F ormation

450'

500'

Mt. Simon Sandstone

Lloyd Williams #4 Prehn #3 and Steinhaus #1

Eau Claire Ironton–Galesville Formation Sandstone

550'

600'

650'

700'

Mt. Simon Cap Rock A

750'

36

Figure 16. Distribution of porosity and permeability of the Mt. Simon Sandstone through Oneota Dolomite in deep bedrock conditions in the Waseca–Waterville area. Matrix porosity and permeability is low except in the coarse clastic component. Secondary porosity is minimally developed except for dissolution cavities developed in specific horizons of carbonate rock in the St. Lawrence Formation and Oneota Dolomite, and in carbonate-rich intraclasts in the Franconia Formation. The Oneota Dolomite cavities were developed about 150 to 200 feet below the bedrock surface, in the transition between shallow and deep bedrock conditions. Plug samples were collected at approximately onefoot intervals. Gray shading on the porosity and permeability logs are estimated values corresponding to intervals of no core recovery, chiefly where coarse clastic beds were present, and are based on plug tests of nearby cores and from Setterholm and others (1991). Cores Pratt 3, Prehn 3, Lloyd Williams 4, and Steinhaus 1.

Conductivity in feet per day

A.

300

Figure 17. Hydraulic conductivity data for the Mt. Simon Sandstone calculated from specific capacity tests. See Figure 11 for an explanation of box plots. A. Scatter plot showing the relationship between the depth of the open-hole interval below the bedrock surface and hydraulic conductivity. Shallower wells tend to have higher conductivity. B. Box plot of hydraulic conductivity values for deep bedrock conditions. Two outlying values greater than 200 feet per day are not shown. C. Box plot of hydraulic conductivity values for shallow bedrock conditions. One outlying value of 990 feet per day is not shown.

250 200 150 100 50 0 0

200

400

600

800

1000

Distance in feet between the bedrock surface and the open-hole top

C.

DEEP BEDROCK CONDITIONS

SHALLOW BEDROCK CONDITIONS

200

200

150

150

Range

Range

B.

100

100

50

50

0

0 25 samples Average 39.5 feet per day

165 samples Average 29.3 feet per day

inches in the fine clastic component.

Hydraulic attributes Deep bedrock conditions—Discrete interval pump tests of the fine clastic component in the Eau Claire Formation in deep bedrock conditions measured a horizontal hydraulic conductivity value of less than 102 foot per day (Miller and Delin, 1993). Fine clastic strata of the Eau Claire Formation in southeastern Wisconsin yielded similar values of horizontal conductivity based on slug tests, about 10-3 to 10-2 foot per day (Bradbury, 2001; Swanson, 2001). Vertical conductivity has been estimated at 10-4 foot per day (Fig. 10; Miller and Delin, 1993), which is consistent with the low intergranular permeability and anisotropy measured in plug samples. Coarse clastic beds in the upper part of the formation have not been individually

37

tested, but can be expected to have a hydraulic conductivity of as much as a few tens of feet per day based on tests of similar coarse clastic strata in deep bedrock conditions. Only seven wells in the County Well Index database draw water from the Eau Claire Formation in deep bedrock conditions. These wells are open to the upper part of the formation in Winona, Houston, and Wabasha Counties, where the formation is known to contain coarse clastic interbeds. Well records did not contain the information necessary to calculate hydraulic conductivity. Shallow bedrock conditions—Scientifically rigorous hydraulic tests of the Eau Claire Formation under shallow bedrock conditions have not been conducted, but detailed, site-specific studies of

hydrostratigraphically similar parts of the Franconia Formation (described later in this report) suggest that the Eau Claire Formation in shallow bedrock settings probably has a great range in hydraulic conductivity, from 10-2 foot to a few tens of feet per day. Hydraulic conductivity calculated from specific capacity tests of 249 wells open to the Eau Claire Formation in shallow bedrock settings typically range from less than one to as much 100 feet per day, and average 36.7 feet per day (Fig. 19). These values are within the typical range for coarse clastic units that have a much greater matrix permeability than the Eau Claire Formation, such as the Jordan aquifer, and therefore support the hypothesis that fracture porosity in the Eau Claire Formation can at least locally result in moderately high bulk hydraulic conductivity.

Hydrogeologic synthesis Previous work based on potentiometric data plotted at both local and regional scales has clearly demonstrated that all or part of the Eau Claire Formation has the ability to function as a confining unit in deep bedrock conditions (for example Delin and Woodward, 1984; Miller, 1984), even though intercalations of the coarse clastic component in its upper part may be utilized as local aquifers in saturated conditions (Runkel, 1996a). The "greensand" facies in Faribault and adjacent counties could yield moderate quantities of water in deep bedrock conditions, based on relatively high values of plug permeability, but it has not been hydraulically tested as a discrete hydrogeologic unit. The hydrogeologic character of the Eau Claire Formation in shallow bedrock conditions may be much more complex than in deep conditions as a result of flow along fractures. Visible seeps occur along beddingplane fractures in bluffside exposures of the Eau Claire Formation in Wabasha County along U.S. Highway 61, 2.5 miles south of its intersection with U.S. Highway 60. In addition, the Eau Claire Formation is used as a source of water for over 400 wells in the County Well Index database, demonstrating that it can yield moderate to even large quantities of water, and be properly classified as an aquifer, in shallow bedrock conditions. The majority of these wells (335) are located in the northern part of the Twin Cities Metropolitan area and along the St. Croix and Mississippi Rivers from Chisago County south to the Iowa border, where the Eau Claire Formation is the uppermost bedrock. Given the low intergranular permeability of most of the Eau Claire Formation, water in these wells is probably drawn chiefly through fracture networks. The relative effectiveness and scale at which the Eau Claire Formation can provide confinement of the underlying

38

Mt. Simon Sandstone in such a setting is in need of further investigation.

IRONTON AND GALESVILLE SANDSTONES Hydrostratigraphic attributes Matrix porosity The Ironton and Galesville Sandstones have historically been combined as a single map unit because both are dominated by the coarse clastic component (Figs. 10, 16, 18). However, this unit can be divided, at least locally, into two subcomponents: a "clean" coarse clastic component and a shaly coarse clastic component (Runkel, 1996a). The latter component contains an appreciable amount of silt and shale as thin interbeds, and as matrix between sand grains. It typically occurs in the upper one-half of the Ironton– Galesville Sandstone, an interval corresponding to the uppermost Galesville and lowermost Ironton lithostratigraphic units. The clean and shaly coarse clastic components have not been studied in the detail necessary to quantitatively compare their permeability. Plug samples of the clean coarse clastic component commonly have a permeability of 10 2 to 10 3 md in both a horizontal and vertical direction (Figs. 10, 16, 18). The shaly component may have a much lower permeability, especially in a vertical direction, because of the presence of clay and silt-sized particles filling pore spaces between sand grains and as thin laminations.

Secondary porosity Deep bedrock conditions—Our knowledge of secondary porosity in deep bedrock conditions is extremely limited. Inasmuch as the Ironton–Galesville Sandstone is a friable, high porosity layer of strata covered by younger bedrock of contrasting material properties, networks of interconnected, open vertical fractures are believed to be poorly developed at best (Price and Cosgrove, 1990; Helgeson and Aydin, 1991; Narr and Suppe, 1991). Thin (a few inches or less) bedding-plane fractures can be seen on video logs of some boreholes (for example unique number 255768, Minnesota Department of Health Borehole Video Library), but their abundance and lateral extent is not known. An unproven assumption by previous investigators is that fractures and dissolution features are uncommon in the Ironton–Galesville Sandstone in deep bedrock settings and therefore porosity and permeability is determined chiefly by intergranular attributes. Shallow bedrock conditions—The

100%

0

Vertical permeability

Porosity

Fractures per foot

Visual porosity

Depth below bedrock surface

Matrix hydrostratigraphic component

Lithostratigraphic unit

0

5

0

40%

10-6

100 md

Jordan Sandstone

Oneota Dolomite

500'

550'

Coarse clastic component St. Lawrence Formation

Fine clastic component 600'

Carbonate component

Interval of no core recovery

Franconia Formation

650'

700'

Ironton–Galesville Sandstone

750'

800'

900'

Schroeder #5 Schroeder #1

"Greensand facies"

Eau Claire Formation

850'

950'

1000'

1050'

Mt. Simon Sandstone

Gamma log Increasing counts

1100'

39

Figure 18. Distribution of porosity and permeability of the Mt. Simon Sandstone through Oneota Dolomite in deep bedrock conditions at Bricelyn. Matrix porosity and permeability is low except in the coarse clastic component. Secondary porosity is minimally developed except for dissolution cavities developed in specific beds of carbonate rock in the St. Lawrence and lower Franconia Formations, the Oneota Dolomite, and in carbonate-rich intraclasts in the upper part of the Franconia Formation. Plug samples were collected at approximately one foot intervals. Gray shading on the porosity and permeability logs are estimated values corresponding to intervals of no core recovery, chiefly where coarse clastic beds were present, and are based on plug tests of nearby cores and from Setterholm and others (1991). Cores Schroeder 1 and Schroeder 5.

hydrostratigraphic character of the Ironton–Galesville Sandstone in shallow bedrock conditions has not been studied in detail. Systematic and nonsystematic fractures with apertures as wide as a few inches are common in the only laterally extensive exposures in Minnesota—a series of road cuts on the west side of U.S. Highway 61 between the cities of Wabasha and La Crescent.

Hydraulic attributes Deep bedrock conditions—Discrete interval and standard aquifer pump tests in southeastern Minnesota and nearby parts of southwestern Wisconsin yielded a relatively narrow range of hydraulic conductivity values from a few feet per day to as high as 15 feet per day (Young, 1992; Miller and Delin, 1993). Discrete interval packer tests of nine boreholes in southeastern Wisconsin and Illinois are consistent with those values with the exception of one value of 31 feet per day (Figs. 10, 13). Vertical hydraulic conductivity values are estimated to be about one-tenth of horizontal values (Miller and Delin, 1993). Discrete interval packer tests by Miller and Delin (1993) at the Aquifer Thermal Energy Storage Project (ATES) site in St. Paul (Fig. 1) indicated that an interval dominated by the clean coarse clastic component has a horizontal hydraulic conductivity value about 58 percent greater than an interval dominated by the shaly coarse clastic

B.

A. 400

400

350

350

300

300

250

250 Range

Conductivity in feet per day

component (Fig. 10). Borehole flowmeter investigations of the Ironton– Galesville Sandstone near Savage, Hastings, and Prior Lake (Fig. 1) demonstrate that the coarse clastic strata have a relatively high intergranular permeability compared to adjacent fine clastic strata of the Eau Claire and Franconia Formations (Fig. 20; Paillet and others, 2000; this study). At each of these sites, ambient and induced borehole flow was not concentrated in thin, discrete intervals within the Ironton–Galesville Sandstone, but instead was evenly distributed across the coarse clastic component, a characteristic consistent with flow chiefly through intergranular pore spaces. Hydraulic conductivity values calculated from specific capacity tests of the Ironton–Galesville Sandstone are shown in Figure 21. In deep bedrock conditions, values typically range from 1.5 to 28 feet per day, with an average of 10.8 feet per day. These calculations are consistent with those determined through higher-quality pump tests and typical in magnitude for fine- to coarse-grained sandstone aquifers. Shallow bedrock conditions—Controlled pump test data for the Ironton–Galesville Sandstone in shallow bedrock conditions of Minnesota are not available. Bradbury (2001) and Swanson (2001) reported the results of discrete interval slug tests and borehole

200

200

150

150

100

100

50

50 0

0 0

50

100

150

200

250 samples Average 36.7 feet per day

Distance in feet between the bedrock surface and the open-hole top

Figure 19. Hydraulic conductivity data for the Eau Claire Formation calculated from specific capacity tests. See Figure 11 for an explanation of box plots. A. Scatter plot showing the relationship between the depth of the open-hole interval below the bedrock surface and hydraulic conductivity. Shallower wells tend to have higher conductivity. B. Box plot of hydraulic conductivity values for shallow bedrock conditions.

40

flowmeter logs of the Ironton–Galesville Sandstone in shallow conditions of southeastern Wisconsin, demonstrating a range in conductivity from 10 to 100 feet per day, with a geometric mean of 20 feet per day. They identified two discrete intervals of preferential flow within the Ironton–Galesville Sandstone. Hydraulic conductivity values calculated from specific capacity tests of wells constructed in shallow bedrock conditions of southeastern Minnesota range from less than one to 60 feet per day, with an average value of 26.8 feet per day (Fig. 21). The relatively high average conductivity and large number of outlying values greater than 60 feet per day likely reflect some component of fracture flow.

Hydrogeologic synthesis The Ironton–Galesville Sandstone can be treated as a single aquifer in which water moves chiefly through large, well connected intergranular pore spaces in deep bedrock settings. The relative proportion of the shaly coarse clastic component to the clean coarse clastic component may have a measurable control on aquifer productivity, much as the relative abundance of distinct hydrostratigraphic components within the Jordan Sandstone are known to control its productivity across southeastern Minnesota (Runkel, 1996b; Runkel and others, 1999; this report). The relative proportion of these two components in the Ironton–Galesville Sandstone can be estimated, with some difficulty, through gamma logs and cuttings at individual well sites. The presence of abundant fractures no doubt plays an important role in the hydraulic behavior of the Ironton–Galesville Sandstone in shallow bedrock conditions, although fracture flow in this unit has not been studied in detail in Minnesota. Preferential flow paths along narrow intervals of the Ironton–Galesville Sandstone in Wisconsin (Bradbury, 2001; Swanson, 2001) may reflect the presence of hydraulically significant bedding-plane fractures that can dominate flow systems of siliciclastic aquifers where they occur in shallow bedrock conditions (Michalski and Britton, 1997; Morin and others, 1997).

FRANCONIA FORMATION Hydrostratigraphic attributes Matrix porosity The intergranular attributes of the Franconia Formation are better documented than those of any other Paleozoic lithostratigraphic unit because porosity and permeability have been calculated for hundreds of plugs sampled from cores collected as far north as St. Paul and as far south as Faribault County (Figs. 10, 16, 18; MUGSP, 1980; Miller and Delin, 1993). Matrix

41

porosity in outcrops along the Mississippi River has also been studied (Runkel and Tipping, 1998). Collectively, this work demonstrates that the bulk of the Franconia Formation across southeastern Minnesota is composed of the fine clastic component, with subordinate beds of carbonate strata. As such the Franconia Formation is similar in grain size, stratification, and degree of cementation to the dominant component of the Eau Claire Formation. Plug samples of fine clastic and carbonate rock components in the Franconia Formation have a vertical permeability that typically ranges from 10-6 to 10-2 md, similar to values calculated for the Eau Claire Formation. Coarse clastic beds of relatively high permeability are common in the upper part of the Franconia Formation, and are of sufficient thickness in the northern and eastern Twin Cities Metropolitan area to be defined as a mappable member, the Mazomanie Member (Fig. 22). Stratigraphic cross-sections and isopachs of the Mazomanie Member indicate that it forms a substantial part of the Franconia Formation (greater than 20 percent of the formation) in parts of Anoka, Chisago, Hennepin, Isanti, Ramsey, Sherburne, Washington, and Wright Counties. Outside of this eight-county area, thin to medium coarse clastic beds are a relatively minor component of the Franconia Formation, most commonly present in its uppermost part, and in strata that are transitional with the underlying Ironton Sandstone. Such interbeds tested as high as 340 md in a vertical direction (Runkel, unpub. data, 1998).

Secondary porosity Deep bedrock conditions—Most of what is known about secondary porosity in fine clastic strata in southeastern Minnesota is based on comprehensive, site specific investigations of the Franconia Formation conducted over the past ten years (Delta Environmental Consultants, Inc., 1992; Miller and Delin, 1993; Wenck and Associates, Inc., 1997; core logging for this report). In deep bedrock conditions, secondary porosity is believed to be relatively uncommon based on examination of core and borehole video and geophysical logs (Figs. 10, 16, 18). Bedding-plane fractures in the upper and lower parts of the formation are documented, however (Fig. 20). Additionally, subvertical mesoscopic fractures occur in some intervals but are rare, and typically are open less than one millimeter or are closed. Dissolution cavities are common in carbonate intraclasts and in a relatively thin, carbonate-dominated interval in the lower part of the formation. Shallow bedrock conditions—The Franconia Formation in shallow bedrock conditions is well documented because of an ash disposal site evaluation near Red Wing (Wenck and Associates, Inc., 1997), and

a similar but less comprehensive study near Lakeland (Delta Environmental Consultants, Inc., 1992). Wenck and Associates, Inc. (1997) used borehole video logs, cores, and surface exposures at the proposed ash disposal site near Red Wing to demonstrate that the Franconia Formation is characterized by a dense network of nonsystematic vertical and horizontal fractures in shallow bedrock conditions. In one monitor well at this site (unique number 575374), the Franconia Formation had 10 horizontal and 2 vertical fractures that were of sufficient width to be recognized on a borehole video log along a 131-foot section of open hole. A test pit at the bedrock surface revealed 11 subvertical fractures, ranging from 36 to 110 inches in length, over a surface area of only 162 square feet. Outcrops along the Mississippi River and its tributaries from Red Wing south to the Iowa border have vertical and horizontal fractures developed to a comparable or greater degree than that documented at the proposed ash disposal site. Borehole caliper logs collected in Anoka and Washington Counties similarly show the

presence of fractures in the lower Franconia Formation (Fig. 23). These features are a classic manifestation of stress-relief unloading and weathering in thinly bedded, moderately cemented, sedimentary rocks and should be expected in abundance anywhere the Franconia Formation is within tens of feet of the bedrock surface, particularly along valley walls.

Hydraulic attributes Deep bedrock conditions—Values of horizontal hydraulic conductivity in the fine clastic component of the Franconia Formation in deep bedrock conditions range from 10-3 to 10-2 foot per day (Fig. 10) based on discrete interval packer tests conducted by Miller and Delin (1993). Vertical hydraulic conductivity is roughly two orders of magnitude less than horizontal hydraulic conductivity. Several 20-foot intervals of Franconia Formation tested by Miller and Delin (1993) yielded no discharge and must have significantly lower hydraulic conductivity values. Coarse clastic beds of the

Figure 20. Borehole geophysical logs of the Eau Claire through St. Lawrence Formations from sites in Savage, Prior Lake, and Hastings. All three logs demonstrate that the Ironton and Galesville Sandstones together have the properties of an aquifer, whereas the overlying lower to middle Franconia Formation strata serve as a confining unit. Bedding-plane fractures in the upper Franconia and the St. Lawrence Formations are hydraulically active, even though strata between the fractures may serve as confining units. See Figure 1 for locations of these sites, and Figure 5 for an explanation of flowmeter logs. A. Flowmeter data were collected while trolling up at 10 feet per minute during injection, and also with the tool stationary at various depths under ambient and 9 gallons per minute injection conditions. The percent of borehole outflow through the open-hole interval of the well during injection is noted to the right of the flowmeter logs. Note that the bulk of the outflow is through the coarse clastic component of the Ironton–Galesville Sandstone. The flowmeter log has a relatively even signature across that interval, consistent with outflow through intergranular pore spaces. The fine clastic component accounts for only 18 percent of the outflow, from a single discrete horizon in the lower part of the Franconia Formation. Savage, unique number 593579; modified from Paillet and others (2000). B. Trolling (at 10 feet per minute) and stationary flowmeter logs indicate that bedding-plane fractures in the lowermost St. Lawrence and upper part of the Franconia Formations yield water that travels down the borehole at a mimimum rate of about 7 gallons per minute under ambient conditions. This downflow exits the hole gradually, in intergranular fashion, in the Ironton–Galesville Sandstone. The intervening middle to lower Franconia Formation serves as a confining unit. Prior Lake, unique number 672729. C. On page 44. Trolling (at 10 feet per minute) and stationary flowmeter logs indicate that the Ironton–Galesville Sandstone yields water that travels up the borehole at a minimum rate of about 1 gallon per minute under ambient conditions. Additional borehole upflow is added through bedding-plane fractures at the St Lawrence– Franconia Formation contact and in the lower part of the St. Lawrence Formation. Water exits the borehole through bedding-plane fractures in the upper St. Lawrence Formation, and through a hole in the casing at 84 feet (not shown on this figure). The lower to middle Franconia Formation, and the middle part of the St. Lawrence Formation serve as confining units at this site. Hastings, unique number 255768.

42

A. Depth in feet below the land surface and bedrock surface (italics)

Matrix hydrostratigraphic Gamma log component (counts per second) 0

100

200

Caliper (inches) 0

4

Trolling flow, during injection (gallons per minute)

Televiewer 8

-5

-10

Stationary flow (gallons per minute) -15

0

-5

-10

0

5

Interpretation No data

450

St. Lawrence Formation

400

500

550

Franconia Formation

300

Casing bottom Ambient flow

Location and percent of borehole transmissivity of dominant permeable intervals

400 600

650

Ironton–Galesville Sandstone

18%

82% Injection flow

Eau Claire Formation

B. Depth in feet below the land surface and bedrock surface (italics)

Matrix hydrostratigraphic component

300

St. Lawrence Formation

0

400

100

Trolling flow, ambient conditions (gallons per minute)

Caliper (inches)

Gamma log (API units) 200

4.7

5.3

-6

-4

-2

0

Stationary flow, ambient conditions (gallons per minute) -6

-4

Casing bottom

Interpretation

500 400

Franconia Formation

450

600 500 Eau Claire Formation

Ironton–Galesville Sandstone

550

650 Coarse clastic component Fine clastic component Carbonate component

Figure 20 continued on page 44.

43

-2

0

C.

Depth in feet below the land surface and bedrock surface (italics)

Matrix hydrostratigraphic component 0

100

Trolling flow, ambient conditions (gallons per minute)

Caliper (inches)

Gamma log (API units) 200

6

9

12

0

1

2

Stationary flow, ambient conditions (gallons per minute) 0

1

2

250

Casing bottom

St. Lawrence Formation

200 200

Jordan Sandstone

400

Interpretation

350

Franconia Formation

300 300

450

Eau Claire Formation

Ironton–Galesville Sandstone

400 400

500 500

Explanation to Figure 20C on page 42.

Mazomanie Member in the upper part of the Franconia Formation, which are about 40 feet thick at the site, have a moderately high horizontal conductivity of about 1.4 to 7.5 feet per day, and a bulk conductivity of about 2 feet per day. Hydraulic conductivity calculated from specific capacity data for the Franconia Formation (Fig. 24) under deep bedrock conditions reflects its intergranular hydrostratigraphic variability. In the northern and eastern Twin Cities Metropolitan area, where the Mazomanie Member forms a substantial part of the Franconia Formation (Fig. 22), conductivity typically ranges from less than one to about 65 feet per day, and averages 27.8 feet per day. These values are analogous to those calculated for the Jordan Sandstone, which is similarly dominated by the coarse clastic component. In contrast, outside of the northern and eastern metropolitan areas, where the Mazomanie Member is thin or absent, the Franconia Formation typically ranges in conductivity from less than one to 10 feet per day, with an average value of 5.9 feet per day.

44

The comprehensive borehole geophysical study of an observation well near Savage (Paillet and others, 2000) yielded information about the relationship between hydrostratigraphic character and ground-water hydraulics in the lower part of the Franconia Formation and subjacent Ironton–Galesville Sandstone (Fig. 20). Flowmeter logs across 150 feet of open hole that exposes the upper Eau Claire through lower Franconia Formations indicated that all the borehole outflow during injection occurred through intergranular pore spaces in the Ironton–Galesville Sandstone and to a much lesser degree through a four-foot interval within the lowermost part of the Franconia Formation. In contrast, the open-hole intervals that intersect fine clastic strata, which are the dominant component of both the Eau Claire and Franconia Formations, were of insufficient conductivity to accommodate measurable outflow of injected water. Shallow bedrock conditions—Hydraulic conductivity values under shallow bedrock conditions

Conductivity in feet per day

A.

250

Figure 21. Hydraulic conductivity data for the Ironton–Galesville Sandstone calculated from specific capacity tests. See Figure 11 for an explanation of box plots. A. Scatter plot showing the relationship between the depth of the open-hole interval below the bedrock surface and hydraulic conductivity. Shallower wells tend to have higher conductivity. B. Box plot of hydraulic conductivity values for deep bedrock conditions. C. Box plot of hydraulic conductivity values for shallow bedrock conditions.

200

150

100

50

0 0

100

200

300

400

500

600

700

Distance in feet between the bedrock surface and the open-hole top

C.

DEEP BEDROCK CONDITIONS

SHALLOW BEDROCK CONDITIONS

250

250

200

200

150

150

Range

Range

B.

100

100

50

50

0

0

332 samples Average 26.8 feet per day

26 samples Average 10.8 feet per day

are variable because of the effect of flow along discrete fractures. At the ash disposal site near Red Wing, hydraulic conductivity values of the fine clastic component ranged over four orders of magnitude, from 10 -2 to 85 feet per day (Wenck and Associates, Inc., 1997). Such a large range in values reflects the variable size and degree of connectivity of fractures across the site. The hydraulic conductivity of the lower one-half of the Franconia Formation where it occurs as the uppermost bedrock in Lakeland in southern Washington County (Fig. 1) was measured at between 0.14 and 0.21 foot per day (Delta Environmental Consultants, Inc., 1992). Flowmeter logging of an injected well in Anoka County (Fig. 24) indicated that a single fracture had a conductivity of 90 feet per day, but the overlying 35 feet of fine clastic Franconia Formation were of insufficient conductivity to accommodate measurable

45

outflow of the injected water. The Franconia Formation has recently been studied in southeastern Wisconsin where it shows a similar large range in hydraulic conductivity in shallow bedrock conditions (Swanson, 2001). It is composed of interbedded fine and coarse clastic strata in that area, and the bulk of the formation has an average conductivity of about 3 feet per day. Ground water flows preferentially along thin (less than 5 feet) beddingplane parallel intervals that average 220 feet per day in conductivity. These intervals appear to be laterally continuous across an entire watershed that exceeds 10 square miles (Swanson, 2001). Hydraulic conductivity of the Franconia Formation calculated from specific capacity tests in southeastern Minnesota also reflects the importance of fracture porosity in shallow bedrock conditions (Fig. 24), especially in areas where the Mazomanie Member is

thin or absent and wells are open chiefly to fine clastic strata. Outside of the northern and eastern Twin Cities Metropolitan areas, the Franconia Formation typically ranges in conductivity from less than one to 40 feet per day, with an average value of 32.3 feet per day, roughly five times greater than its conductivity in conditions where it is deeply buried by younger bedrock. Within the counties where the Mazomanie Member forms a substantial part of the formation, conductivity typically ranges from less than one to 75 feet per day, with an average value of 31.7 feet per day.

Hydrogeologic synthesis The Franconia Formation has previously been characterized as a single, more or less homogeneous hydrogeologic unit across all of southeastern Minnesota: either as a moderately productive aquifer in good hydraulic connection with the underlying Ironton– Galesville aquifer (for example Kanivetsky, 1978), or as a confining unit combined with the overlying St. Lawrence Formation (for example Delin and Woodward, 1984). Our characterization of the Franconia Formation, which interprets hydraulic data within the context of its hydrostratigraphic properties, demonstrates that it is very heterogeneous in its hydrogeologic properties. It therefore is best divided into two distinct hydrogeologic units: the upper part of the Franconia aquifer is a widely-used aquifer that yields water through high permeability coarse clastic strata and along bedding-plane fractures across much of the Twin Cities Metropolitan area. In contrast the middle to lower Franconia Formation in the metropolitan area, and the entire formation to the south, is analogous to other low permeability fine clastic units described in this report, serving as a regionally extensive confining unit above the Ironton–Galesville aquifer. In deep bedrock conditions, the hydrostratigraphic and hydraulic character of the bulk of the Franconia Formation is similar to that of the Eau Claire Formation. The plug analyses (Figs. 10, 16, 18), packer tests, and borehole flowmeter logs described above collectively demonstrate that from a regional perspective, the bulk of both the Eau Claire and Franconia Formations are composed of low conductivity, fine clastic rock that has the ability to serve as a confining unit (MUGSP, 1980; Miller, 1984; Miller and Delin, 1993; Paillet and others, 2000; core logging for this report). In contrast, the Mazomanie Member, which from a regional perspective forms a subordinate component of the Franconia Formation (Fig. 22), is a high-conductivity hydrostratigraphic body that is best considered a discrete, local aquifer within the regional Franconia confining unit. It is largely composed of the coarse clastic component, with hydraulic properties roughly

46

analogous to those of the Jordan, St. Peter, and Ironton– Galesville Sandstones. Flowmeter logging of the Franconia Formation and adjacent units in deep bedrock conditions at Prior Lake (Fig. 20B) measured dynamic, strong, ambient flow that substantiates the ability of its middle to lower part to function as a confining unit, and also demonstrated that bedding-plane fractures in its upper few tens of feet can be hydraulically active. Bedding-plane fractures in the Franconia Formation (and possibly the lowermost St. Lawrence Formation) collectively yield a minimum of seven gallons per minute of water to the borehole under ambient conditions. This water travels down the borehole, exiting gradually across the Ironton–Galesville Sandstone. The relatively strong ambient flow that bypasses the middle and lower Franconia Formation, without significant contribution or subtraction of flow, is evidence that this part of the formation is of sufficiently low vertical conductivity to serve as a confining unit. Borehole tests of a number of wells open to the Franconia Formation in the western Twin Cities metropolitan area (conducted as this report was completed) in both shallow and deep bedrock conditions have yielded similar results. In shallow bedrock conditions, the Franconia Formation is much more hydrogeologically complex. At the proposed ash disposal site near Red Wing, ground water moves in intergranular fashion as well as through discrete hydraulic fractures (Wenck and Associates, Inc., 1997). Borehole videos at the site reveal water that seeps through intergranular spaces in thin beds of fineto medium-grained sandstone, particularly in the upper Franconia Formation. Water can also be observed flowing at a much higher rate along fractures developed in fine clastic strata. For example, water cascades from a horizontal fracture 100 feet below the ground surface in one borehole. The importance of fracture flow is also reflected by the wide-ranging and locally high values of hydraulic conductivity. Three of the five intervals of Franconia Formation rock tested at the ash disposal site had hydraulic conductivities at least two orders of magnitude higher than the same fine clastic component tested in a borehole with no open fractures (Miller and Delin, 1993). Recently collected flowmeter logs of injected boreholes open to the fine clastic strata of the Franconia Formation in Anoka and Washington Counties show similar conditions whereby the intergranular permeability is of insufficient conductivity to accommodate measurable outflow of injected water. Narrow fractures of a few inches, or coarse clastic interbeds of the underlying Ironton–Galesville Sandstone dominate the open-hole hydraulics (Fig. 23). Potentiometric data indicate that the fine clastic

47

0 feet

50

100

0

A

MOWER COUNTY

Increasing counts

221179

C.

A

B.

A.

A'

30 miles

25

150 100 Thickness in feet of 75 Mazomanie Member 50

226610

Coarse clastic component

185810

Fine clastic component

Site of ATES project

225611

Carbonate component

TWIN CITIES METROPOLITAN AREA

556149

Mazomanie Member

Ironton–Galesville Sandstone

Franconia Formation

St. Lawrence Formation

Ironton–Galesville Sandstone

Franconia Formation

St. Lawrence Formation

A'

Figure 22. Matrix hydrostratigraphic attributes of the Franconia Formation across southeastern Minnesota showing the distribution of coarse clastic and fine clastic components. Figure is based on measured sections by Berg (1954) along the St. Croix and Mississippi Rivers, and on natural gamma logs and cuttings in the subsurface to the west. A. Location of the cross-section and contoured thickness of the Mazomanie Member. B. Cross-section showing that the bulk of the formation consists of fine clastic strata of low permeability and hydraulic conductivity. Coarse clastic strata, referred to as the Mazomanie Member, are common only in its middle to upper parts in the Twin Cities Metropolitan area. C. Representative gamma logs that were among dozens used in conjunction with cuttings to map the distribution of the Mazomanie Member.

60 360

40 340

20 320

0 300

Carbonate component

Fine clastic component

Coarse clastic component

0

Depth in feet below Matrix the land surface and hydrostratigraphic bedrock surface (italics) component

B.

320

80

300

60

280

40

260

0

Depth in feet below Matrix the land surface and hydrostratigraphic bedrock surface (italics) component

A.

Franconia Formation

Franconia Formation

Ironton–Galesville Sandstone

48 150

Gamma log (API units)

150

Gamma log (API units)

Casing bottom

Casing bottom

300

300

4.5

4

Caliper (inches)

Caliper (inches)

5.5

5

-4

-3

-2 -1

0

-2

-1

0

1

Trolling flow— during injection (gallons per minute)

-3

Trolling flow— during injection (gallons per minute) 1

-4

-1

-2

-1

Ambient flow

Interpretation lines Injection flow

-3

Stationary flow (gallons per minute)

Ambient flow

Interpretation lines

Injection flow

-2

Stationary flow (gallons per minute)

0

0

49%

28%

23%

Location and percent of transmissivity of dominant permeable interval

90 feet per day

Location and hydraulic conductivity of dominant permeable interval

strata of the Franconia Formation have the ability to serve as a confining unit regardless of burial depth beneath younger bedrock. Differential static water levels of 10 to 15 feet between the Ironton–Galesville Sandstone and upper part of the Franconia Formation (Mazomanie Member) in nested well screens measured as part of the ATES project investigation demonstrates that the fine clastic component of the Franconia Formation serves as a confining unit in deep bedrock conditions (Walton and others, 1991). The lower to middle, fine clastic-dominated part of the Franconia Formation in shallow bedrock conditions near Red Wing is a confining unit that hydraulically separates a fractured upper Franconia Formation aquifer from a lowermost Franconia/Ironton–Galesville aquifer (Wenck and Associates, Inc., 1997). Potentiometric levels in the upper part of the formation mimic the land surface and show a pattern of local recharge and discharge, traits characteristic of water-table aquifers (Figs. 25, 26). The lowermost Franconia Formation and Ironton–Galesville Sandstone, perhaps connected by fractures at this site, have potentiometric levels as much as 60 feet lower in elevation than the water-table aquifer, and consistent with levels for the larger-scale, regional ground-water flow system. A site remediation study of a generally similar geologic setting near Lakeland in southern Washington County by Braun Intertec (1992) and Delta Environmental Consultants, Inc. (1992) yielded results analogous to those at Red Wing: the middle to lower part of the Franconia Formation hydraulically separates an upper water-table aquifer of fractured bedrock from an Ironton–Galesville aquifer, creating differential heads between the two aquifers of more than 40 feet. Pumping of a remediation well open only to the Ironton– Galesville aquifer caused only a 0.010-foot drawdown in the water-table aquifer at this site.

Field observations of outcrops and springs in southeastern Minnesota are compatible with subsurface results that suggest ground-water flow in shallow bedrock conditions occurs chiefly along a few discrete intervals with well developed secondary pores (Fig. 26, for example). Secondary pore networks parallel to bedding planes are particularly common in the carbonate-rich lowermost part of the Franconia Formation in Minnesota, and are principal water sources for bluffside springs along the Mississippi River, many of which emit ground water that has a "mixed" tritium chemistry (Runkel, 1996a; Runkel and Tipping, 1998; E.C. Alexander, Jr., unpub. data). Based on the characterization of the Franconia Formation made by Wenck and Associates, Inc. (1997) at Red Wing, such springs may be supplied by two distinct ground-water sources, which produce a mixed-tritium chemistry when mixed (Fig. 26): one source is vintage confined water from the regional Ironton–Galesville system that has upwelled through fractures in the lowermost Franconia Formation. A second source is recent water locally supplied through near-surface fractures along the sides of bluffs. Field observations suggest that the coarse clastic strata of the Mazomanie Member may also be dominated by fracture flow in shallow bedrock conditions. Bluffside springs along the St. Croix River are emitted from bedding-plane fractures in the Mazomanie Member, best exemplified by the "Boomsite" spring in the north end of Stillwater. The hydrostratigraphic variability of the Franconia Formation is reflected by the variable locations and geologic settings in which it is used as a domestic source of water. Its hydraulic conductivity is great enough to be utilized as an economic source of water chiefly where it occurs in shallow bedrock settings and where the Mazomanie Member is of substantial thickness in

Figure 23. Borehole geophysical logs of the Franconia Formation in shallow bedrock conditions. Flowmeter data were collected while trolling up at 10 feet per minute during injection, and also with the tool stationary at various depths under ambient and injection conditions. The logs of both holes demonstrated that the openhole hydraulics (measurable outflow in these examples) was dominated by thin (less than 1 foot) intervals of preferential flow in the Franconia Formation. The remaining parts of the Franconia Formation were of insufficient conductivity to accommodate measurable outflow. See Figure 5 for an explanation of flowmeter logs. A. Water injected at a rate of about 2.5 gallons per minute exits the borehole through a thin bedding-plane fracture that occurs within fine clastic strata of the middle part of the Franconia Formation. Hydraulic conductivity of the fracture is about 90 feet per day based on flow rate and static water level increase during injection. Anoka County unique number 165573. B. Water injected at a rate of about 3.5 gallons per minute exits the borehole through narrow bedding-plane fractures or thin coarse clastic interbeds in the lower Franconia Formation and upper Ironton–Galesville Sandstone. Washington County unique number 227031.

49

A.

400

Conductivity in feet per day

350 300 250 200 150 100 50 0 0

100

200

300

400

500

600

Distance in feet between the bedrock surface and the open-hole top

C.

DEEP BEDROCK CONDITIONS

SHALLOW BEDROCK CONDITIONS

200

200

150

150

Range

Range

B.

100

100

50

50

0

0 68 samples Average 27.8 feet per day

D.

1,872 samples Average 31.7 feet per day

DEEP BEDROCK CONDITIONS

SHALLOW BEDROCK CONDITIONS

E. 400

200

350 300

150

Range

Range

250 100

200 150 100

50

50 0

0

131 samples Average 32.3 feet per day

64 samples Average 5.9 feet per day

50

both shallow and deep bedrock conditions. Only 4.6 percent of the 5,126 wells open to the Franconia Formation are constructed in deep bedrock settings where the Mazomanie Member is thin or absent (County Well Index database). Examination of well construction records in nine counties outside of the Twin Cities Metropolitan area by Runkel (2000) indicated that thin to medium, coarse clastic beds in the uppermost part of the Franconia Formation and carbonate rock having small dissolution cavities and mesoscopic fractures in its lower part apparently are the most conductive parts of the formation in a deep bedrock setting, and supply relatively small to moderate yields of water for domestic use. The remaining 95 percent of Franconia Formation wells in southeastern Minnesota are constructed where the formation occurs in shallow bedrock conditions or where the Mazomanie Member reaches thicknesses greater than 20 feet. The extreme hydrostratigraphic and hydraulic conductivity variability within the Franconia Formation and its ability to serve as a confining unit, even in areas where the formation can yield economic quantities of water, should be a consideration in ground-water management practices. The thickness and aerial extent of the fine clastic component of the Franconia Formation that provides confinement is equal to or greater than that of most other historically recognized confining units in southeastern Minnesota. Additionally, the increasingly recognized importance of preferential flow along bedding-plane fractures (this study) has significant implications for predicting ground-water paths and travel times.

ST. LAWRENCE FORMATION

Hydrostratigraphic attributes Matrix porosity The St. Lawrence Formation consists of interbeds of the fine clastic and carbonate rock components. Across most of southeastern Minnesota, it is dominated by the carbonate rock component in its lower part, and by fine clastic strata in its upper part (Figs. 10, 16, 18). An exception is in the St. Croix River Valley, where the St. Lawrence Formation consists almost entirely of fine clastic strata. Individual shale beds in the St. Lawrence Formation as thick as several inches are common. Vertical permeability measured in plug samples of the carbonate and fine clastic components is very low to low, commonly ranging from 10 -6 to 102 md.

Secondary porosity Deep bedrock conditions—Mesoscopic fractures and dissolution cavities are common in the St. Lawrence Formation regardless of depth of burial. Examination of seven cores of the St. Lawrence Formation collected from deep bedrock conditions in Faribault, Fillmore, Freeborn, Ramsey, and Rice Counties, and video and caliper logs of open boreholes in the Twin Cities Metropolitan area revealed that secondary pores are common along discrete intervals (Figs. 3, 9, 16, 18, 20B). Dissolution cavities are as large as two inches, and typically elongate in a direction parallel to bedding. The pores exhibit features typical of hydraulic pores, including oxidation and deposits of minerals such as pyrite and calcite. Shallow bedrock conditions—Outcrops and borehole video log investigations demonstrate that in shallow bedrock conditions the St. Lawrence Formation

Figure 24. Hydraulic conductivity data for the Franconia Formation calculated from specific capacity tests. See Figure 11 for an explanation of box plots. A. Scatter plot showing the relationship between the depth of the open-hole interval below the bedrock surface and hydraulic conductivity. Shallower wells tend to have higher conductivity. B. Box plot of hydraulic conductivity values for deep bedrock conditions where the coarse clastic-dominated Mazomanie Member is greater than 25 feet thick (Anoka, Chisago, Hennepin, Isanti, Ramsey, Sherburne, Washington, and Wright Counties). C. Box plot of hydraulic conductivity values for shallow bedrock conditions where the Mazomanie Member is greater than 25 feet thick. Nineteen outlying values greater than 200 feet per day are not shown. D. Box plot of hydraulic conductivity values for deep bedrock conditions where the Mazomanie Member is less than 25 feet thick. E. Box plot of conductivity values for shallow bedrock conditions where the Mazomanie Member is less than 25 feet thick. One outlying value greater than 400 feet per day not shown (note the vertical scale change in this plot).

51

contains dissolution features as well as systematic and nonsystematic fractures (Fig. 8). At the proposed ash disposal site near Red Wing, the St. Lawrence Formation is broken by a dense network of fractures where it occurs as the uppermost bedrock (Wenck and Associates, Inc., 1997). A 200-square-foot surface exposure contained 28 subvertical fractures that ranged in length from 18 to 60 inches. A single borehole with 60 feet of open hole at the site contained two beddingplane parallel fractures large enough to be visible on a video log. Outcrops of the St. Lawrence Formation along the Mississippi River and its tributaries from Red Wing south to the Iowa border have nonsystematic fractures developed to a comparable or greater degree than those documented at the proposed ash disposal site (Fig. 26). Accessible sites where classic stress-relief fractures and dissolution features typical of the St. Lawrence Formation in shallow bedrock conditions can be examined are along Winona County Road 15 near Homer (T. 106 N., R. 6 W., sec. 16, NE, NW, NE), and in a road cut and abandoned quarry along U.S. Highway 16 in Houston County near Mound Prairie (T. 104 N., R. 5 W., sec. 34, NE, SW). Large, laterally extensive outcrops of the St. Lawrence Formation are uncommon, but one such exposure on the northwest side of Barn Bluff in Red Wing (T. 113 N., R. 14 W., sec. 20) shows that the formation contains vertical flat-sided fractures with large apertures (Fig. 8) that are similar to those known to be part of systematic regional-scale joint sets in younger strata, such as in the Platteville Formation and Galena Group.

Hydraulic attributes Deep bedrock conditions—The bulk vertical hydraulic conductivity of the St. Lawrence Formation was calculated at 10 -5 to 10 -4 foot per day in deep bedrock conditions based on pump tests and thermal profile data as part of the ATES project study (Kanivetsky, 1989). Individual shale beds may have a vertical hydraulic conductivity as low as 10-7 foot per day (Freeze and Cherry, 1979). Bulk horizontal

conductivity was not calculated as part of the ATES project, but a single packer test of a discrete interval in the lower part of the St. Lawrence Formation that contains dissolution cavities indicated a conductivity for that part of the formation of 6.7 feet per day (Fig. 10), a moderately high value analogous to the conductivity measured in the coarse clastic component in deep bedrock settings. Packer tests of two boreholes open to the St. Lawrence Formation in southwestern Wisconsin, where the formation is similar in hydrostratigraphic attributes to southeastern Minnesota, yielded values of 9.3 and 20 feet per day (Young, 1992). Hydraulic conductivity calculated from specific capacity data for wells that draw water from only the St. Lawrence Formation in deep bedrock conditions typically ranges from less than one to 50 feet per day, and averages 14 feet per day (Fig. 27). Yield to these wells must occur chiefly through bedding-plane dissolution cavities, considering the very low to low intergranular permeability of the fine clastic and carbonate rock components that compose the St. Lawrence Formation. Shallow bedrock conditions—Discrete interval and standard aquifer test data for the St. Lawrence Formation in shallow bedrock conditions are not available at this time. A limited amount of borehole data and field measurements indicate that its hydraulic properties in shallow bedrock conditions are markedly variable, and characteristic of an aquifer with low intergranular permeability in which water travels chiefly through secondary pores. Preliminary information from an ongoing site-remediation project near Blaine includes pump tests that indicate a range in specific capacity of over four orders of magnitude, with values as high as 294 gallons per minute per foot (Blaine Municipal Well Field State Superfund Site; H. Neve, unpub. data, 2001). Springs are emitted from the St. Lawrence Formation through enlarged fractures in several places in southeastern Minnesota, such as the Old House Spring in Wabasha County (T. 110 N., R. 11 W., sec. 20, NE, SE, SW), where rates of over 100 gallons per minute have been measured (Tipping and others, 2001).

Figure 25. Contour map of the potentiometric surface of the Ironton–Galesville and lower Franconia aquifers (boxed values), and of the local water-table aquifer in the fractured upper Franconia Formation strata at the proposed ash disposal site near Red Wing, Goodhue County. Note that the potentiometric surface for the lowermost few feet of the Franconia Formation and Ironton–Galesville aquifer, which are part of a regional flow system, is as much as 60 feet lower than the water table. Additionally, ground-water flow in the regional aquifer occurs in a direction that varies from that in the water-table aquifer. This indicates that the fine clastic strata in the middle to lower Franconia at this site provide confinement. See Figure 1 for location. Modified from Wenck and Associates, Inc. (1997).

52

0

78

53 710

86 0

88 0

0 D-1 D-2

76 753

E-1

754

0 84

N

755

820

C-1 C-2

766

0

water-table aquifers have similar levels

708 Measured water level where regional and

(fractured upper Franconia Formation)

766 Measured water level for water-table aquifer

(Ironton–Galesville Sandstone/lowermost Franconia Formation)

705 Measured water level for regional aquifer

0

770

920

0

5

0

79 0

94

70

88

90

797 MW-1 P-1 709

88 0

786

80

750

0

740

840

860

0

90

710

860

200 feet

720

770

P-2 MW-104

940

705

MW-105 766

Piezometric surface contour and inferred 760 flow direction for the water-table aquifer

Line of cross-section in Figure 26

Land surface contour

Monitoring well

880

860

MW-2 705

MW-102 & 202 756 705

700

697

P-3

Piezometric surface contour and inferred

780

708

0

700 flow direction for the regional aquifer

MW-103

730

820

760

0

88

92

900 800

820

840

860

880

MW-3 700

700

800

900

St. Lawrence Jordon Formation Sandstone

Franconia Formation

P-3 (426823)

ble cascading water

seeps

MW-102 (575374)

Measured water level where regional and water-able aquifers have similar levels

Carbonate component

Potentiometric surface

Measured water level for regional aquifer

Fine clastic component

Surficial deposits

Measured water level for water-table aquifer

EXPLANATION

MW-202 (578913)

Coarse clastic component

Regional aquifer potentiometric surface

Water ta

MW-105 (577642)

MW-104 (577641)

Inferred flow direction

Screened/open borehole

Dissolution features—cavities and bedding-plane fractures

Systematic fractures

Nonsystematic fractures

P-2 (426822)

MW-2 (426806)

Common stratigraphic position of springs along the lower bluffs of the Misissippi River

Figure 26. Hydrogeologic character of the Ironton–Galesville Sandstone, and Franconia and St. Lawrence Formations in shallow bedrock conditions at a proposed expansion of an ash disposal site near Red Wing in Goodhue County. A water-table aquifer occurs in the fractured, fine clastic strata of the upper Franconia Formation. The water table aquifer is separated from a regional aquifer by relatively unfractured, middle to lower Franconia Formation fine clastic strata that act as a confining unit in this setting. The cross-section is based largely on the work of Wenck and Associates, Inc. (1997), supplemented with the results of other hydrogeologic studies of the Franconia Formation in southeastern Minnesota (Figs. 20 and 23; Delta Environmental Consultants, Inc., 1992; E.C. Alexander, Jr., unpub. data, 1999). Unique numbers are listed above the boreholes. See Figure 1 for location of the ash disposal site, and Figure 25 for line of cross-section.

Ironton–Galesville Sandstone

Confining unit

Elevation in feet

54

Hydraulic conductivity calculated from specific capacity data for wells that draw water from only the St. Lawrence Formation in shallow bedrock conditions typically ranges from less than one to 75 feet per day, and averages 46 feet per day, about three times greater than the conductivity in deep bedrock conditions (Fig. 27).

Hydrogeologic synthesis In deep bedrock settings, the St. Lawrence Formation is best characterized as a unit that has a low bulk hydraulic conductivity in a vertical direction, and can therefore serve as a confining unit. Published potentiometric maps (for example Delin and Woodward, 1984) cannot be used to test the effectiveness of the St. Lawrence Formation alone to function as a confining unit in deep bedrock conditions because water levels in the upper part of the Franconia Formation are not plotted separately from those in the Ironton–Galesville aquifer. However, the relatively low vertical bulk conductivity values measured as part of the ATES project study, and differential static heads between the St. Lawrence and upper Franconia Formations at the same site suggest that the St. Lawrence Formation can provide vertical confinement. In contrast, bulk horizontal conductivity of the St. Lawrence Formation is apparently as much as four or five orders of magnitude greater than vertical conductivity, based on the moderately high values measured with discrete interval packer and specific capacity tests (Young, 1992; Miller and Delin, 1993; County Well Index database; this study). Therefore, even though the St. Lawrence Formation can serve as a confining unit in deep bedrock settings, discrete intervals with interconnected secondary pores can yield moderate to large quantities of water (Howard and Nolen-Hoeksema, 1990). The results of borehole flowmeter investigations of the St. Lawrence Formation under ambient groundwater conditions in deep bedrock settings (Figs. 20C, 28) reflect the formation's variable hydrogeologic attributes (Paillet and others, 2000; this study). Hydraulically active bedding-plane fractures can accommodate relatively strong ambient flow, yet parts of the St. Lawrence Formation have the ability to produce confinement in a vertical direction. At Winona, a substantial amount of measurable inflow to the well occurs along bedding-plane fractures or dissolution cavities in the upper and middle part of the St. Lawrence Formation (Fig. 28). Outflow of this water occurs along St. Lawrence–Franconia Formation contact strata. The lowest part of the St. Lawrence Formation at this site apparently serves as an aquitard, separating hydraulically active conduits in the upper and middle parts of the formation from the underlying Franconia

55

Formation. At Hastings, bedding-plane fractures in the lower part of the St. Lawrence Formation contributed a minimum of about a half-gallon per minute of flow that traveled up the borehole to exit along fractures in the upper part of the formation (Fig. 20C). A borehole investigation by Tipping and Runkel (unpub. data) near Hugo yielded evidence that similarly demonstrates the presence of hydraulically significant secondary pores in the St. Lawrence Formation. Substantial downhole flow emitted from the Jordan aquifer exited near the bottom of the borehole chiefly through a dissolutionenlarged bedding-plane fracture in the upper St. Lawrence Formation (Minnesota Department of Health borehole video library, unique number 645394). The St. Lawrence Formation in shallow bedrock conditions exhibits several features characteristic of younger carbonate rock layers generally considered to be karstic aquifers in southeastern Minnesota: it contains dissolution-enlarged nonsystematic fractures, cavities, and systematic fractures that may be part of a regional network. Borehole flowmeter logs collected from municipal test wells at Chaska (unique number 665713), Watertown (unique number 658174), and Greenfield (unique number 658157) as part of an ongoing project conducted as this report was completed demonstrate that bedding-plane fractures are hydraulically active in the St. Lawrence Formation in ambient conditions, and are among the principal contributors to well yield in stressed conditions of pumping or injection. High capacity springs originate from these secondary pores along the sides of bluffs. The St. Lawrence Formation differs from the three major karst systems described later in this report mostly because it occurs as the uppermost bedrock over a much smaller area of southeastern Minnesota, and it lacks a classic landsurface expression of a karstic terrain. In part this may reflect the relatively limited subcrop extent of the St. Lawrence Formation where it occurs close to the land surface, compared to the thicker karst systems that occur across broad plateaus that have only a thin cover of unconsolidated material. The St. Lawrence Formation clearly has all the attributes of a moderate to high yielding aquifer where it occurs in shallow bedrock conditions, and was recently classified as such in Rice County where it occurs as the uppermost bedrock across a large part of the county (Campion, 1997). Over 400 wells in the County Well Index database draw water from the St. Lawrence Formation in shallow bedrock conditions in southeastern Minnesota, commonly constructed in areas where the underlying Franconia Formation is dominated by the fine clastic component and therefore has poor yields.

The relative effectiveness of the St. Lawrence Formation to provide confinement in shallow bedrock conditions needs to be evaluated. It has never been demonstrated to act as a confining unit across a significant geographic extent where it occurs as the uppermost bedrock and is known to have high bulk hydraulic conductivity. An unpublished investigation conducted by the Minnesota Department of Health in Blue Earth, LeSueur, and Nicollet Counties measured differences in water chemistry above and below the St. Lawrence Formation locally where it occurs in shallow bedrock conditions, suggesting that some part(s) may provide confinement even where it is commonly used as an aquifer. The St. Lawrence Formation may therefore be analogous in hydrogeologic character to the fine-clastic dominated Franconia Formation (Fig. 23), and younger Paleozoic strata composed chiefly of the carbonate rock component. Discrete intervals with minimal development of secondary pores can at least locally provide confinement if they are laterally extensive, whereas other intervals with a greater abundance of interconnected fractures and dissolution features are of high enough hydraulic conductivity to supply moderate to large quantities of water in saturated conditions.

miles wide and have a crescent shape in cross-sections oriented perpendicular to strike. The open-hole interval interpreted as the "Jordan aquifer" on many water-well drilling records in the County Well Index database includes in its uppermost part the Coon Valley Member of the Oneota Dolomite, a unit composed of an interbedded mixture of the fine clastic, coarse clastic, and carbonate rock components. The bulk of the Coon Valley Member is inferred to have a relatively low to moderate permeability based on a small number of plug tests and water chemistry data that suggest it serves as an aquitard (Setterholm and others, 1991; Runkel, 1996b). There is a large variation from place to place in the relative proportions of the Coon Valley Member, fine clastic component, and coarse clastic component across the Jordan aquifer in southeastern Minnesota. For example, the Jordan aquifer is 80 to 100 feet thick and is internally composed of 50 to 70 feet of the coarse clastic component, and 20 to 40 feet of the fine clastic component in some parts of the city of Rochester (Fig. 30; Runkel, 1996b). The Coon Valley Member is 20 to 30 feet thick in these areas. Elsewhere in Rochester, the Jordan aquifer consists entirely of the fine clastic component and Coon Valley Member.

Secondary porosity

JORDAN SANDSTONE Hydrostratigraphic attributes Matrix porosity The Jordan Sandstone is composed of coarse clastic and fine clastic components. Plug samples of the coarse clastic component commonly have permeabilities of greater than 1,000 md (Figs. 16, 18; MUGSP, 1980; Setterholm and others, 1991). The fine clastic component is moderately to tightly cemented, very finegrained sandstone with minor siltstone and shale. Plug samples have a horizontal permeability that typically ranges from 10-5 to 10 -1 md and vertical permeability of 10-5 to 10-3 md. Cross-sections based on correlated natural gamma logs and cuttings samples (Fig. 29) show that the Jordan Sandstone internally consists of varying proportions of coarse clastic to fine clastic material (Runkel, 1996b; Runkel and others, 1999). The lower 5 to 50 feet of Jordan Sandstone is typically composed of the fine clastic component, and the upper 50 to 80 feet typically consists of the coarse clastic component. Additionally, the two components are intercalated within the Jordan Sandstone in a stratigraphically complex manner. Tongues of fine clastic component up to 40 feet thick rise diagonally up-section from the base of the Jordan Sandstone. Individual tongues are as much as several

56

Deep bedrock conditions—An unproven assumption by previous investigators is that porosity and permeability in the Jordan Sandstone is determined chiefly by intergranular attributes. However, limited core, video, and caliper log data suggest that secondary pores may be more common than widely believed. Cores and borehole video logs of the Jordan Sandstone in deep bedrock conditions indicate that cavities are rarely present in the fine clastic component of the Jordan Sandstone, developed locally where carbonate-rich intraclasts and fossils have been dissolved (Fig. 18). Mesoscopic fractures occur in the well-cemented fine clastic beds, but are open less than 1 millimeter. Caliper and video logs have also revealed bedding-plane fractures in the Jordan Sandstone where it occurs about 270 feet below the bedrock surface (Minnesota Department of Health borehole video library, unique number 658966). Borehole video logs from three sites in the Twin Cities Metropolitan area also have revealed the presence in deep bedrock conditions of vertical, systematic fractures with apertures of several inches in the upper part of the Jordan Sandstone (Minnesota Department of Health borehole video library, unique numbers 200519, 206169, 205821). Fracture apertures at these sites were apparently widened when the borehole was

A.

Conductivity in feet per day

500

Figure 27. Hydraulic conductivity data for the St. Lawrence Formation calculated from specific capacity tests. See Figure 11 for an explanation of box plots. A. Scatter plot showing the relationship between the depth of the open-hole interval below the bedrock surface and hydraulic conductivity. Shallower wells tend to have higher conductivity. B. Box plot of hydraulic conductivity values for deep bedrock conditions. C. Box plot of hydraulic conductivity values for shallow bedrock conditions. One outlying value of greater than 500 feet per day is not shown.

400

300

200

100

0 0

100

200

300

400

500

600

Distance in feet between the bedrock surface and the open-hole top

C.

DEEP BEDROCK CONDITIONS

SHALLOW BEDROCK CONDITIONS

500

500

400

400

300

300

Range

Range

B.

200

200

100

100

0

0 25 samples Average 14 feet per day

109 samples Average 46.0 feet per day

blasted with dynamite and bailed in an attempt to increase productivity. It appears the Jordan Sandstone at these sites has planes of weakness that were enlarged by these well development procedures. A tentative interpretation based on these observations is that systematic fractures are present in the Jordan Sandstone (and perhaps other units as well) in deep bedrock settings, but individual fractures may have relatively narrow apertures or are closed, and perhaps poorly connected at a large scale, compared to their character in shallow bedrock settings. Shallow bedrock conditions—Fractures are common in the Jordan Sandstone in outcrop, particularly in the uppermost 20 feet of the formation where it is locally well-cemented by calcite. Fractures range from small, irregular, closed mesoscopic fractures, to vertical flat

57

joints with inch-scale apertures that extend vertically across outcrops several tens of feet in height (Fig. 8). Fractures in the Jordan Sandstone have not been rigorously studied, but cursory observations of individual outcrops indicate that vertical fractures are typically more widely spaced than those in carbonate strata above and below the Jordan Sandstone.

Hydraulic attributes Deep bedrock conditions—High quality, discreteinterval tests have not been conducted on the Jordan Sandstone in southeastern Minnesota. Tests of three boreholes in southwestern Wisconsin yielded hydraulic conductivity values ranging from 8.0 to 24 feet per day. Standard aquifer tests of 26 wells across southeastern Minnesota indicated that hydraulic conductivity in the Jordan Sandstone ranges from about 0.1 to 100 feet

Depth in feet below the land surface and bedrock surface (italics)

Matrix hydrostratigraphic component

per day and averages 48.5 feet per day. These Minnesota values are based on aquifer tests conducted by many different scientists who presumably employed a number of different pumping procedures and methods of analysis. The lowest values of less than 10 feet per day are based on tests of the Jordan Sandstone in the

300

300 350

10

20

-2

-1

0

1

Oneota Dolomite

0

Jordan Sandstone

250

200

static water level

Interpretation

St. Lawrence Formation

200

100

Stationary flow— ambient conditions (gallons per minute)

Caliper (inches)

Gamma log (counts per second) 0

200

Rochester area where the aquifer is known to have a lenticular distribution. The fine clastic component of the Jordan Sandstone is inferred to have a horizontal hydraulic conductivity of about 10-3 to 10-1 foot per day and a vertical hydraulic conductivity between 10-5 and 10-3 foot per day based on tests of similar facies in the

400 450

Franconia Formation

400

500

500 550 Ironton–Galesville Sandstone 600 Coarse clastic component

Fine clastic component

Carbonate component

Figure 28. Results of a borehole geophysical investigation in deep bedrock conditions near Winona by Paillet and others (2000). The flowmeter data were obtained under ambient conditions. Note that ambient inflow to the well is from interbedded carbonate and fine clastic rock in the upper part of the St. Lawrence Formation, and outflow occurs in the upper part of the Franconia Formation. These hydraulically active intervals are separated by a confining unit that corresponds approximately to the St. Lawrence–Franconia Formation contact. Unique number 235704. See Figure 1 for location, and Figure 5 for an explanation of flowmeter logs.

58

Franconia and Eau Claire Formations (Miller and Delin, 1993). Hydraulic conductivity values calculated from specific capacity tests of Jordan Sandstone wells are graphically depicted in Figure 31. Under deep bedrock conditions, conductivity typically ranges from less than one to as much as 35 feet per day, with an average value of 17.4 feet per day. These values fall within the range obtained by standard aquifer pump tests, but are much lower in average value, which may reflect the ability of standard tests to more fully measure contribution of water from secondary pore networks. The relatively large range in productivity and hydraulic conductivity values for the Jordan Sandstone in deep bedrock conditions (Fig. 31) in part reflects considerable variability in the thickness of the coarse clastic component, even at the scale of an individual municipality (Runkel, 1996b; Runkel and others, 1999). The coarse clastic component has permeability values orders of magnitude higher than the fine clastic component. Therefore, under deep bedrock conditions, variations from place to place in coarse clastic component thickness have a measurable impact on the productivity and transmissivity of wells that draw water from the Jordan aquifer (Runkel, 1996b; Runkel and others, 1999). Additionally, the large number of high outlying values may be representative of wells that intersect deep secondary pores such as bedding-plane fractures or networks of systematic fractures. Shallow bedrock conditions—Hydraulic properties of the Jordan Sandstone in shallow bedrock settings are much more variable than in deep bedrock settings (Runkel and others, 1999). For example, two municipal wells open to the Jordan Sandstone in Bloomington have hydraulic conductivities of 56 and 533 feet per day based on standard aquifer pump tests, an order of magnitude disparity that is characteristic of aquifers that have some component of fracture flow (Runkel and others, 1999). Springs with flow rates of hundreds of gallons per minute are emitted from individual fractures in the Jordan Sandstone (such as along U.S. Highway 76 in Houston County, T. 103 N., R. 6 W., sec. 27). The presence of calcite cement along the walls of systematic fractures in many outcrops of Jordan Sandstone in southeastern Minnesota, and its absence in adjacent, nonfractured strata, indicates that fractures were once common preferential pathways for paleoground-water flow (McBride and others, 1994). Hydraulic conductivity values calculated from specific capacity tests for Jordan Sandstone typically range from less than one to as much as 95 feet per day, with an average value of 43.2 feet per day (Fig. 31). The greater range and overall higher average hydraulic

59

conductivity compared to the Jordan Sandstone in deep bedrock conditions reflects the greater importance of fracture flow.

Hydrogeologic synthesis In deep bedrock conditions, the lithostratigraphic unit known as the Jordan Sandstone contains components best defined as an aquifer as well as components best defined as a confining unit. The Jordan aquifer should refer to the coarse clastic component that commonly composes 20 to as much as 80 feet of the Jordan Sandstone. Defined in such a manner it is more analogous in definition and properties to the high porosity intergranular Ironton–Galesville and St. Peter aquifers. The fine clastic component of the Jordan Sandstone is more similar in hydrostratigraphic properties to the upper Eau Claire, middle Franconia, and parts of the St. Lawrence Formations. Where it occurs in the lowermost part of the Jordan Sandstone it is best considered a confining unit together with the underlying St. Lawrence Formation. Where it occurs as part of the Coon Valley Member on top of the Jordan Sandstone it is part of the overlying Oneota confining unit. Even though the Jordan Sandstone is often described as an intergranular aquifer, flow along fractures may actually be volumetrically dominant in certain settings. Flow along fractures should be expected in shallow bedrock conditions and may occur at least locally in deep bedrock conditions. The known presence of fractures in deep bedrock conditions, and the locally high hydraulic conductivity of the Jordan aquifer compared to conductivity measured by discrete interval packer tests of similar material in the Ironton– Galesville aquifer (Miller and Delin, 1993), suggest that significant yield to some wells may occur through fractures.

PRAIRIE DU CHIEN GROUP Hydrostratigraphic attributes Matrix porosity The Prairie du Chien Group consists of two formations: the Oneota Dolomite and the overlying Shakopee Formation (Fig. 2; Plates 1, 2). Both formations consist largely of carbonate rock that has a low matrix porosity of less than 10 percent, and very low to low vertical permeability that ranges from 10-4 to 10-1 md (Figs. 16, 18; MUGSP, 1980; Setterholm and others, 1991). Horizontal permeability is typically as much as ten times greater than vertical permeability. Fine and coarse clastic interbeds are common in the lowest part of the Oneota Dolomite, the Coon Valley

Member, and as a subordinate component throughout the stratigraphic extent of the Shakopee Formation. Individual siliciclastic beds are typically less than 3 feet thick except in the lower part of the Shakopee Formation, where the coarse clastic component locally is as thick as 40 feet and commonly referred to as the "Root Valley" or "New Richmond" Sandstone. The New Richmond Sandstone has been mapped across much of southeastern Minnesota outside of the Twin Cities Metropolitan area (Plates 1, 2; Squillace, 1979) and is the only regionally distributed, substantially thick interval within the Prairie du Chien Group that has a high intergranular permeability.

Secondary porosity Deep bedrock conditions—The Prairie du Chien Group is similar to the St. Lawrence Formation in that macroscopic secondary pores are common along specific intervals of strata even where buried by several hundred feet of overlying bedrock in Minnesota, and by thousands of feet of bedrock in Iowa (Figs. 9, 32; Des Moines Water Works, 1995). The Shakopee Formation and uppermost Oneota Dolomite have especially well developed secondary porosity, more than that known for any other Paleozoic unit in deep bedrock conditions of southeastern Minnesota. Typical features include dissolution-enlarged horizontal and vertical fractures up to a centimeter in width, decimeter-scale cavities, and oomoldic pores. Macroscopic secondary pores in the Oneota Dolomite are much less abundant overall, and appear to be restricted to relatively few discrete horizons compared to their more ubiquitous distribution in the Shakopee Formation. A high density of large secondary pores is common in a stromatolitic facies that is as much as 70 feet thick in the uppermost part of the Oneota Dolomite, and along individual beds as much as a few feet thick in the middle to lower parts of the formation (Fig. 32; Plates 1, 2). These high porosity intervals in the Oneota Dolomite are separated from one another by strata as much as several tens of feet thick that have relatively few, small secondary pores. Shallow bedrock conditions—In shallow bedrock conditions, the Prairie du Chien Group is ubiquitously fractured, and secondary pores are more extensively developed compared to its character in deep bedrock conditions (Fig. 32; Plates 1, 2). All outcrops of the Prairie du Chien Group contain nonsystematic, stressrelief fractures along bedding planes and at high angles to bedding. Planar, vertical fractures with inch-scale apertures are also known in the Prairie du Chien Group and may be part of regional-scale orthogonal systems (Ruhl, 1995; Runkel, 1996a). Dissolution features in the Prairie du Chien Group in shallow bedrock

60

conditions appear to be most pervasive in the stratigraphic intervals that contain the greatest density of secondary pores in deep bedrock conditions; such as within many of the carbonate beds in the Shakopee Formation, in the uppermost part of the Oneota Dolomite, and in the middle to lower Oneota Dolomite (Figs. 4, 8) along discrete horizons a few feet thick. The Prairie du Chien Group hosts dozens of caves. Most of these "megascopic pores" are three-dimensional phreatic maze caves and are developed through the lower Shakopee Formation, New Richmond Sandstone, paleokarstic Shakopee Formation–Oneota Dolomite contact strata, and in the top portion of the Oneota Dolomite. The largest Prairie du Chien Group cave in the region is Crystal Cave, a commercial cave located about thirty miles east of the St. Croix River near Spring Valley, Wisconsin. Crystal Cave extends vertically through the lower Shakopee Formation and New Richmond Sandstone and is about one mile long. At least two other extensive Prairie du Chien Group caves are known in Wabasha County, Minnesota (Tipping and others, 2001). Several smaller Prairie du Chien Group caves occur as abandoned phreatic tubes near the top of Oneota Dolomite cliffs from Hastings south to Houston County along the Mississippi River.

Hydraulic attributes Hydraulic conductivity in the Prairie du Chien Group reflects the relative development of secondary pores because the unit is composed chiefly of carbonate rock with very low intergranular permeability. As a relatively thick body of strata with variability in the size, abundance, and interconnectivity of fractures and dissolution cavities, the Prairie du Chien Group can be expected to have a range in conductivity of nine or more orders of magnitude, in part depending on the scale at which it is measured (Liesch, 1973; Graese and others, 1988; Gianniny and others, 1996; Hoffman and Alexander, 1998). The hydraulic data discussed below are biased to some extent because most hydraulic conductivity values are calculated from tests of boreholes constructed so that relatively high porosity water-producing intervals are exposed across the openhole interval, or are based on field investigations that measure rates of flow along well-developed conduit systems. Conductivity values of greater than 1,000 feet per day and flow speeds measured in miles per day in the Prairie du Chien Group are well documented through these kinds of investigations (Wheeler, 1993; Paillet and others, 2000). However, the hydraulic properties of relatively tight carbonate rock, which can be tens of feet thick in parts of the Prairie du Chien Group, are not well represented in our database because they have rarely been tested. Studies of similarly dense

61

nce awre St. L rmation Fo

an Jord one st d n Sa

235375 ota A e n O mite Dolo

M

C

A

B

110485

G

B

E F

C

100 API units

Gamma logs

C

G

235565

I

H

D

E

J

236122

I

F

K

206798

D

416091

L

50

30

0 10

0

feet

M

239765

1 mile

N Fine clastic component

Coarse clastic component

HENNEPIN

M

A

H 122210

E

B

I

G H

F

C

J

D

J

L

DAKOTA

K

RAMSEY

ANOKA

I

K 225881

G 135503

200164

200154

L

200496

Figure 29. Geometry of fine clastic tongues within the Jordan Sandstone in the central part of the Twin Cities Metropolitan area. Similar tongues are common across southeastern Minnesota. Gamma log signatures used to construct the cross-section are included. Modified from Runkel and others (1999).

WASHINGTON

62

South

S

F

C

V

434041 (cuttings only)

Interval typically interpreted as Jordan Sandstone on drilling records

1 mile

10 feet

S

S

F

V

220666

S

F

V

S

S

St. Lawrence Formation

Coon Valley Member (V): interbedded coarse clastic, fine clastic, and carbonate rock

Carbonate component

Fine clastic component (F)

F

C

F

V

F

C

F

C

V

223266 (cuttings only)

St. Lawrence Formation

North

S

F

C

F

V

420668

Figure 30. South to north cross-section through the city of Rochester, Olmsted County, showing the distribution of matrix hydrostratigraphic components in the interval generally considered the Jordan Sandstone by water-well drillers and geologists. Note that the Jordan Sandstone consists of variable proportions of fine clastic, coarse clastic, and carbonate rock components. The cross-section was constructed on the basis of interpretations of gamma logs (shown) and cuttings samples (unique numbers are listed above the logs). See Figure 1 for location. Modified from Runkel (1996b).

222525

F C F

V

220664 220831

Coarse clastic component (C)

St. Lawrence Formation

S

F

C

V

161425

increasing counts

carbonate rock have demonstrated hydraulic conductivity values that typically range from 10 -6 to 10-4 foot per day (for example Liesch, 1973; Libra and Hallberg, 1985; Graese and others, 1988; Gianniny and others, 1996). Deep bedrock conditions—Hydraulic properties obtained from pump tests (Fig. 13), dye-trace and borehole flowmeter investigations (Fig. 33) of the Prairie du Chien Group in deep bedrock settings demonstrate that flow occurs along discrete intervals, many with very high conductivity, that are preferentially located in the Shakopee Formation and uppermost Oneota Dolomite strata. A dye-trace study in Fillmore County documented flow speeds of greater than 6.5 miles per day, largely along Shakopee Formation–Oneota Dolomite contact conduits that locally occurred in deep bedrock conditions along the trace. At Faribault, two relatively narrow horizons with a high concentration of cavities in Shakopee Formation–Oneota Dolomite contact strata at a depth between 175 and 200 feet below the bedrock surface had conductivities measured at 93 and 837 feet per day based on flowmeter logs analyzed in conjunction with pumping and drawdown data (Fig. 33; Paillet and others, 2000). An ongoing investigation by Tipping and Runkel (unpub. data) also documents similar discrete intervals in the Shakopee Formation and uppermost Oneota Dolomite with high conductivity at a number of other sites in southeastern Minnesota. An approximately 150-foot-thick interval of upper Prairie du Chien Group rock under deep bedrock conditions at Spring Grove has a bulk hydraulic conductivity of 9 feet per day (Eder and Associates, 1997). Hydraulic conductivity of the middle and lower parts of the Oneota Dolomite is much lower than that of its upper part. Hydraulic models based on pump and slug tests in the New Brighton and Arden Hills areas indicated a relatively low vertical conductivity of 1.5 x 10-4 foot per day and horizontal conductivity of 7.5 x 10-3 foot per day at this site, where lower to middle Oneota Dolomite strata occur at a depth that is transitional between shallow and deep bedrock conditions (Camp, Dresser and McKee, 1991). Strata in the overlying Shakopee Formation are five orders of magnitude higher in conductivity at the same site. A cooperative study by the Minnesota Department of Natural Resources and Minnesota Department of Health indicated a similarly low vertical conductivity of 10-4 foot per day at Plainview (Fig. 13). Hydraulic conductivity values based on specific capacity tests of wells that draw water from the Prairie du Chien Group in deep bedrock conditions typically range from less than one to as high as 50 feet per day,

63

and average 33.53 feet per day (Fig. 34). The large number of high outlying values and overall moderate to high average conductivity compared to other parts of the Paleozoic section is consistent with borehole and core observations that large secondary pores are common along specific intervals in the Prairie du Chien Group in deep bedrock conditions. Runkel (2000) used specific capacity values in nine southeastern Minnesota counties to demonstrate that relative productivity of water wells is at least in part dependent upon the stratigraphic position of the open hole: wells that draw water from the lowermost Shakopee Formation and uppermost Oneota Dolomite contact strata were three times more productive than those open only to the upper Shakopee Formation or middle to lower part of the Oneota Dolomite. The substantially higher average productivity of wells open across the contact between these formations apparently reflects the relatively great abundance of secondary pores and perhaps the local presence of the New Richmond (Root Valley) Sandstone. Shallow bedrock conditions—A wide range in hydraulic conductivity has been measured in the Prairie du Chien Group where it occurs in shallow bedrock conditions, with most measurements collected as part of site-remediation projects (Fig. 13). Pump and slug tests of the Shakopee Formation in the Arden Hills and New Brighton areas indicated a horizontal conductivity of 163 feet per day and vertical conductivity at 1.75 feet per day. Bulk values of 18 and 5.3 feet per day were calculated based on borehole flowmeter and pumping test data collected from two wells open to the uppermost Oneota Dolomite and Shakopee Formation near Faribault and Rochester, respectively (Paillet and others, 2000). The same investigation demonstrated that contribution to these wells was accommodated through a few discrete horizontal fractures and dissolution features having hydraulic conductivity values that ranged from 2.2 to 1,023 feet per day (Figs. 33, 35). Unfractured rock between major water-producing horizons has much lower hydraulic conductivity, and was demonstrated at both sites to provide hydraulic separation of conduits. Variability in hydraulic conductivity was also recorded at the Oronoco Landfill site near Rochester where discrete 10-foot intervals had conductivities that ranged from 1.6 to 65 feet per day (RMT, Inc., 1992). An ongoing investigation conducted by the Minnesota Geological Survey (Tipping and Runkel, 2001) has yielded results consistent with those of RMT, Inc. (1992) and Paillet and others (2000). Flowmeter and packer tests of scientific boreholes at Northfield and Cottage Grove (Figs. 36, 37) demonstrated that water in the Prairie du Chien Group travels chiefly along

A.

B. 1.6 x 104 Transmissivity in feet squared per day

500 400 300 200 100

1.4 x 104 1.2 x 104 1 x 104 8000 6000 4000 2000

0

0 100

0

200

300

400

500

600

700

30

C.

40

50

60

70

80

Thickness of coarse clastic component in feet

Distance in feet between the bedrock surface and the open-hole top

D.

DEEP BEDROCK CONDITIONS

SHALLOW BEDROCK CONDITIONS

250

250

200

200

150

150

Range

Range

Conductivity in feet per day

600

100

100

50

50

0

0 702 samples Average 17.4 feet per day

851 samples Average 43.2 feet per day

Figure 31. Hydraulic conductivity data for the Jordan Sandstone calculated from specific capacity tests. See Figure 11 for an explanation of box plots. A. Scatter plot showing the relationship between the depth of the open-hole interval below the bedrock surface and hydraulic conductivity. Shallower wells tend to have higher conductivity. B. Scatter plot showing positive correlation between the thickness of the coarse clastic component and transmissivity for wells in the Jordan Sandstone in the Twin Cities Metropolitan area (modified from Runkel and others, 1999). C. Box plot of hydraulic conductivity values for shallow bedrock conditions. Twenty-five outlying values greater than 250 feet per day are not shown. D. Box plot of hydraulic conductivity values for deep bedrock conditions. Two outlying values greater than 250 feet per day are not shown.

64

90

a few discrete bedding-plane parallel conduit systems separated from one another by intervals of carbonate rock of sufficiently low vertical conductivity to serve as confining units. Static head variability within the Prairie du Chien Group is strong enough locally to drive ambient flow in a borehole at rates greater than 12 gallons per minute. Dye-trace investigations have measured rapid flow speeds in the Prairie du Chien Group where it occurs in shallow bedrock conditions. Horizontal flow speeds of ground water at the Oronoco site were as rapid as 800 feet per day (Donahue and Associates, Inc., 1991). The Fillmore County study by Wheeler (1993) measured flow speeds as rapid as 6.5 miles per day within conduits that occurred largely in conditions of shallow burial beneath younger bedrock. These and other dye traces in shallow bedrock conditions of southeastern Minnesota all yield relatively wide breakthrough curves that have a fine-scale structure. These breakthrough curves are different from the narrow breakthrough curves seen in the conduits of the Galena–Spillville karst system (described later in this report) and are consistent with movement through complex anastomosing, turbulent flow systems. Hydraulic conductivity values calculated from specific capacity tests for Prairie du Chien Group wells in shallow bedrock conditions (Fig. 34) typically range from less than one to 125 feet per day and have an average value of 60.8 feet per day, about twice the average value for deep bedrock conditions. Wells constructed to draw water from Shakopee Formation– Oneota Dolomite contact strata are substantially more productive than wells that are not (Runkel, 1999), a relationship similar to that found in deep bedrock conditions.

Hydrogeologic synthesis The Prairie du Chien Group occurs as the uppermost bedrock across a wide expanse of southeastern Minnesota where it exhibits all of the usual attributes of karst, including sinkholes, springs, caves, stream sinks, and dry valleys. As such it is considered the stratigraphically lowermost of three major "karst systems." Recharge occurs through fractures and dissolution cavities, and ground water can travel along bedding-plane parallel conduits at rates that have been measured in miles per day. The differential stratigraphic distribution of secondary pores is apparently reflected in some of the karstic characteristics of the Prairie du Chien Group: relatively high densities of sinkholes and springs, corresponding to major bedding-plane parallel conduit systems, occur where the Shakopee Formation– Oneota Dolomite and St. Peter Sandstone–Shakopee Formation contact strata lie directly beneath the land

65

surface (Dalgleish and Alexander, 1984; Tipping and others, 2001). Four catastrophic failures of three wastewater treatment lagoons (Alexander and Book, 1984; Jannik and others, 1992; Alexander and others, 1993) all occurred in lagoons built on top of Shakopee Formation–Oneota Dolomite contact strata. Discrete interval pump tests, dye-trace investigations, plug tests of permeability, and borehole flowmeter studies collectively demonstrate that ground water travels in the Prairie du Chien Group chiefly through fractures and solution features having moderate to extremely high values of hydraulic conductivity. Significant flow may also occur along relatively permeable coarse clastic interbeds, although this has not been documented. Intervals of rock that lack these features are orders of magnitude lower in conductivity and serve as confining units. The distribution of these preferential flow paths in deep and shallow bedrock conditions appears to be largely stratigraphically controlled (Fig. 32) and therefore generally predictable on a regional scale. The principal intervals with particularly high porosity and conductivity include: the Shakopee Formation where individual carbonate beds with macroscopic secondary pores and thin, coarse clastic interbeds are common; uppermost Oneota Dolomite strata where large cavities are abundant; and within the coarse clastic New Richmond Sandstone. In contrast, the middle to lower Oneota Dolomite has a lower porosity, and hydraulic tests demonstrate a corresponding lower conductivity (Camp, Dresser and McKee, 1991). On the basis of this distribution of porosity and related conductivity, the Prairie du Chien Group can be generally divided into two hydrogeologic units: a Shakopee aquifer of relatively high hydraulic conductivity that roughly corresponds to the Shakopee Formation and upper one-third (as much as 70 feet) of the Oneota Dolomite; and an underlying Oneota confining unit. The most comprehensive site-specific hydrogeologic study of the Prairie du Chien Group, conducted at a landfill site near Oronoco in Olmsted County (Donahue and Associates, Inc., 1991; RMT, Inc., 1992), indicated that its hydrogeologic character is generally similar to that described for many other karstic carbonate rock aquifers studied over the past 10 years (Fig. 38A; for example Gianniny and others, 1996; Zanini and others, 1998), and in particular reflects the hydrogeologic differences between its upper and lower parts. Donahue and Associates, Inc. (1991) and RMT, Inc. (1992) used borehole videos, dye tracing, groundwater chemistry, discrete-interval packer tests, and potentiometric maps to demonstrate that the Prairie du Chien Group is divisible into two hydrogeologic units that correspond to the Shakopee aquifer and Oneota

66

A

200

100

Freeborn County 223082

300

8 inches

LOCATION OF CROSS-SECTION

A

Dissolution features— cavities and enlarged bedding-plane fractures

Carbonate component

Fine clastic component

Coarse clastic component

Jordan Sandstone

100

200

100

225652

Increasing borehole diameter

100

206169

Hennepin County 203782

300

250

200

Matrix hydrostratigraphic component Visual porosity log 0 100%

Core log

185808

Ramsey County 255736

100

225622

J o rd

100

stone

Oneota confining unit

Shakopee Formation Oneota Dolomite

Shakopee aquifer

ter Sa n d s t o ne

an San d

St. Pe

Washington County 645394

A'

Figure 32. Representative caliper and core logs showing the distribution of secondary pores in the carbonate-dominated Prairie du Chien Group across southeastern Minnesota. The depth of burial beneath the bedrock surface is listed beside each log. The sites range from deep bedrock to shallow bedrock conditions. The distribution of pores, largely dissolution cavities, is stratigraphically controlled: the Shakopee Formation and upper part of the Oneota Dolomite have a high density of large cavities. In contrast, secondary porosity in the middle to lower parts of the Oneota Dolomite is much lower. Hydraulic data discussed in this report indicate that the more porous Shakopee Formation and upper Oneota Dolomite are best classified as an aquifer, whereas the relatively tight Oneota Dolomite behaves as a confining unit. See Figure 1 for the location of counties listed for each log.

20 inches 10 Caliper log scale (except for Olmsted County)

205821

200

stone

Rice County Dakota County 658966

St. Peter Sand

Steele County 236018

Shakopee Formation Oneota Dolomite

600

500

Scale for Olmsted County

6

A'

500

658967

Core log Matrix hydrostratigraphic component Visual porosity log 0 100% Olmsted County

Winona County 603060

Shakopee aquifer

Oneota confining unit

confining unit (Fig. 38A). The Shakopee aquifer contains a well-connected network of fractures and dissolution cavities that provide recharge to horizontal conduits across which water flows as rapidly as 800 feet per day. In contrast, the Oneota confining unit lacks a well-connected system of vertical fractures and dissolution features and serves as a hydraulic barrier that separates karstic carbonate rock above, from the coarse clastic Jordan aquifer below. Considerable evidence supports the interpretation of a confined Jordan aquifer at the Oronoco landfill site (Donahue and Associates, Inc., 1991; RMT, Inc., 1992). The Jordan and Shakopee aquifers differ from one another in potentiometric head by as much as 9 feet, and in hydraulic gradients and inferred flow directions. Furthermore, pumping of the Jordan aquifer caused no measurable drawdown in water levels in the Shakopee aquifer. Lastly, the two aquifers are hydrogeochemically isolated from one another: water in the Shakopee aquifer was extensively impacted by surface contamination and water in the Jordan aquifer was not. The site-specific hydrogeologic depiction in Figure 38 can be extrapolated to a regional scale because the landfill at Oronoco lies in a plateau setting that is typical of an enormous area of southeastern Minnesota where the Prairie du Chien Group occurs as the uppermost bedrock. The results of many other site-specific, as well as regional-scale studies are consistent with the model for the Oronoco landfill. For example, Hall and others (1911) recognized a twofold hydrogeologic division of the Prairie du Chien Group that roughly corresponds to the Shakopee aquifer and Oneota confining unit as we define them in this report. Their work included documentation of the ability of the Oneota Dolomite to confine artesian water in the underlying Jordan Sandstone across a large expanse of southeastern Minnesota, and similar conditions remain today in places such as Northfield, Cannon Falls, and Preston despite heavy water withdrawals from the Jordan Sandstone in those areas. A dye-trace investigation by Wheeler (1993) indicated that the hydrogeologic model for the Oronoco landfill site is applicable to a much larger scale in nearby Fillmore County (Fig. 38B). Several other investigations conducted over the past ten years are also best explained by the presence of a confining unit in the Oneota Dolomite. This includes independent hydrologic evidence outside of that collected from the Oronoco landfill study: potentiometric data that indicate differential heads above and below the Oneota Dolomite (Kanivetsky, 1988; Tipping, 1992; Barr Engineering, 1996), pumping tests and barometric data that document

67

hydraulic confinement of the Jordan Sandstone (Camp, Dresser and McKee, 1991; Barr Engineering, 1996; Runkel and others, 1999), and by ground-water chemistry (for example Setterholm and others, 1991; Tipping, 1992; Wall and Regan, 1994). Even though these studies were conducted on the Prairie du Chien Group in different parts of Minnesota and in different conditions of burial depth and topography, they are all compatible with a hydrogeologic model in which the Shakopee Formation and upper part of the Oneota Dolomite are a karstic aquifer with moderate to high hydraulic conductivity, whereas the middle to lower Oneota Dolomite is of markedly lower conductivity and confines the Jordan aquifer. The results of recent borehole geophysical studies by Tipping and Runkel (unpub. data) at a number of sites in southeastern Minnesota (Fig. 37, for example) also demonstrate the presence of an Oneota confining unit. An evaluation of potentiometric data in Minnesota and surrounding areas does not support the common assertion of good hydraulic connection between the Prairie du Chien Group and Jordan Sandstone (for example Kanivetsky, 1978; Delin and Woodward, 1984). A study in northern Iowa by Horick (1989) provided evidence that the lower part of the Prairie du Chien Group provides effective confinement at a regional scale (Fig. 39). He showed that the potentiometric surface of water in the St. Peter aquifer is 50 to 200 feet higher than that of the Jordan Sandstone across much of northeastern Iowa along its border with Minnesota. In the absence of any known beds in the lower St. Peter aquifer that could create such confinement, Horick concluded that some part of the Prairie du Chien Group is a "confining interval." In Minnesota, differences in static levels between the upper part of the Prairie du Chien Group and Jordan aquifer have been noted in at least seven counties (Hall and others, 1911; Kanivetsky and Palen, 1982; Kanivetsky, 1984, 1988; Kanivetsky and Cleland, 1990, 1992; Donahue and Associates, Inc., 1991; RMT, Inc., 1992; Barr Engineering, 1996; Zhang and Kanivetsky, 1996). These differences in potentiometric head levels have typically been attributed to impermeable beds of "limited" extent in the Prairie du Chien Group (for example Kanivetsky and Palen, 1982; Kanivetsky, 1984, 1988; Kanivetsky and Cleland, 1990, 1992), but no stratigraphic evidence is provided to support the interpretation that such beds are genuinely only locally distributed. The results of this study support a different interpretation: that the differences in head levels noted in these many different areas reflect the ability of an interval of rock that is of regional extent across southeastern Minnesota, the Oneota Dolomite, to provide confinement. The data summarized above for the Prairie du Chien

Matrix hydrostratigraphic component

Depth in feet below the land surface and bedrock surface (italics)

Gamma log (counts per second) 0

100

200

Caliper (inches) 0

3

Stationary flow (gallons per minute)

Televiewer 6

-10

-5

Location and hydraulic conductivity of dominant permeable intervals (feet per day)

0

St. Peter Sandstone 1023

50 Casing bottom

Ambient flow 140

50

Pumped flow

100

100

150

Shakopee Formation

140

Interpretation 837 279 150

200 93

837 200

Oneota Dolomite

Coarse clastic component

Fine clastic component

Carbonate component

Figure 33. Results of a borehole geophysical investigation of the carbonate dominated strata of the Shakopee Formation and uppermost Oneota Dolomite strata at Faribault, Rice County (Paillet and others, 2000). Flowmeter data were collected under ambient and 2 gallons per minute pumping conditions. Note that water is emitted through discrete horizons, typically less than one foot thick, with hydraulic conductivity as high as 1,023 feet per day. Ambient inflow to the well occurs in the upper half of the Shakopee Formation and outflow occurs in the lower Shakopee Formation and uppermost Oneota Dolomite, demonstrating hydraulic separation of conduits within the Shakopee Formation. Unique number 625327. See Figure 1 for location, and Figure 5 for an explanation of flowmeter logs.

68

Conductivity in feet per day

A.

1200 1000 800 600 400 200 0 0

100

200

300

400

500

600

700

Figure 34. Hydraulic conductivity data for the Prairie du Chien Group calculated from specific capacity tests. See Figure 11 for an explanation of box plots. A. Scatter plot showing the relationship between the depth of the open-hole interval below the bedrock surface and hydraulic conductivity. Shallower wells tend to have higher conductivity. B. Box plot of hydraulic conductivity values for deep bedrock conditions. Two outlying values greater than 1,000 feet per day are not shown. C. Box plot of hydraulic conductivity values for shallow bedrock conditions. Twelve outlying values greater than 1,000 feet per day are not shown.

Distance in feet between the bedrock surface and the open-hole top

C.

DEEP BEDROCK CONDITIONS

SHALLOW BEDROCK CONDITIONS

1000

1000

800

800

600

600

Range

Range

B.

400

400

200

200

0

0 448 samples Average 33.5 feet per day

2,195 samples Average 60.8 feet per day

Group indicate that it should no longer be treated as a single aquifer combined with the underlying Jordan Sandstone. Instead it should be divided into a Shakopee aquifer and Oneota confining unit. The Oneota confining unit can be traced across all of southeastern Minnesota area using outcrops, cores, borehole logs, and cuttings. Although discrete beds with a relatively great density of secondary pores in the Oneota confining unit may be of high horizontal hydraulic conductivity, there is strong evidence that bulk vertical conductivity is low enough to provide confinement, analogous in that respect to the hydrogeologic attributes of the St. Lawrence Formation. As with all other confining units in southeastern Minnesota, it provides confinement where it is not breached by interconnected networks of secondary pores, a situation that most commonly

69

occurs locally in shallow bedrock conditions.

ST. PETER SANDSTONE Hydrostratigraphic attributes Matrix porosity The St. Peter Sandstone is composed chiefly of the coarse clastic component across most of southeastern Minnesota, but it contains thick (up to 30 feet), laterally extensive fine clastic beds in its lower one-half in the Twin Cities Metropolitan area, and along the western subcrop of Paleozoic rocks south of the metropolitan area (Fig. 40). Fine clastic beds in the lowermost part of the St. Peter Sandstone are known to occur locally in other counties across southeastern Minnesota, but

Matrix hydrostratigraphic component

Depth in feet below the land surface and bedrock surface (italics)

are typically only a few feet thick or less, and their lateral continuity is poorly understood. The coarse clastic component in the St. Peter Sandstone is especially well-sorted and friable. It is one of the most texturally homogeneous units of sandstone known. Laboratory analyses of plug samples indicated it has

Gamma log (counts per second) 0 50 100

2

Caliper (inches) 5

a high porosity and permeability (Norvitch and others, 1973). Interbeds of the fine clastic component have not been tested for permeability, but the attributes observable in outcrop and core samples suggest that the fine clastic rocks are similar to beds of older

Televiewer 12

Location and hydraulic conductivity of dominant Stationary flow permeable intervals (gallons per minute) (feet per day) -2 0 2 4

No data

Quaternary

50

Casing bottom

Ambient flow

440 100

50

150

Pumped flow

Shakopee Formation

Interpretation

165

28

100

200 11 Coarse clastic component

Fine clastic component

Carbonate component

Figure 35. Results of a borehole geophysical investigation of the carbonate-dominated strata of the Shakopee Formation at Rochester, Olmsted County (Paillet and others, 2000). Flowmeter data were collected under ambient and 2 gallons per minute pumping conditions. Note that water is emitted through discrete horizons, typically less than one foot thick, with hydraulic conductivity as high as 440 feet per day. The ambient and pumped flowmeter logs indicate that conduits are hydraulically separated: there is a small head difference driving inflow up the borehole from the conduit that lies 50 feet beneath the bedrock surface. Relatively minor inflow in the two conduits below is not markedly affected by pumping because their inflow is driven up the borehole by a head difference that is larger than the drawdown produced by the pumping. Unique number 485610. See Figure 1 for location, and Figure 5 for an explanation of flowmeter logs.

70

Paleozoic strata that have a vertical permeability from 10 -7 to 10 -5 md.

Secondary porosity Deep bedrock conditions—The conventional view is that interconnected networks of open vertical fractures are poorly developed in a friable, high porosity unit such as the St. Peter Sandstone where it is covered by younger bedrock layers of contrasting material properties. An unproven assumption by previous investigators is that fractures and dissolution features are rare to absent in the St. Peter Sandstone in deep bedrock settings and therefore porosity and permeability is determined chiefly by intergranular attributes. Shallow bedrock conditions—The St. Peter Sandstone in shallow bedrock settings contains systematic fractures with apertures as wide as a few inches. They are similar to fractures developed in younger carbonate strata, except they are more widely spaced, are not solution enlarged, and are commonly filled with friable sand. Easily accessible examples of such fractures occur in roadcut exposures in the Rochester area (such as a roadcut in the southwest corner of T. 107 N., R. 14 W., sec. 32), and in bluffside exposures along the Mississippi River in downtown St. Paul. Norvitch and Walton (1979) also noted the presence of fractures in shallow excavations of the St. Peter Sandstone beneath the overlying Glenwood and Platteville Formations. Large voids routinely develop in the St. Peter Sandstone. Caves up to hundreds of feet long are known in the sandstone from the Twin Cities Metropolitan area south to the Iowa border. The voids are believed to be the result of either mechanical erosion of the sandstone by turbulent water flow or of collapse into solutional voids in the underlying Shakopee Formation carbonates. These voids are sufficiently common in the Rochester area to have affected the foundation engineering of several buildings.

Hydraulic attributes Deep bedrock conditions—Hydraulic conductivity of the St. Peter Sandstone in deep bedrock in Illinois and southeastern Wisconsin falls within a relatively narrow range of values from 1.3 to 8.4 feet per day (Nicholas and others, 1987; Graese and others, 1988; Young, 1992). The St. Peter Sandstone in those areas is similar in texture to the formation in Minnesota, but may have a greater degree of cementation (Hoholick and others, 1984). Hydraulic conductivity values calculated from specific capacity data (Fig. 41) collected from wells in deep bedrock conditions of southeastern Minnesota average 15.9 feet per day, with a typical range of about 2 to 50 feet per day. Fine clastic beds

71

in the lower St. Peter Sandstone have not been tested for hydraulic conductivity, but leakage rates calculated from pump tests and mathematical models suggest an overall low vertical conductivity of 10-3 foot per day (Schoenberg, 1990). Shallow bedrock conditions—Standard aquifer tests of several boreholes in the Twin Cities Metropolitan area and of a single well in Rice County yielded conductivity values that ranged from 20 to 30 feet per day for the St. Peter Sandstone in shallow bedrock conditions (Barr Engineering, 1976, 1986; Madsen and Norvitch, 1979). Conductivity values based on specific capacity tests of wells in shallow bedrock conditions (Fig. 41) typically range from less than 1 to 75 feet per day, with an average value of 38.7 feet per day. The wide range and overall higher average conductivity compared to the St. Peter Sandstone in deep bedrock conditions reflects the greater significance of contribution from fracture flow.

Hydrogeologic synthesis The common depiction of the upper part of the St. Peter Sandstone as a relatively homogeneous, intergranular aquifer of moderate to high hydraulic conductivity is possibly an adequate model for deep bedrock conditions. County-scale potentiometric maps demonstrate that fine clastic strata in the lower part of the formation serve as confining units that hydraulically separate the St. Peter aquifer from the underlying Shakopee aquifer (for example Kanivetsky and Cleland, 1992) in the Twin Cities Metropolitan area. The presence of similar fine clastic beds in the lower St. Peter Sandstone outside of the metropolitan area indicates that the Shakopee aquifer may be hydraulically separated from the St. Peter aquifer on a larger scale across southeastern Minnesota, particularly along the western subcrop extent of Paleozoic strata (Fig. 40). To the east, fine clastic strata are much thinner, and their ability to serve as a confining unit is not known. Lindgren (2001) suggested that the lower St. Peter Sandstone in the Rochester area locally provides confinement where the fine clastic beds are at most only a few feet thick. The St. Peter and Shakopee aquifers have historically been inferred to be hydraulically well connected where these fine clastic interbeds are absent, and local hydrogeologic studies demonstrate some connectivity (Delin, 1991). The hydrogeologic significance of fractures and large voids in shallow bedrock conditions has not been investigated in systematic fashion. The large apertures and abundance of fractures in outcrop, and the relatively high hydraulic conductivity of the St. Peter aquifer in shallow bedrock conditions across southeastern Minnesota compared to deep conditions of burial

100 100

150

150

8

12 16

20

0

1

2

3

765

770

775

Trolling flow— Stationary flow— ambient conditions ambient conditions (gallons per minute) (gallons per minute) -10

-5

0

-15

-10

static water level

Interpretation

Jordan Sandstone

200

200

-5

casing bottom

Prairie du Chien Group Shakopee Formation

50

4

Static water elevation (pump off)

Oneota Dolomite

50

100

Pump rate (gallons per minute)

Hole diameter (inches)

Quaternary

0

0

Major cavities (based on video and caliper logs)

Matrix hydrostratigraphic component

Depth in feet below the land surface and bedrock surface (italics)

Gamma log (API units)

Packer tests

Packer status: top and bottom closed

Flow measurement taken

Top open Bottom open

Coarse clastic component Fine clastic component Carbonate component

Figure 36. Results of a borehole geophysical investigation of the carbonate-dominated strata of the Prairie du Chien Group and underlying coarse clastic strata of the Jordan Sandstone at Hamlet Park in Cottage Grove, Washington County. Flowmeter data were collected under ambient ground-water conditions with the tool moving up-hole at a rate of 10 feet per minute ("trolling" log), as well as with the tool at stationary positions. Note that water enters the borehole chiefly through large secondary pores in the upper part of the Shakopee Formation. That water travels down the hole at a rate measured at nearly 15 gallons per minute, and exits the borehole through large cavities at a discrete horizon in the lower part of the Shakopee Formation. Such a borehole flow pattern, in conjunction with information from the caliper log, static water levels, temperature profile, and chloride concentrations, all indicate that flow within the Shakopee aquifer occurs along discrete, bedding-plane parallel conduits separated from one another by confining units. Similar borehole tests in other parts of southeastern Minnesota (Tipping and Runkel, unpub. data) are consistent with such an interpretation. Measurable downflow of lesser magnitude occurs between the lower Shakopee Formation and Jordan Sandstone. Such flow is driven by differential heads between the Shakopee and Jordan aquifers. Static water levels measured at discrete intervals indicate that the Oneota Dolomite serves as a confining unit separating the two aquifers. Unique number 658965. See Figure 1 for location, and Figure 5 for an explanation of flowmeter logs.

72

0

50

100

250 250

200

200

150

150

100

static water level

casing bottom

100

Carbonate component

Coarse clastic component

0 5 6 7 8 9 10

Hole diameter (inches) 0 1

2 925 7

pH

8 0

Figure 37

Temperature

9

Temperature (C) and pH 5 10 15

Chloride (mg/l)

Explanation to Figure 37 is on page 74.

Bottom open

Top open

920

Static water Pump rate (gal/min) elevation (pump off)

Below pump limit

Gamma log (API units)

Packer tests

Flow measurement taken

Depth in feet below the land surface and bedrock surface (italics)

50

Quaternary

Prairie du Chien Group

0

Shakopee Formation

Oneota Dolomite

Jordan Sandstone

73

20 0 5

static water level

casing bottom

10

Trolling flow—ambient (gallons per minute) 15

0

10 15

Pumping

Ambient

Interpretation

5

Stationary flow (gallons per minute)

51%

14%

34%

Location and percent of transmissivity of dominant permeable intervals

Packer status: top and bottom closed

Major cavities (based on video and caliper logs)

Matrix hydrostratigraphic component

suggest that fracture flow may be regionally significant. Fractures in the St. Peter Sandstone are known to at least locally provide conduits through which ground water can move much more rapidly than rates predicted under the assumption of intergranular flow only. Preferential flow through fractures was noted in shallow excavations of the St. Peter Sandstone in the Twin Cities Metropolitan area (Norvitch and Walton, 1979), and small streams in Fillmore County and surrounding areas are known to completely empty into fractures in the St. Peter Sandstone. A good example of such a feature is a stream sink near the headwaters of an intermittent tributary to Watson Creek about 0.25 mile south of Fillmore County Highway 8 (T. 103 N., R. 10 W., sec. 18, BCCAC).

GLENWOOD FORMATION Hydrostratigraphic attributes

Matrix porosity The Glenwood Formation is composed chiefly of the fine clastic component, mostly shale and siltstone. Plug samples from a core in Faribault County had vertical permeability of 10-5 to 10-4 md (MUGSP, 1980). Friable coarse clastic sandstone beds that are a minor component of the Glenwood Formation have not been tested, but similar sandstone in other parts of the Paleozoic section is of high porosity and permeability.

Secondary porosity Deep bedrock conditions—As a relatively ductile, high porosity unit in a layered sequence of bedrock strata, open fractures in deep bedrock settings are probably uncommon in the Glenwood Formation. None were visible in the single core examined as part of our investigation (Fig. 42). Shallow bedrock conditions—Open fractures are common in the Glenwood Formation in shallow bedrock

Figure 37. Figure appears on page 73. Results of a borehole geophysical investigation of the carbonate-dominated strata of the Prairie du Chien Group and underlying coarse clastic strata of the Jordan Sandstone at Carleton College in Northfield, Rice County. Flowmeter data were collected under ambient ground-water conditions with the tool moving up-hole at a rate of 10 feet per minute ("trolling" log), as well as with the tool at stationary positions. Note that water enters the borehole chiefly through the coarse clastic strata of the Jordan Sandstone near the bottom of the hole. That water travels up the hole, past the entire Oneota Dolomite with negligible loss, at rates greater than 12 gallons per minute. This upflow exits the borehole through large dissolution cavities at a discrete horizon that approximates the Shakopee Formation–Oneota Dolomite contact. Such a borehole flow pattern, in conjunction with information from the caliper log, static water levels, temperature profile, and chloride concentrations, all strongly indicate that the lower part of the Oneota Dolomite confines the Jordan Sandstone at this site. Similar borehole tests in other parts of southeastern Minnesota (Tipping and Runkel, unpub. data) are consistent with such an interpretation. A borehole video log of this well reveals that the Jordan aquifer contains bedding-plane and narrow, subvertical fractures. The gradual change in rate of flow across the Jordan aquifer on the flowmeter logs apparently records inflow through intergranular pore spaces, whereas the abrupt shifts in rate records contribution through fractures. Unique number 658966. See Figure 1 for location, and Figure 5 for an explanation of flowmeter logs.

Figure 38. Hydrogeologic character of the Prairie du Chien Group and Jordan Sandstone in shallow to deep bedrock conditions at a landfill near Oronoco (Olmsted County), and in eastern Fillmore County. See Figure 1 for location. A. At Oronoco, a water-table aquifer lies in the Shakopee Formation and uppermost Oneota Dolomite, which have abundant fractures and dissolution features. The Jordan Sandstone at this site is an intergranular aquifer with a regional groundwater system. The two aquifers are hydraulically separated by Oneota Dolomite with relatively few secondary pores. This depiction is based on borehole videos, gamma and caliper logs, cuttings, dye tracing, water chemistry, and potentiometric levels. Modified from Donahue and Associates, Inc. (1991). B. Conduit flow in the Prairie du Chien Group in eastern Fillmore County based on dye-trace investigations. Note the similarity of hydrogeologic conditions to those at Oronoco. Modified from Alexander and Lively (1995).

74

75

600

700

800

900

1000

1100

1200

100 feet

200 feet

Oneota Dolomite

Decorah Shale Platteville and Glenwood Formations St. Peter Sandstone Shakopee Formation– Willow River Member Shakopee Formation– New Richmond Member

Galena Group– Cummingsville Formation

B.

A.

Elevation in feet

0

C

0

0.5 mile

500

DYE INTRODUCTION

seeps

surface stream

C

Carbonate

Fine clastic

Coarse clastic

DYE INTRODUCTION

stream sink

Dissolution features—cavities and enlarged bedding-plane fractures

Systematic fractures (some dissolution enlarged)

Non-systematic fractures (some dissolution enlarged)

SECONDARY POROSITY

EXPLANATION

LOCATION OF CROSS-SECTION

C'

FILLMORE COUNTY

MATRIX HYDROSTRATIGRAPHIC COMPONENTS

1000 feet

DYE PLUME

Dye plume flow direction

Path of surface and subsurface water flow

Springs

Dye plume

Surficial deposits

seeps

C'

Duschee Creek

Lanesboro Fish Hatchery

Jordan Sandstone

Oneota Dolomite

Shakopee Formation

conditions, and some fractures extend vertically across the entire formation (outcrop observations for this study). Their abundance, interconnectivity, and presence in the subsurface have not been studied.

Hydraulic attributes The Glenwood Formation has not been subjected to discrete interval packer testing, but based on its high shale content it can be expected to have a vertical hydraulic conductivity of about 10-7 to 10-5 foot per day (Freeze and Cherry, 1979) where secondary pores are absent, and variable conductivity where it contains open fractures such as those known to occur in outcrop.

Hydrogeologic synthesis The Glenwood Formation is a low hydraulic conductivity unit that is known to function as a confining bed in shallow as well as deep bedrock conditions. Perched water-table aquifers are common on top of the Glenwood Formation in shallow bedrock settings, and its contact with the overlying Platteville Formation is a common source of springs (for example Hall and others, 1911; Brick, 1997). However, the Glenwood Formation is commonly thin enough that minor vertical fractures will entirely breach it, locally allowing hydraulic connection between the overlying Platteville Formation and the underlying St. Peter Sandstone. It is believed that such fractures account for large volumes of water locally recharged into the St. Peter Sandstone where it is capped by the Glenwood Formation (Delin and Almendinger, 1993; Lindgren, 2001).

PLATTEVILLE FORMATION Hydrostratigraphic attributes Matrix porosity The Platteville Formation is composed chiefly of the carbonate rock component. Thin shale laminae are common, including regionally traceable bentonites. Plug tests of the carbonate rock indicate a very low to low permeability ranging from 10-7 to 10-4 md (MUGSP, 1980).

Secondary porosity Deep bedrock conditions—Little is known about the Platteville Formation in deep bedrock conditions in southeastern Minnesota. A single core examined as part of this study had no visible fractures or dissolution features (Fig. 42). Open joints and fractures are uncommon in cores of the Platteville Formation collected from deep bedrock conditions in Illinois (Kempton and others, 1987; Curry and others, 1988). Shallow bedrock conditions—Outcrop and

76

subsurface studies in Minnesota and Wisconsin have demonstrated that in shallow bedrock conditions the Platteville Formation contains discrete intervals with relatively well developed secondary porosity, separated by intervals with much lower porosity (for example Barr Engineering, 1994, 2000; Brick, 1997; Stocks, 1998). The Platteville Formation is well known for containing bedding-plane and vertical fractures typical of stressrelief conditions, and it also has vertical, flat fractures that are part of a large-scale, orthogonal system (for example Runkel, 1996a; Barr, 2001). Individual fractures commonly cut across the entire formation vertically, and are open as much as a few inches. Dissolution-enlarged fractures and cavities are locally large enough to permit human exploration and to lead to the development of sinkholes (Spong, 1980; Hoffman and Alexander, 1998). A 2,000-foot maze cave has developed in the Platteville Formation in Fillmore County (Spong, 1980). In areas where the Platteville Formation is the uppermost bedrock, an extensive epikarst system is commonly developed that presents a variety of challenges to construction activity.

Hydraulic attributes Deep bedrock conditions—Discrete interval packer tests of the Platteville Formation in deep bedrock conditions of Minnesota are not available. A comprehensive hydraulic investigation of the formation in Illinois, where it is generally similar in hydrostratigraphic properties to the Platteville Formation in Minnesota, indicated that its hydraulic conductivity typically ranges from 10 -3 to 10-2 foot per day. Two boreholes intersected discrete horizons with small, open fractures and packer tests indicated conductivities as high as 10-1 foot per day (Kempton and others, 1987; Curry and others, 1988). In Illinois, pump tests conducted across an open-hole interval exposing hundreds of feet of carbonate rock in the Galena Group and Platteville Formation caused a drawdown of over 256 feet, and failed to yield 15 gallons per minute, the minimum pump capacity. This implies a very low bulk hydraulic conductivity of less than 10-2 foot per day. Similar hydraulic conductivity values were obtained in analogous geologic settings of deep burial in Indiana and Wisconsin (Nicholas and others, 1987). A monitor well installed by Rowden and Libra (1990) in the Platteville Formation in northeast Iowa remained dry. Shallow bedrock conditions—The hydraulic conductivities of the Platteville Formation where it occurs in shallow bedrock conditions (Fig. 13) have been measured in detail at a number of sites in the Twin Cities Metropolitan area, such as at Minnesota Pollution Control Agency Superfund remediation sites (for example ERT, 1987; Barr Engineering, 1991; ENSR

Mower Co.

Houston Co. MINNESOTA IOWA

Fillmore Co. 00 10

90

10

00

900

00

e Riv r

0 90

Mississippi

11

900

0

800

800

0 0

20 mi 30 km

600

700

0 70

Jordan aquifer potentiometric surface St. Peter aquifer potentiometric surface

Figure 39. Potentiometric surfaces of the St. Peter aquifer (dashed line) and Jordan aquifer (solid line) in northeastern Iowa, demonstrating that part(s) of the intervening Prairie du Chien Group strata serve as a confining unit that creates hydraulic separation of the two aquifers. Modified from Horick (1989).

International, 1991), and at a proposed tunnel excavation (Liesch, 1973). Hoffman and Alexander (1998) reported that the hydraulic conductivity of the Platteville Formation at these and other Twin Cities Metropolitan area sites ranges over at least six orders of magnitude. It has extremely high hydraulic conductivity where secondary porosity is well developed, and extremely low conductivity where such features are poorly developed. At individual sites, large-scale permeability values based on pump tests have been calculated to be so high as to be considered infinite for specific intervals of the Platteville Formation, whereas other intervals tested were below the measurement threshold and therefore may be substantially lower than 10-1 foot per day (Fig. 13). Flow speeds of ground water in the

77

Platteville Formation in a shallow bedrock setting have been calculated to be faster than one mile per day in Wisconsin (Hoffman and Alexander, 1998) and in the Twin Cities Metropolitan area (Alexander and others, 2001). The packer testing conducted by Kempton and others (1987) and Curry and others (1988) on the Platteville Formation in Illinois demonstrated that its conductivity is roughly two orders of magnitude higher in shallow bedrock conditions than in conditions of deep burial, particularly where the rock is within the uppermost 40 feet of the bedrock surface. In those settings hydraulic conductivity ranged from 10-2 to 3 feet per day. Conductivity calculated from specific capacity tests of water wells in southeastern Minnesota typically

78

0 feet

50

100

A

Prairie du Chien Group

Lower St. Peter Sandstone

Upper St. Peter Sandstone

Glenwood Formation

NE Ramsey County 236088

LOCATION OF CROSS-SECTION

A'

A Dakota County 540212

Carbonate component

Fine clastic component

Coarse clastic component

Increasing counts

S. Ramsey County 200044

Blue Earth County 463778 Waseca County 215698

Prairie du Chien Group

Lower St. Peter Sandstone

Upper St. Peter Sandstone

Glenwood Formation

Faribault County 217033

Figure 40. Matrix hydrostratigraphic components within the St. Peter Sandstone across part of southeastern Minnesota. Representative natural gamma logs show that the upper part of the St. Peter Sandstone is dominated by the coarse clastic component. The lower St. Peter Sandstone contains fine clastic interbeds as thick as 30 feet.

Rice County 518699

A'

ranges from 2 to 130 feet per day, with an average value of 72 feet per day (Fig. 43).

Hydrogeologic synthesis The hydrogeologic properties of the Platteville Formation vary tremendously. In deep bedrock conditions, its low bulk hydraulic conductivity and relatively minor development of secondary pores suggest that it is most properly considered a confining unit. However, the discrete interval hydraulic tests conducted in Illinois (Kempton and others, 1987; Curry and others, 1988) demonstrated that relatively thin and widely spaced intervals of the formation can yield moderate quantities of water in such conditions of deep burial. In shallow bedrock conditions, the Platteville Formation is clearly a classic karstic aquifer, similar to the properties of the Galena aquifer described later in this report. It contains hydraulically significant fractures and dissolution features, as well as sinkholes and large caverns characteristic of classic karstic aquifers (Liesch, 1973; Barr Engineering, 1983, 1991, 1994; Hoffman and Alexander, 1998). Secondary pores appear to be most densely concentrated along a few specific stratigraphic intervals of the Platteville Formation, forming bedding-plane parallel conduit networks that provide preferential flow paths. Brick (1997) used outcrop attributes and the stratigraphic position of springs to demonstrate that bedding-plane parallel conduits occur at predictable stratigraphic positions within the Platteville Formation in the Twin Cities Metropolitan area. Discrete bedding-plane parallel conduit systems were also identified in the subsurface at the East Hennepin Avenue–General Mills Solvent Disposal Superfund site in Minneapolis (Barr Engineering, 1983, 1994, 2000). Individual conduit systems at the site differ from one another in static head and in inferred flow direction. They are separated from one another by bentonite beds or intervals of unfractured carbonate rock that serve as local confining units. Stocks (1998) described similar subsurface attributes in an investigation in Wisconsin, suggesting that ground water traveled chiefly along discrete intervals of relatively high secondary porosity, separated from one another by low conductivity aquitards. A recently completed dye-trace study by Alexander and others (2001) at the Camp Coldwater Spring site in the Twin Cities Metropolitan area demonstrated that ground water can travel at rates faster than one mile per day along such bedding-plane systems. Dye pulses were recorded for weeks to months after initial injection, reflecting an anastomosing pore system with variable flow speeds among the individual interconnected conduits. Recent ground-water models developed for site-

79

specific settings in the Twin Cities Metropolitan area and near Rochester have demonstrated that the Platteville Formation in those areas does not by itself effectively serve as a confining unit in shallow bedrock conditions. In the St. Louis Park area, contaminants were transported through the Platteville Formation in a well-developed network of secondary pores where the formation occurs as the uppermost bedrock (Lindgren, 1995). Near Rochester, a comprehensive ground-water model (Lindgren, 2001) demonstrated that a substantial amount of recharge to the St. Peter Sandstone occurs via fractures that are interconnected vertically through the Platteville Formation and underlying Glenwood Formation where those units occur near the sides of bluffs (Fig. 44). The common hydrogeologic depiction of the Platteville Formation as a confining bed (for example Kanivetsky, 1978; Delin and Woodward, 1984) is apparently accurate only for areas where it is deeply buried by younger bedrock and has negligible development of secondary porosity. In such a setting the term "Decorah–Platteville–Glenwood confining unit" is appropriately applied to that part of the stratigraphic section. In shallow bedrock conditions, however, the Platteville Formation is more hydrogeologically complex. It is a karstic carbonate aquifer that serves as a source of water for over 500 wells in the County Well Index database. It has the potential to provide confinement locally only where it is not breached by interconnected vertical fracture networks (Barr Engineering, 1983; Lindgren, 2001). The Platteville Formation is generally similar in hydrogeologic properties to other carbonate rock layers in southeastern Minnesota, such as the Prosser Limestone, which have historically been accepted as karstic aquifers. The Platteville Formation is not considered a "major" karst system in this report because it is relatively thin and has a limited distribution as uppermost bedrock compared to the major karst systems.

DECORAH SHALE Hydrostratigraphic attributes Matrix porosity The Decorah Shale is composed chiefly of the fine clastic component, with subordinate interbeds of carbonate rock. It is over 90 feet thick in the Twin Cities Metropolitan area, and thins to less than 30 feet at the Iowa border. Most of the Decorah Shale is actually shale, with a vertical permeability that ranges from 10-5 to 10-4 md based on plug tests of similar shale in the Glenwood Formation. The carbonate interbeds in the Decorah Shale are similar to those in the

A.

400

Figure 41. Hydraulic conductivity data for the St. Peter Sandstone calculated from specific capacity tests. See Figure 11 for an explanation of box plots. A. Scatter plot showing the relationship between the depth of the open-hole interval below the bedrock surface and hydraulic conductivity. Shallower wells tend to have higher conductivity. B. Box plot of hydraulic conductivity values for deep bedrock conditions. C. Box plot of hydraulic conductivity values for shallow bedrock conditions. Nineteen outlying values greater than 250 feet per day are not shown.

Conductivity in feet per day

350 300 250 200 150 100 50 0 0

100

200

300

400

500

600

Distance in feet between the bedrock surface and the open-hole top

DEEP BEDROCK CONDITIONS

SHALLOW BEDROCK CONDITIONS

C.

250

250

200

200

150

150 Range

Range

B.

100

50

100

50

0

0 72 samples Average 15.9 feet per day

853 samples Average 38.7 feet per day

Platteville Formation, and therefore probably have similarly low matrix permeabilities.

Secondary porosity Deep bedrock conditions—Secondary porosity characteristics in the Decorah Shale have not been studied in deep bedrock conditions. As a relatively ductile, high porosity unit in a layered sequence of bedrock strata, open fractures are assumed to be uncommon to absent. Shallow bedrock conditions—Open fractures in the Decorah Shale are known to occur in shallow bedrock settings (Fig. 8; Hall and others, 1911). These include nonsystematic stress relief fractures and orthogonal fracture sets that may be part of a regional system.

80

Hydraulic attributes Deep bedrock conditions—The Decorah Shale has not been subjected to discrete interval packer tests, but based on the large shale content it can be expected to have a vertical hydraulic conductivity of about 10-7 to 10-5 foot per day (Freeze and Cherry, 1979). Shallow bedrock conditions—Hydraulic conductivity values calculated from specific capacity data for six wells in Ramsey and Steele Counties range from about 2 to nearly 160 feet per day, and average 60.1 feet per day (Fig. 45). This suggests that fracture porosity can result in the development of moderate hydraulic conductivity.

Hydrogeologic synthesis

Many studies have demonstrated that the Decorah Shale serves as an effective confining bed, even in shallow bedrock conditions, and we classify it as a confining unit together with the interbedded shale and carbonate strata of the overlying Cummingsville Formation. Perched water-table aquifers are common above the Decorah Shale, and springs are preferentially located along hillsides at elevations that correspond to the top of the formation or to the lower part of the Cummingsville Formation. These springs approximate the position of an important hydrogeologic boundary that results in a process called "focused recharge" (Fig. 44). Focused recharge occurs along hillsides where large volumes of water emitted from the Galena aquifer travel rapidly downward across the eroded edge of the Decorah Shale and Platteville and Glenwood Formations, eventually entering the St. Peter aquifer in a relatively limited, or "focused" area of recharge (for example Lindgren, 2001). Water moving from the Galena aquifer to the St. Peter aquifer in these areas travels at the surface, through thin surficial deposits, and along secondary pore networks in the shallow bedrock. The hydrogeologic significance of open fractures that are known to occur in shallow bedrock conditions has not been evaluated, but 25 wells in southeastern Minnesota use the Decorah Shale as a water source, suggesting that such fractures can yield economic quantities of water.

GALENA THROUGH CEDAR VALLEY GROUPS The Paleozoic strata from the base of the Galena Group through the preserved thickness of the Cedar Valley Group are composed mostly of carbonate rock, with subordinate beds of the fine clastic component, chiefly shale (Fig. 46). Although the entire stratigraphic section has historically been treated as a single aquifer, the "upper carbonate aquifer," investigations in Minnesota and adjoining states over the past 15 years have demonstrated that it contains distinct hydrostratigraphic components that differ substantially from one another in their hydraulic properties. These investigations are ongoing (for example Campion, 2002), and until they are completed our understanding of the hydrogeologic attributes for this part of the Paleozoic section remains relatively limited. The limited amount of hydrogeologic information available for Galena through Cedar Valley Groups strata prevents a reasonably thorough description of hydrostratigraphic and hydraulic attributes for each of the many individual lithostratigraphic units to be

81

presented in this report. Therefore these strata are described together, highlighting the general stratigraphic distribution of their two major matrix hydrostratigraphic components, carbonate rock and shale, as well as a summary of the available information on secondary porosity and hydraulic attributes. This is followed by a synthesis of these data in which we subdivide the stratigraphic section into discrete hydrogeologic units and karst systems.

Hydrostratigraphic attributes Matrix porosity The dominant component of the Galena through Cedar Valley Groups is carbonate rock, mostly finely crystalline to dense, microcrystalline dolostone or limestone. It has not been tested with laboratory methods to calculate porosity and permeability, but can be expected to have very low to low porosity and permeability based on field examination and plug tests of generally similar older carbonate strata in the Paleozoic section. Shale is abundant in only a few intervals (Fig. 46; Olsen, 1988a; Mossler, 1995a, 1998). The Chickasaw Member of the Little Cedar Formation, the lower part of the Pinicon Ridge Formation, and the upper Dubuque Formation are the thickest intervals of strata composed largely of shale. The lower Maquoketa Formation in western Fillmore County and across Mower County is composed of interbedded shale and shaly dolostone. Additionally, beds of shale as thick as a few feet are intercalated with carbonate beds of similar thickness in the Coralville Formation, the upper part of the Little Cedar Formation, and the lower part of the Galena Group (Cummingsville Formation). Porosity and permeability have not been calculated for plug samples collected in Minnesota, but core samples of shale from the Maquoketa Formation in Wisconsin yielded very low values (Eaton and others, 2000).

Secondary porosity Deep bedrock conditions—The abundance of macroscopic secondary pores is known to be variably distributed and stratigraphically controlled in the carbonate rock of the Galena through Cedar Valley Groups strata (Fig. 46) in deep bedrock conditions. Site specific studies indicate that secondary pores in the Galena Group are apparently concentrated in discrete, relatively thin intervals separated by much thicker bodies of relatively tight carbonate rock (for example Curry and others, 1988; Delta Environmental Consultants, Inc., 1995, 2002). A site-remediation study near Spring Valley revealed that dissolution features are concentrated in the lower part of the Prosser

Fractures per foot of core

Visual porosity

Depth in feet below bedrock surface

Matrix hydrostratigraphic component

Lithostratigraphic unit

0

100%

0

5

150

Prosser Limestone

200

250

Decorah Shale

Cummingsville Formation

Galena Group

Stewartville Formation

Figure 42. Visual porosity and mesoscopic fracture abundance in core H-1, Freeborn County. Macroscopic secondary pores are abundant in the Shakopee Formation in deep bedrock conditions, and were also noted along Cummingsville Formation–Prosser Limestone contact strata and in the lower part of the Stewartville Formation in shallower conditions. See Figure 1 for location.

300

Limestone and uppermost part of the Cummingsville Formation in deep bedrock conditions. A core from Freeborn County was examined as part of this investigation (Fig. 42), and the distribution of visible pores was similar: in deep bedrock conditions cavities and mesoscopic vertical fractures were present only in relatively thin horizons clustered near the Cummingsville Formation–Prosser Limestone contact. Witzke and Bunker (1984) described the stratigraphic distribution of secondary pores in carbonate strata above the Galena Group where it occurs in deep bedrock conditions in northern Iowa. They noted particularly high densities of cavities in the uppermost part of the Maquoketa Formation, the middle part of the Pinicon Ridge Formation, the lower part of the Coralville Formation, and throughout much of the Spillville Formation and Bassett Member of the Little Cedar Formation. Secondary porosity in beds dominated by shale under deep conditions of burial has not been described. Fractures known to occur in shallow bedrock conditions (for example Eaton and others, 2000) have not been described under deeper conditions of burial in Minnesota, although they could be present locally (Ryder, 1996). Shallow bedrock conditions—The carbonatedominated strata of the Galena through Cedar Valley Groups contain all of the porosity attributes typical of classic karsted rock (Alexander and Lively, 1995; Alexander and others, 1996; Witthuhn and Alexander, 1996), including nonsystematic fractures along bedding planes and at high angles to bedding, as well as systematic fractures that are part of a large-scale, orthogonal system. Individual vertical fractures with

350

Platteville Formation Glenwood Formation

St. Peter Sandstone (incomplete)

400

450

Shakopee Formation

Prairie du Chien Group

500

550

600

Coarse clastic component Fine clastic component Carbonate component

82

apertures of a few inches are known locally to span entire outcrops that are tens of feet in height. Dissolution has widened fracture apertures and has also produced vuggy pores of wide-ranging abundance and size. The horizons with relatively high secondary porosity in deep bedrock conditions (Witzke and Bunker, 1984) appear also to be of relatively high porosity in shallow bedrock conditions based on outcrop and borehole investigations (Delta Environmental Consultants, Inc., 1995, 2002; Mossler, 1998; Paillet and others, 2000). In addition, the presence of large cavern systems in the lower Dubuque, Stewartville, and upper Cummingsville Formations and relatively high density of sinkholes in the Prosser Limestone, Lithograph City and Stewartville Formations suggest that those formations may be especially susceptible to the development of large-scale, interconnected networks of pores. The intervals dominated by shale are known to be fractured in shallow bedrock conditions. Eaton and others (2000) described open fractures in the Maquoketa Formation shale in southeastern Wisconsin where the formation is buried beneath less than 200 feet of overlying bedrock. In Minnesota, nonsystematic stressrelease fractures are common, and larger-scale,

A.

systematic fractures are known to cut entirely across relatively thick shaly intervals such as the upper Dubuque and lower Maquoketa Formations based on observations of caves and outcrops (for example Alexander and Lively, 1995).

Hydraulic attributes Deep bedrock conditions—The results of hydraulic tests of the Galena through Cedar Valley Groups (Fig. 13) reflect the differential development of secondary porosity: intervals of strata where secondary pores are relatively large and abundant are of moderate to high conductivity, whereas thick intervals of tight strata are orders of magnitude lower in conductivity. Nicholas and others (1987) tested a saturated open-hole interval exposing about 200 feet of Galena Group strata in Illinois and were unable to achieve the minimum pump capacity of 15 gallons per minute, while creating a drawdown of over 256 feet. This indicated a hydraulic conductivity of less than 10-2 foot per day. At a different site in Illinois, Graese and others (1988) conducted discrete interval pump tests that indicated the Galena Group typically has a conductivity of 10 -3 to 10-2 foot per day or less, with the exception of three packed intervals with conductivities that ranged from 1.4 to

B.

400 150 Range

Conductivity in feet per day

SHALLOW BEDROCK CONDITIONS 500

200

100

300

200

50

100

0

0 0

10

20

30

40

50

60

43 samples Average 72.0 feet per day

Distance in feet between the bedrock surface and the open-hole top

Figure 43. Hydraulic conductivity data for the Platteville Formation calculated from specific capacity tests. See Figure 11 for an explanation of box plots. A. Scatter plot showing the relationship between the depth of the open-hole interval below the bedrock surface and hydraulic conductivity. B. Box plot of hydraulic conductivity values for shallow bedrock conditions. One outlying value greater than 500 feet per day is not shown.

83

14 feet per day (Kempton and others, 1987). Hydraulic conductivity data for the Galena Group in deep bedrock conditions within Minnesota are scarce. Three wells in our specific capacity database ranged in conductivity from 4 to 10 feet per day and averaged 6.5 feet per day (Fig. 47). Some Galena Group variance wells permitted by the Minnesota Department of Health in Mower County have failed to yield adequate supplies for domestic purposes in deep bedrock conditions. Packer tests at the Spring Valley site remediation investigation in western Fillmore County yielded variable results. Two wells at the site were pumped dry in 10 minutes or less when pumping at 5 to 6 gallons per minute, indicating a low hydraulic conductivity, while a third well had a moderate hydraulic conductivity of between 2 and 7 feet per day (Delta Environmental Consultants, Inc., 1995, 1998). Hydraulic conductivity of the carbonate strata that overlie the Galena Group in deep bedrock conditions is known chiefly from packer tests conducted by Libra and Hallberg (1985) in northern Iowa, along with one test conducted on the Spillville Formation by Green and others (1997) near LeRoy, Minnesota. These authors reported values that range over three orders of magnitude. Three tests of the Spillville Formation range from 0.5 to 39 feet per day (Fig. 13). The Bassett Member of the Little Cedar Formation tested as low as 0.8 foot per day at one site, whereas another borehole had a hydraulic conductivity of more than 190 feet per day, greater than any other individual formation tested in deep bedrock conditions. A single test of the Coralville and upper part of the Little Cedar Formations yielded a bulk hydraulic conductivity value of 26 feet per day. Intervals dominated by shale have not been tested in Minnesota, but packer tests in Illinois under deep bedrock conditions resulted in hydraulic conductivity values that typically were less than 10-3 foot per day, with several tests below the measurement limit of 104 foot per day (Curry and others, 1988; Graese and others, 1988). Green and others (1997) demonstrated that the vertical conductivity of shaly strata in the lowermost part of the Pinicon Ridge Formation was so low as to provide effective hydraulic confinement at LeRoy, Minnesota, on the basis of a pump test conducted on the underlying Spillville Formation and observations of monitor wells higher in the stratigraphic section. Shallow bedrock conditions—It is well known that in shallow bedrock conditions the carbonate strata in the Galena through Cedar Valley Groups contain large, interconnected conduits that can accommodate enormous volumes of water traveling at rates measured

84

in hundreds of feet per day to several miles per day (for example Alexander and Lively, 1995; Green and others, 1997; Delta Environmental Consultants, Inc., 1998) and therefore at a relatively large scale commonly have a high bulk hydraulic conductivity. Specific capacity tests of the Galena Group in Minnesota average 64.6 feet per day. The statistically acceptable range of conductivity calculated from specific capacity tests of the strata above the Galena group has a maximum value of 170 feet per day, and the average conductivity is 67 feet per day. Several packed intervals tested by Libra and Hallberg (1985) in northern Iowa had conductivities greater than 50 feet per day, and the productivity of wells that draw water from the Coralville and Lithograph City Formations in two boreholes in Iowa, and of the Spillville Formation at Austin, Minnesota was so great that measurable drawdown did not occur while pumping (Libra and Hallberg, 1985; Paillet and others, 2000). At the Austin site, 93 percent of the water was contributed from a discrete 5-foot horizon with large secondary pores (Fig. 48). At a smaller scale, packer tests reveal extreme variability in hydraulic conductivity, and the presence within the Galena through Cedar Valley Groups of intervals of carbonate rock with much lower conductivity than that calculated for boreholes with longer open-hole intervals. Packer testing of 0.6-meter intervals of the Galena Group in Wisconsin demonstrated that individual, carbonate-dominated, hydrostratigraphic units have conductivities as low as 10-3 foot per day (Stocks, 1998). Rigorous hydraulic tests of this kind, which are necessary to recognize discrete horizons of low conductivity that may serve as confining units, have not been conducted in Minnesota. Intervals dominated by shale have been tested at only one site in Minnesota, near Spring Valley, where hydraulic conductivity and ground-water flow speeds varied dramatically. The Dubuque Formation had a horizontal hydraulic conductivity ranging from 8.9 x 10 -4 to 3 x 10 -2 foot per day (Delta Environmental Consultants, Inc., 1995). The lower Maquoketa Formation at the same site yielded variable test results. Some wells were dry after a few minutes of pumping at a rate of 5 to 7 gallons per minute whereas other intervals yielded hydraulic conductivity values as high as 1 to 2 feet per day. Dye tracers in fractured Maquoketa and Dubuque Formation strata traveled laterally at rates ranging from 0.23 mile per year to 1.8 miles per day, and reached depths of 260 feet into the underlying Galena Formation in less than 7 months (Delta Environmental Consultants, Inc., 1998). The range in conductivity measured at Spring Valley is

85 2 mi

Non-systematic fractures (some dissolution enlarged) Systematic fractures (some dissolution enlarged) Dissolution features—cavities and enlarged bedding-plane fractures Sinkhole Spring

Fine clastic component Carbonate component Surficial deposits Ground and surface water flow direction

St. Peter Sandstone

Platteville Formation Glenwood Formation

Galena Group– Cummingsville Formation Decorah Shale

Galena Group– Prosser Limestone

Coarse clastic component

EXPLANATION

Relatively rapid infiltration of precipitation and horizontal flow along fractures and cavities

Figure 44. Ground- and surface-water flow in a typical hillside setting where the St. Peter Sandstone through Galena Group are the uppermost bedrock. Water in the Galena aquifer moves downward through vertical fractures, and laterally along bedding-plane conduits where it is perched above relatively unfractured shale and carbonate rock of the Cummingsville Formation and Decorah Shale. Galena aquifer water is emitted at springs along hillsides, and travels at the surface or through thin surficial deposits downward to eventually reach the St. Peter aquifer. A significant component of Galena aquifer water also travels downward through shallow bedrock fractures in the Decorah Shale and Platteville and Glenwood Formations to reach the St. Peter aquifer. Recharge of the St. Peter aquifer is commonly focused in discrete areas in such a setting. Based on flowpath lines modeled by Lindgren (2001) for the Rochester area, and supplemented with information based on outcrop observations (for example Runkel and Tipping, 1998), sinkhole maps (for example Witthuhn and Alexander, 1996), dye-trace and related karst studies (Alexander and Lively, 1995; Alexander and others, 1996), and borehole investigations (Delta Environmental Consultants, Inc., 1995) of this part of the stratigraphic section in nearby areas of Olmsted and Fillmore Counties.

0

100 ft

Focused recharge to St. Peter aquifer

Water discharges from Galena Group bedrock surfaces as springs or travels downward through fractures

generally consistent with that reported for the Maquoketa Formation in southeastern Wisconsin and Illinois, where fractured shaly beds commonly have conductivities of a few feet per day, and relatively unfractured shaly intervals are commonly 10-3 foot per day or less (Curry and others, 1988; Graese and others, 1988; Eaton and others, 2000).

Hydrogeologic synthesis The Galena through Cedar Valley Groups are composed of discrete intervals having moderate to extremely high hydraulic conductivity separated by confining units composed of unfractured carbonate rock or shale that are several orders of magnitude lower in conductivity (Fig. 46). The stratigraphic section is divided into hydrogeologic units on the basis of matrix hydrostratigraphic properties: the shaly strata of the upper Dubuque/lower Maquoketa Formations, lower Pinicon Ridge Formation, and Chickasaw Member separate four intervals dominated by carbonate rock. In deep bedrock conditions, hydraulic tests and potentiometric levels in nested well screens have demonstrated that each of the shaly intervals have the ability to serve as confining units (Libra and Hallberg, 1985; Rowden and Libra, 1990; Green and others, 1997), an interpretation additionally supported by documentation of ground-water ages that are stratified

B.

A.

SHALLOW BEDROCK CONDITIONS

160

160

140

140

120

120

100

100

Range

Conductivity in feet per day

in a manner consistent with hydraulic separation (Green and others, 1997; Tipping, 1997; Campion, 2002). The individual bodies of carbonate rock separated by shaly confining units can be considered aquifers as bulk units because they all yield economic quantities of water to wells in Minnesota, even in deep bedrock conditions. These are the Galena aquifer (including the Prosser Limestone, Stewartville and lower Dubuque Formations), the upper Maquoketa–Spillville aquifer, and the lower and upper Cedar Valley aquifers (Green and others, 1997). Water is produced from these bodies of rock chiefly along discrete intervals of relatively high porosity. It is noteworthy that individual aquifers may contain thick intervals of carbonate rock much lower in conductivity and that such intervals have been proven to serve as confining units in some areas (Buchmiller and others, 1985; Nicholas and others, 1987; Graese and others 1988; Delta Environmental Consultants, Inc., 2002). For example, at Spring Valley, where the Galena Group occurs in a transitional setting between shallow and deep bedrock, such intervals provide confinement sufficient to separate discrete bedding-plane conduits that differ from one another in potentiometric level and inferred flow direction (Delta Environmental Consultants, Inc., 2002). In Illinois and Iowa, where the Galena Group is much more deeply buried by

80

80

60

60

40

40

20

20 0

0 0

5

10

15

20

25

6 samples Average 60.1 feet per day

Distance in feet between the bedrock surface and the open-hole top

Figure 45. Hydraulic conductivity data for the Decorah Shale calculated from specific capacity tests. See Figure 11 for an explanation of box plots. A. Scatter plot showing the relationship between the depth of the open-hole interval below the bedrock surface and hydraulic conductivity. B. Box plot of hydraulic conductivity values for shallow bedrock conditions.

86

younger bedrock, it is dominated by thick intervals of carbonate rock with low conductivity and therefore the entire group is typically classified as a confining unit (for example Buchmiller and others, 1985; Nicholas and others, 1987). The hydrogeologic system in shallow bedrock conditions is more complex, with enormous variations in hydraulic conductivity and flow features that are characteristic of "triple porosity" karstic aquifers (Figs. 46, 49). The much greater degree to which secondary porosity is developed, especially in the size, spacing, and interconnectivity of fractures, is reflected in an overall greater range in conductivity, and at least locally can compromise the ability of low permeability units to provide confinement. For example, outcrop and cave investigations have documented the presence of large vertical fractures that pass continuously through the Maquoketa and Dubuque Formations, and dye-trace studies have demonstrated that ground water travels across them at rapid rates (Alexander and Lively, 1995; Delta Environmental Consultants, Inc., 1995, 1998). Tipping (1997) and Campion (2002) noted that stratification in ground-water ages in Mower County is less pronounced in shallow bedrock conditions, particularly in areas with a relatively thin cover of surficial deposits and near bedrock faults. Nevertheless, even in such a setting, Campion (2002) delineated separate potentiometric surfaces for the upper Cedar Valley, lower Cedar Valley, and Spillville–Maquoketa aquifers, indicating that the confining units between them provide some degree of hydraulic separation. In shallow bedrock conditions, the strata between the base of the Galena Group and the top of the Lithograph City Formation contain two of southeastern Minnesota's major karst systems: the Galena–Spillville karst and the Cedar Valley karst (Figs. 46, 49). These karst systems are separated by the shaly strata of the Pinicon Ridge Formation. However, each karst system contains intervals of shaly strata that have been demonstrated to act as confining units in relatively deep bedrock conditions, and therefore they may be further subdivided with continued study. Investigations thus far have indicated that in shallow bedrock conditions, sinkholes, caves, and dye traces pass through shaly intervals within each of the individual karst systems. Together the Galena–Spillville karst system and Cedar Valley karst system occur as the uppermost bedrock across a large part of southeastern Minnesota, although the latter has been recognized only in Mower County at the time of this report. Karstic features such as sinkholes, springs, stream sinks, and dry valleys are best expressed and most abundant where overlying unconsolidated glacial drift is relatively thin (less than 50 feet). The differential stratigraphic distribution of

87

secondary pores is apparently reflected in some of the karstic characteristics: relatively high densities of sinkholes occur in the Prosser Limestone, Stewartville and Cedar Valley Formations, and springs most commonly lie at the approximate contacts between carbonate rock and shale, such as in the lower Cummingsville Formation. Lateral movement of ground water can be expected to occur preferentially along a few discrete bedding-plane parallel intervals with welldeveloped secondary pore systems, such as those described for the Galena karst system in northeastern Iowa (Keeler, 1997). Recharge to karstic aquifers such as the Galena– Spillville occurs relatively rapidly through fractures and dissolution cavities, and ground water can travel laterally at rates commonly measured in miles per day (Alexander and Lively, 1995; Alexander and others, 1996). Flow paths in this aquifer system have been demonstrated to commonly cross surface watershed divides. Furthermore, the direction of ground-water movement along conduit systems does not consistently correspond to ground-water flow directions inferred from regional-scale potentiometric maps. The Dubuque–Galena part of the Galena–Spillville karst system contains numerous caves where it occurs as the uppermost bedrock. There are two fundamentally different types of caves present. The caves that develop in the Dubuque and Stewartville Formations tend to be fracture-controlled, high-gradient, water-inlet maze caves (for example Mystery Cave, Spring Valley Caverns, Goliath Cave). These caves serve hydraulically to conduct surface water into the aquifer via sinkholes. The second type of caves typically develops in the mid to lower Cummingsville Formation, and their attributes in part may reflect the presence of laterally continuous shale beds in the Cummingsville Formation. Such caves (Coldwater Cave in Iowa, Pine Cave, Tyson's Spring Cave, and Deep Lake Cave) are large, low-gradient dendritic conduits that collect and drain water from stratigraphically higher levels of the Galena Group. In planview these caves resemble surface drainages. They are typically flat and terminate at springs, draining large conduits. Vertical hydraulic gradients are locally larger than the horizontal gradient in this part of the karst system.

DISCUSSION: CLASSIFICATION OF AQUIFERS AND CONFINING UNITS Our classification of aquifers and confining units (Plates 1, 2) recognizes eleven aquifers and ten confining units at a regional scale. This new hydrogeologic classification is based on hydraulic data interpreted within the context of a hydrostratigraphic framework that depicts the distribution of porosity and

88

Odcr

Ogcm

Ogpr

Ogsv

Odub

Omaq

Confining unit

100 feet

0

b v r m

r Odc

Ogc

Ogp

Ogs

Odu

Dclp

Dspl

maq

O

l

Dsp

Dclc

Dcum

Dcuu

5 miles

Surfical deposits

Carbonate component

Setting of Mystery Cave

Odcr—Decorah Shale Ogcm—Cummingsville Formation Ogpr—Prosser Limestone Odub—Dubuque Formation Omaq—Maquoketa Formation Dspl—Spillville Formation Dclc—Chickasaw Member of the Little Cedar Formation Dcum—Coralville Formation and Hinkle and Eagle Center Members of the Little Cedar Formation

Lithostratigraphic units:

Springs common

Sinkholes

Dissolution features—cavities and enlarged bedding-plane fractures

Systematic fractures (some dissolution enlarged)

Nonsystematic fractures (some dissolution enlarged)

EXPLANATION

Setting of Spring Valley site-remediation study

Fine clastic component

Galena–Spillville karst system

S o u t h we s t e r n F i l l m o r e C o u n t y

Ogsv—Stewartville Formation Dclp—Pinicon Ridge Formation and Bassett Member of the Little Cedar Formation Dcuu—Lithograph City Formation

Figure 46. Cross-section of the Galena through Cedar Valley Groups in Mower and Fillmore Counties showing lithostratigraphic units, hydrostratigraphic attributes, aquifers and confining units, and two major karst systems. These strata contain four aquifers that are composed chiefly of carbonate rock with fractures and/or dissolution cavities. Note that some intervals of carbonate rock in the Galena aquifer likely have the ability to provide confinement in deep bedrock settings where secondary porosity is poorly developed. The four aquifers are separated by three confining units composed largely of shale. Where the shales occur near the fractured bedrock surface, their relative effectiveness to provide confinement is likely highly variable. The Cedar Valley and Galena–Spillville karst systems occur where those units are near the bedrock surface. Lithostratigraphic units and geologic structures are modified from Mossler (1995a, 1998). Hydrostratigraphic attributes are chiefly from Libra and Hallberg (1985), Witzke and Bunker (1985), Green and others (1997), and Mossler (1998).

Galena aquifer

Maquoketa– Spillville aquifer

Lower Cedar Valley aquifer

Upper Cedar Valley aquifer

LeRoy karst study

Cedar Valley karst system

S o u t h e a s t e r n M owe r C o u n t y

permeability in the Paleozoic bedrock. It therefore provides a more realistic depiction of aquifers and confining units in a variety of geologic settings across southeastern Minnesota than previous classifications, and more accurately characterizes the hydraulic properties of each of these hydrogeologic units at varying conditions of burial. Our classification follows standard conventions in the use of the terms "aquifer" and "confining unit" (for example Fetter, 1988; Subsurface-Water Glossary Working Group, 1989). An aquifer is a body of rock that is sufficiently permeable to yield economic quantities of water to wells and springs. A confining unit is a body of rock of relatively low permeability that is stratigraphically adjacent to one or more aquifers. The standard definitions of "aquifer" and "confining unit" are not entirely mutually exclusive—for example bodies of rock that can yield economic quantities of water through bedding-plane parallel fractures can be of sufficiently low vertical conductivity to hydraulically confine an underlying aquifer. The first step in our classification of hydrogeologic units in southeastern Minnesota therefore was to delineate major confining units in deep bedrock conditions, where secondary pores are relatively diminished. The "major" confining units we recognize are regionally extensive, relatively thick intervals of fine clastic and carbonate rock that have been demonstrated to be of sufficiently low bulk vertical conductivity to provide confinement under particular conditions of hydraulic stress, and where they are not breached by vertical fractures. They meet all the standard criteria characteristic of confining units (for example Fetter, 1988; Subsurface-Water Glossary Working Group, 1989) such as having a vertical hydraulic conductivity of less than 10-2 foot per day. Furthermore, each of our regional confining units except the "middle Mt. Simon Sandstone" has been demonstrated in this report to hydraulically separate aquifers with differential static heads in southeastern Minnesota or extreme northern Iowa. The middle Mt. Simon Sandstone confining unit has been demonstrated to do so in southeastern Wisconsin and Illinois. The aquifers we define are the bodies of rock dominated by coarse clastic strata or relatively thick intervals of carbonate rock with abundant secondary pores that are known to yield moderate to large volumes of water in deep bedrock settings. The coarse clastic aquifers typically have a horizontal hydraulic conductivity between 5 and 60 feet per day in deep bedrock conditions. The carbonate rock aquifers are much more variable in conductivity, and typically consist internally of relatively narrow intervals of high

89

to very high conductivity (tens to thousands of feet per day) separated by thick intervals of tight carbonate rock that is orders of magnitude lower in conductivity. The regional scale at which we have defined our aquifers and confining units results in a generalized classification in which some individual hydrogeologic units are internally variable in hydrostratigraphic and hydraulic properties. For example, the St. Lawrence and lower St. Peter confining units internally contain discrete, bedding-plane parallel intervals of secondary pores or coarse clastic interbeds that have moderate to high conductivity. Conversely, some aquifers we have defined at a regional scale are known to contain internal confining units, such as fine clastic beds in the upper Mt. Simon aquifer, and carbonate rock with few secondary pores in the Galena aquifer. Our limited understanding of the lateral distribution of these heterogeneities prevents us from depicting them at a regional scale. Regional-scale classification of aquifers and confining units in shallow bedrock conditions is difficult because of the ubiquitous presence of fractures, relatively great abundance of dissolution features, and the limited number of comprehensive ground-water studies conducted in such settings in Minnesota. All of the confining units we recognize in deep conditions of burial have been demonstrated to locally have moderate to high bulk conductivities in shallow bedrock settings. More importantly, it has been clearly shown that these confining units are in places hydraulically breached by fractures and dissolution features where they occur close to the bedrock surface (Alexander and Lively, 1995; Lindgren, 2001). On the other hand, parts of some confining units, such as the Decorah Shale and fine clastic strata in the Franconia Formation, have been demonstrated to provide confinement at some scale in shallow bedrock conditions even though elsewhere they may be breached vertically by fractures. With consideration of these complexities, we have tentatively applied a single hydrogeologic classification of aquifers and confining units for both shallow and deep bedrock conditions. Although each of the confining units have the potential to provide hydraulic separation at some scale, in shallow bedrock conditions, the relative effectiveness and scale at which they can do so is practically untested in southeastern Minnesota. Therefore as a practical matter for environmental investigations, the ability of these confining units to provide hydraulic separation in shallow bedrock conditions has to be established at individual sites— it cannot be assumed.

SUMMARY

A.

800

Conductivity in feet per day

700 600 500 400 300 200 100 0 0

50

100

150

200

250

300

Distance in feet between the bedrock surface and the open-hole top

DEEP BEDROCK CONDITIONS

SHALLOW BEDROCK CONDITIONS

C.

500

500

400

400

300

300

Range

Range

B.

200

200

100

100

0

0 3 samples Average 6.5 feet per day

170 samples Average 64.6 feet per day

E.

SHALLOW BEDROCK CONDITIONS

1000

1000

800

800

600

600

Range

Conductivity in feet per day

D.

400

400

200

200

0

0 0

50

100

150

200

250

300

Distance in feet between the bedrock surface and the open-hole top

90

217 samples Average 67.0 feet per day

This study demonstrates that individual lithostratigraphic units in the Paleozoic bedrock of southeastern Minnesota have great variability in their internal hydrostratigraphic character (Plates 1, 2). Variations in matrix and secondary pores result in measurable and predictable variability in hydraulic properties within individual lithostratigraphic units. The hydrogeologic system is best understood when studied within the context of the hydrostratigraphic attributes of these rocks, such as within the context of the threedimensional distribution of porosity and permeability. The Paleozoic bedrock of southeastern Minnesota can be divided into three principal matrix hydrostratigraphic components: 1. Coarse clastic rock of high porosity and permeability; 2. Fine clastic rock of low porosity and permeability; and 3. Carbonate rock, also of low porosity and permeability. The groundwater system appears to be relatively simple and predictable in conditions of deep burial by younger bedrock. Under these conditions, coarse clastic strata are of relatively high hydraulic conductivity, typically ranging from a few feet per day to a few tens of feet per day, presumably reflecting flow through large, wellconnected intergranular pore spaces. In contrast, the matrix conductivity of the fine clastic and carbonate rock components is sufficiently low in a vertical direction (10 -7 to 10 -3 foot per day) that intervals dominated by these components can provide hydraulic confinement. The abundance and distribution of secondary pores overprinted on matrix hydrostratigraphic attributes substantially affects hydraulic properties. We demonstrate that in deep bedrock conditions carbonate strata contain discrete, stratigraphically controlled horizons with abundant secondary pores, separated from one another by relatively tight carbonate rock. These

horizons have a wide range in conductivity, in part depending on the scale at which they are tested, but can be as great as hundreds of feet per day. Our understanding of the hydraulic importance of fractures in deep bedrock conditions is much more limited. Bedding-plane parallel fractures and systematic fractures are known to exist in all rock types at least locally, but their abundance and connectivity is not documented. Standard aquifer and specific capacity test data tentatively indicate that fracture networks may at least locally be hydraulically significant. Regional-scale connectivity of such networks may provide an enhanced, large-scale conductivity to the aquifers and confining beds in southeastern Minnesota that has not been measured by the standard hydraulic tests performed thus far. In shallow conditions of burial, secondary pores including abundant fractures are common in all three matrix hydrostratigraphic components. Individual layers composed of coarse clastic, fine clastic, or carbonate components in relatively shallow conditions are very different hydrogeologically from the same layers in relatively deep bedrock conditions because secondary porosity is vastly different. In the shallow setting they have a higher bulk hydraulic conductivity, a greater range in conductivity, and may transmit the greatest volumes of ground water through conduit networks. Water in the conduit network is typically recharged through vertical fractures and transported laterally through an interconnected system of bedding-plane parallel secondary pores with high hydraulic conductivity. These preferential intervals of flow are separated from each other by blocks with substantially lower conductivity, which reflects the matrix permeability. Flow paths in such conditions are much less predictable than in deep conditions of burial and flow speeds have been documented to be faster than

Figure 47. Hydraulic conductivity data for Galena Group through Cedar Valley Group strata calculated from specific capacity tests. See Figure 11 for an explanation of box plots. A. Scatter plot showing the relationship between the depth of the open-hole interval below the bedrock surface and hydraulic conductivity for wells open only to the Galena Group. Shallower wells tend to have higher conductivity. B. Box plot of Galena Group hydraulic conductivity values for deep bedrock conditions. C. Box plot of Galena Group hydraulic conductivity values for shallow bedrock conditions. Two outlying values greater than 500 feet per day are not shown. D. Scatter plot showing the relationship between the depth of the open-hole interval below the bedrock surface and hydraulic conductivity for wells open to strata above the Galena Group. Shallower wells tend to have higher conductivity. E. Box plot of hydraulic conductivity values for strata above the Galena Group in shallow bedrock conditions.

91

Matrix hydrostratigraphic component

Depth in feet below the land surface and bedrock surface (italics)

Gamma log (counts per second) 0

50

Caliper (inches) 100

0

Stationary flow (gallons per minute)

Televiewer

3

6

-4

-2

0

2

Location and percent borehole transmissivity of dominant permeable intervals 4

No data

Pinicon Ridge Formation

60

Casing bottom

50 75

Interpretation

Spillville Formation

Ambient flow

93% Pumped flow

75 100

Fine clastic component

Carbonate component

Figure 48. Results of a borehole geophysical investigation at Austin, Minnesota (Paillet and others, 2000). Flowmeter data were collected under ambient and 2 gallons per minute pumping conditions. Note that 93 percent of the contribution to the borehole is from a discrete horizon within the Spillville Formation. Ambient downflow from this horizon moves out of the borehole into secondary pores just a few feet below. Pumping produced no measurable drawdown. Unique number 613746. See Figure 1 for location, and Figure 5 for an explanation of flowmeter logs.

those calculated on the assumption of intergranular flow only. The effectiveness and scale at which bedrock layers with low matrix permeability can provide confinement is diminished in shallow bedrock conditions. Such units are ubiquitously fractured, and have a relatively great range in hydraulic conductivity in shallow conditions of burial. Individual hydrostratigraphic units that are proven to serve as confining units in deep bedrock conditions are in places hydraulically breached by fractures and dissolution features where they occur close to the bedrock surface (Alexander and Lively, 1995; Lindgren, 2001).

92

RECOMMENDATIONS Our synthesis of the hydrogeologic attributes of Paleozoic bedrock in southeastern Minnesota has revealed a number of topics that warrant further investigation. For example, among the major aquifers we have defined, the upper and lower Mt. Simon aquifers and all of the aquifers above the Decorah Shale are relatively poorly characterized. In addition, virtually all of the confining units need to be rigorously tested in shallow bedrock settings to adequately determine to what degree fractures and dissolution features compromise their inferred confining capability. The presence, abundance, and hydraulic significance of systematic fractures and bedding-plane fractures are

topics that are particularly poorly understood. Such features may affect basin-scale hydraulics, particularly in layers that have a low matrix permeability. Researchers are encouraged to analyze both new and existing data in the context of our new hydrogeologic framework to further its development. Recently developed borehole geophysical techniques such as flowmeter, video, caliper, and acoustic televiewer logs have proven to be particularly useful for recognition of major ground-water conduits and low permeability confining units in subsurface conditions at both local and regional scales. Models of ground-water flow in southeastern Minnesota should take into consideration the fact that flow in some aquifers and confining units occurs chiefly along discrete intervals of high hydraulic conductivity such as bedding-plane fractures. Such flow is increasingly recognized to be important in aquifers and confining units that were formerly treated as more or less homogeneous bodies (for example Gianniny and others, 1996; Michalski and Britton, 1997; Morin and others, 1997; Swanson, 2001). Prior to recognition of preferential flow paths, conductivity for such units was commonly calculated on the basis of standard aquifer tests and with the assumption that the entire thickness of a hydrogeologic unit contributes equally to a borehole. Travel times of ground water calculated under such an assumption are known to be orders of magnitude slower than travel times measured by tracer experiments (for example Bradbury and others, 2000). Wellhead protection plans and other environmental management strategies in which ground-water travel times are of critical importance should take this into consideration. Potentiometric and water chemistry maps should be constructed within the context of the hydrogeologic framework presented in this report. Existing regionalscale maps are of limited value because they were constructed at scales that allow small but potentially important head differences to go unrecognized, and were developed under the incorrect premise that lithostratigraphic units directly correspond to hydrogeologic units, and that apparent similarities in static water levels in adjacent units demonstrates hydraulic connection. Furthermore, previously published potentiometric maps are based in part on measurements of static water levels in boreholes that we now know expose multiple aquifers and confining units.

ACKNOWLEDGMENTS The framework for this report, and its inspiration, are attributed to Bea Hoffman of the Southeast Minnesota Water Resources Board. Her recognition

93

that a comprehensive synthesis of hydrogeologic data for southeastern Minnesota was needed to produce improved wellhead protection plans led early development of this report. The Southeast Minnesota Water Resources Board was funded to initiate such a synthesis by the Minnesota Board of Water Resources through two "Challenge Grants" from its local water resources protection and management program. Much of the information on the Prairie du Chien Group and Jordan Sandstone described in this report was compiled by the Minnesota Geological Survey as part of two projects approved by the Minnesota Legislature for funding as recommended by the Legislative Commission on Minnesota Resources: the 1989 project entitled "Geologic Factors Affecting the Sensitivity of the Prairie du Chien–Jordan Aquifer," and the 1999 project supported specifically through the Minnesota Environment and Natural Resources Trust Fund entitled "Groundwater Flow in the Prairie du Chien Aquifer, Southeastern Minnesota." Additionally, those projects provided part of the funds used to purchase borehole geophysical equipment that has been essential in characterizing the hydrogeologic attributes of all parts of the Paleozoic stratigraphic section. An ongoing investigation of the hydrogeologic attributes of the Franconia Formation and Ironton and Galesville Sandstones funded by the Metropolitan Council also provided important borehole flowmeter data used in this report. A number people associated with local and state government agencies have been particularly helpful over the past 10 years in providing funds and logistical support necessary to collect much of the information used in this report. They include: Terry Lee, Olmsted County Planning Department; Doug Rovang and Barb Huberty, Rochester Public Works; Donna Rasmussen, Fillmore County Soil and Water Conservation District; Bill Buckley, Mower County Environmental Health; Daryl Franklin, Mower County Planning and Zoning; Ross Dunsmore, Winona County Environmental Services; Jim Lundy, Sandeep Burman, and Larry Landherr, Minnesota Pollution Control Agency; Laurel Reeves, Minnesota Department of Natural Resources, Division of Waters; and Bruce Olsen and Patrick Sarafolean, Minnesota Department of Health.

Mahood's Valley

Dubuque Formation

A.

vertical exaggeration x35

Buried valley

Surface stream

York Blind Valley Stream sink

Topographic divide

Figure 49. Characteristic features of the Galena– Spillville karst system highlighting major, dissolution enlarged conduits. Conduit flow is commonly measured at rates as rapid as miles per day, and large cavern systems are well known. Note that the shaly strata of the upper Dubuque Formation do not provide effective confinement everywhere because they are cut by systematic fractures. Figure is based on dye-trace studies and cave explorations (Alexander and Lively, 1995).

Galena Group– Stewartville Formation Galena Group– Prosser Limestone Galena Group– Cummingsville Formation

Dubuque Formation

Maquoketa Formation

Spillville Formation

B.

Galena Group– Stewartville Formation Galena Group– Prosser Limestone Galena Group– Cummingsville Formation Decorah Shale Platteville and Glenwood Formations St. Peter Sandstone vertical exaggeration x11

94 Surficial deposits

Carbonate component

Fine clastic component

Coarse clastic component

EXPLANATION

Surface basin divides Sinkhole

Subsurface basin divide

Ground and surface water flow direction

Spring

Fractures and dissolution features

Sinkholes

0

0

B

A' B'

FILLMORE COUNTY A

Upper Iowa River

Odessa spring

B'

Watson Creek

Stagecoach spring

LOCATION OF CROSS-SECTIONS

1 mile

0.5 mile

A'

95

96

97

REFERENCES Alexander, E.C., Jr., Alexander, S.C., and Barr, K.D., 2001, Dye tracing to Camp Coldwater Spring, Minneapolis, Minnesota: Midwest Ground Water Conference, 46th Annual Meeting, Program and Abstracts, p. 33. Alexander, E.C., Jr., and Book, P.R., 1984, Altura, Minnesota lagoon collapses, in Beck, B.F., ed., Sinkholes: Their geology, engineering, and environmental impact: Rotterdam, A.A. Balkema, p. 311-318. Alexander, E.C., Jr., Broberg, J.S., Kehren, A.R., Graziani, M.M., and Turri, W.L., 1993, Bellechester Minnesota lagoon collapses, in Beck, B.F., ed., Applied karst geology: Proceedings of the Fourth Multidisciplinary Conference on Sinkholes and the Engineering and Environmental Impacts of Karst, Panama City, Florida, 25-27 January: Rotterdam, A.A. Balkema, p. 63-72. Alexander, E.C., Jr., Green, J.A., Alexander, S.C., and Spong, R.C., 1996, Spring sheds, pl. 9 of Falteisek, J., project supervisor, Geologic atlas of Fillmore County, Minnesota: Minnesota Department of Natural Resources, Division of Waters County Atlas C-8, pt. B, 4 pls., scale 1:100,000. Alexander, E.C., Jr., and Lively, R.S., 1995, Karst— Aquifers, caves and sinkholes, in Lively, R.S., and Balaban, N.H., eds., Text supplement to the geologic atlas of Fillmore County, Minnesota: Minnesota Geological Survey County Atlas C-8, pt. C, p. 10-18. Barr Engineering, 1976, Soil and groundwater investigation, coal tar distillation and wood preserving site, St. Louis Park, Minnesota: Phase I Report, 29 p. ———1983, Site characterization study and remedial action plan, General Mills solvent disposal site, June, 1983: On file at the Minnesota Pollution Control Agency, 17 p. ———1986, Response action plan, the Nutting Company Faribault site: On file at the Minnesota Pollution Control Agency, 24 p. ———1991, Magnolia member aquifer pump test report, remedial action design plan, 2010 East Hennepin Avenue site [General Mills], Minneapolis, Minnesota: On file at the Minnesota Pollution Control Agency, 13 p. ———1994, 1993 Annual report, East Hennepin Avenue site [General Mills], Minneapolis, Minnesota, January 1994: On file at the Minnesota Pollution Control Agency, 17 p. ———1996, Dakota County groundwater model

98

summary report: 90 p. ———2000, 1999 Annual report, East Hennepin Avenue site [General Mills]: On file at the Minnesota Pollution Control Agency, 21 p. Barr, K.D., 2001, Field techniques for delineation of fracture flow system, Platteville Formation, Minneapolis, Minnesota: Midwest Ground Water Conference, 46th Annual Meeting, Program and Abstracts, p. 34. Berg, R.R., 1954, Franconia Formation of Minnesota and Wisconsin: Geological Society of America Bulletin, v. 65, no. 9, p. 857-882. Bradbury, K.R., 2001, Springs, sinks, and flowing wells: Hydrogeology at the urban fringe: Midwest Ground Water Conference, 46th Annual Meeting, Field Trip Guide, 16 p. Bradbury, K.R., Rayne, T.W., and Muldoon, M.A., 2000, Field verification of capture zones for municipal wells at Sturgeon Bay, Wisconsin: Wisconsin Geological and Natural History Survey, Informal Final Report to the Wisconsin Department of Natural Resources, 29 p. Bradbury, K.R., and Rothschild, E.R., 1985, A computerized technique for estimating the hydraulic conductivity of aquifers from specific capacity data: Ground Water, v. 23, no. 2, p. 240-245. Braun Intertec, 1992, Ray's North Star Truck Stop site remediation investigation report to the Minnesota Pollution Control agency: Project no. CMKX-910240 A.11. Brick, G., 1997, Along the Great Wall: Mapping the springs of the Twin Cities: Minnesota Ground Water Association, v. 16, no. 1, p. 1-6. Buchmiller, R., Gaillot, G., and Soenksen, P.J., 1985, Water resources of north-central Iowa: Iowa Geological Survey Water Atlas 7, 93 p. Camp, Dresser and McKee, 1991, Phase 1A: Multi-point source groundwater remedial investigation, New Brighton/Arden Hills, Minnesota: Minnesota Pollution Control Agency Report 90-00007, v. 1, 453 p. Campion, M., 1997, Bedrock hydrogeology, pl. 8 of Falteisek, J., ed., Geologic atlas of Rice County, Minnesota: Minnesota Department of Natural Resources, Division of Waters County Atlas C-9, pt. B, 3 pls., scale 1:100,000. ———2002, Bedrock hydrogeology, pl. 8 of Falteisek, J., project manager, Geologic atlas of Mower County, Minnesota: Minnesota Department of Natural Resources, Division of Waters County Atlas C-11, pt. B, scale 1:100,000.

99

Carlson, D.A., and Taylor, R.W., 1999, Evidence that the Mt. Simon Formation has several hydraulic layers in southeastern Wisconsin: Geological Society of America, North-Central Section Meeting, Abstracts with Programs, p. A-8. Curry, B.B., Graese, A.M., Hasek, M.J., Vaiden, R.C., Bauer, R.A., Schumacher, D.A., Norton, K.A., and Dixon, W.G., Jr., 1988, Geological–geotechnical studies for siting the superconducting super collider in Illinois: Results of the 1986 test drilling program: Illinois State Geological Survey Environmental Geology Notes 122, 108 p. Dalgleish, J.B., and Alexander, E.C., Jr., 1984, Sinkholes and sinkhole probability, pl. 5 of Balaban, N.H., and Olsen, B.M., eds., Geologic atlas of Winona County, Minnesota: Minnesota Geological Survey County Atlas C-2, 8 pls., scale 1:100,000. Delin, G.N., 1991, Hydrogeology and simulation of ground-water flow in the Rochester area, southeastern Minnesota, 1987-88: U.S. Geological Survey Water-Resources Investigations Report 904081, 101 p. Delin, G.N., and Almendinger, J.E., 1993, Delineation of recharge areas for selected wells in the St. Peter– Prairie du Chien–Jordan aquifer, Rochester, Minnesota: U.S. Geological Survey Water-Supply Paper 2397, 39 p. Delin, G.N., and Woodward, D.G., 1984, Hydrogeologic setting and the potentiometric surfaces of regional aquifers in the Hollandale Embayment, southeastern Minnesota, 1970-80: U.S. Geological Survey Water-Supply Paper 2219, 56 p. Delta Environmental Consultants, Inc., 1992, Annual groundwater monitoring report to the Minnesota Pollution Control Agency for Ray's North Star Truck Stop, Lakeland, Minnesota: Delta Report 11-92-002, on file at the Minnesota Pollution Control Agency. ———1995, Supplemental remedial investigation report: Amoco Oil Company Spring Valley terminal, Spring Valley, Minnesota: Delta Report A091-088, on file at the Minnesota Pollution Control Agency, 25 p. ———1998, Amoco Oil Company Spring Valley terminal, Spring Valley, Minnesota: MPCA leak number 372: Delta Report 2930-504G, on file at the Minnesota Pollution Control Agency, 13 p. ———2002, Annual monitoring report calendar year 2001, BP terminal, Spring Valley, Minnesota: MPCA leak number 372: Delta Project 29305, on file at the Minnesota Pollution Control Agency, 30 p.

Des Moines Water Works, 1995, Aquifer storage and recovery demonstration project, Phase II: Demonstration test results: Informal report submitted by CH2MHILL to the Des Moines Water Works Department, unpaginated. Donahue and Associates, Inc., 1991, Olmsted County dye-trace investigations [prepared for Olmsted County]—Final report, volume 1: On file at the Minnesota Pollution Control Agency, 155 p. Eaton, T.T., Bradbury, K.R., Anderson, M.P., and Wang, H.F., 2000, Evaluation of the hydraulic properties of a sparsely fractured confining unit: Importance of fracture connectivity: Geological Society of America, Annual Meeting, Reno, Nev., Abstracts with Program, p. 64. Eberts, S.M., and George, L.L., 2000, Regional groundwater flow and geochemistry in the Midwestern Basins and Arches aquifer system in parts of Indiana, Ohio, Michigan, and Illinois (7th ed.): U.S. Geological Survey Professional Paper 1423C, 103 p. Eder and Associates, 1997, 1996 annual monitoring report for Spring Grove, Minnesota: Response action plan: Project 536-01, on file at the Minnesota Pollution Control Agency. ENSR International, 1991, Drift–Platteville aquifer, supplemental remedial investigation for the Reilly Tar and Chemical Corporation site, St. Louis Park, Minnesota: On file at the Minnesota Pollution Control Agency, 41 p. ERT, 1987, Drift–Platteville aquifer source and gradient control well construction and aquifer test report: On file at the Minnesota Pollution Control Agency. Ferguson, H.F., 1967, Valley stress release in the Allegheny Plateau: Bulletin of the International Association of Engineering Geology, v. 4, no. 1, p. 63-71. Fetter, C.W., 1988, Applied hydrogeology (2nd ed.): Columbus, Ohio, Merrill Publishing Company, 592 p. Freeze, R.A., and Cherry, J.A., 1979, Groundwater: Englewood Cliffs, N.J., Prentice-Hall, Inc., 604 p. Gianniny, G.L., Muldoon, M.A., Simo, J.A., and Bradbury, K.R., 1996, Correlation of highpermeability zones with stratigraphic features in the Silurian dolomite, Sturgeon Bay, Wisconsin: Wisconsin Geological and Natural History Survey Open-File Report 1996-7, 102 p., 1 pl. Goldstrand, P.M., and Shevenell, L.A., 1997, Geologic controls on porosity development in the Maynardville Limestone, Oak Ridge, Tennessee:

Environmental Geology, v. 31, no. 3/4, p. 248-258. Graese, A.M., Bauer, R.A., Curry, B.B., Vaiden, R.C., Dixon, W.G., Jr., and Kempton, J.P., 1988, Geological–geotechnical studies for siting the superconducting super collider in Illinois: Regional summary: Illinois State Geological Survey Environmental Geology Notes 123, 100 p. Green, J.A., Mossler, J.H., Alexander, S.C., and Alexander, E.C., Jr., 1997, Karst hydrogeology of Le Roy Township, Mower County, Minnesota: Minnesota Geological Survey Open-File Report 972, 2 pls., scale 1:24,000. Hall, C.W., Meinzer, O.E., and Fuller, M.L., 1911, Geology and underground waters of southern Minnesota: U.S. Geological Survey Water-Supply Paper 256, 406 p. Hatcher, R.D., Jr., Lemiszki, P.J., Dreier, R.B., Ketelle, R.H., Lee, R.R., Leitzke, D.A., McMaster, W.M., Foreman, J.L., and Lee, S.Y., 1992, Status report on the geology of the Oak Ridge Reservation: Oak Ridge National Lab Report ORNL/TM-12074, 260 p. Helgeson, D.E., and Aydin, A., 1991, Characteristics of joint propagation across layer interfaces in sedimentary rocks: Journal of Structural Geology, v. 13, no. 8, p. 897-911. Hoffman, R.S., and Alexander, E.C., Jr., 1998, Evidence of karstic features in the Twin Cities, Minnesota: A summary review of the Ordovician Platteville Formation: Minnesota Water '98, Protecting Minnesota's Water Supplies, Minneapolis, Minn., Collection of Abstracts, p. 111. Hoholick, J.D., Metarko, T., and Potter, P.E., 1984, Regional variations of porosity and cement: St. Peter and Mount Simon Sandstones in the Illinois Basin: American Association of Petroleum Geologists Bulletin, v. 68, no. 6, p. 753-764. Horick, P.J., 1989, Water resources of northeast Iowa: Iowa Department of Natural Resources, Geological Survey Bureau Water Atlas 8, 133 p. Howard, J.H., and Nolen-Hoeksema, R.C., 1990, Description of natural fracture systems for quantitative use in petroleum geology: American Association of Petroleum Geologists Bulletin, v. 74, no. 2, p. 151-162. Hurley, N.F., and Swager, D., 1991, A classic-fracturecontrolled dolomite reservoir; Albion-Scipio Trend, Michigan: American Association of Petroleum Geologists Bulletin, v. 75, no. 11, p. 1806. Jannik, N.O., Alexander, E.C., Jr., and Landherr, L.J., 1992, The sinkhole collapse of the Lewiston,

100

Minnesota waste water treatment facility lagoon, in Quinlan, J., and Stanley, A., eds., Proceedings of the 3 rd National Ground Water Association conference on hydrogeology, ecology, monitoring, and management of ground water in karst terranes: Dublin, Ohio, Water Well Journal Publishing Company, p. 715-724. Kanivetsky, R., 1978, Hydrogeologic map of Minnesota, bedrock hydrogeology: Minnesota Geological Survey State Map S-2, 2 pls., scale 1:500,000. ———1984, Bedrock hydrogeology, pl. 4 of Balaban, N.H., and Olsen, B.M., eds., Geologic atlas of Winona County, Minnesota: Minnesota Geological Survey County Atlas C-2, 8 pls., scale 1:200,000. ———1988, Bedrock hydrogeology, pl. 5 of Balaban, N.H., ed., Geologic atlas of Olmsted County, Minnesota: Minnesota Geological Survey County Atlas C-3, 9 pls., scale 1:100,000. ———1989, Determination of hydraulic parameters of aquifers and confining beds: Hydrological Science and Technology, v. 4, no. 1-4, p. 30-37. Kanivetsky, R., and Cleland, J.M., 1990, Hydrogeology, pl. 5 of Swanson, L., and Meyer, G.N., eds., Geologic atlas of Washington County, Minnesota: Minnesota Geological Survey County Atlas C-5, 7 pls., scales 1:150,000 and 1:250,000. ———1992, Bedrock hydrogeology, pl. 7 of Meyer, G.N., and Swanson, L., eds., Geologic atlas of Ramsey County, Minnesota: Minnesota Geological Survey County Atlas C-7, 10 pls., scales 1:48,000 and 1:150,000. Kanivetsky, R., and Palen, B., 1982, Hydrogeology of Scott County, pl. 6 of Balaban, N.H., and McSwiggen, P.L., eds., Geologic atlas of Scott County, Minnesota: Minnesota Geological Survey County Atlas C-1, 6 pls. Keeler, R.R., 1997, Modeling groundwater flow in a fractured-karst aquifer in the Big Spring basin, Iowa: Iowa City, M.S. thesis, University of Iowa, unpaginated. Kempton, J.P., Bauer, R.A., Curry, B.B., Dixon, W.G., Jr., Graese, A.M., Reed, P.C., and Vaiden, R.C., 1987, Geological–geotechnical studies for siting the superconducting super collider in Illinois: Results of the spring 1985 test drilling program: Illinois State Geological Survey Environmental Geology Notes 120, 88 p. Klimchouk, A.B., and Ford, D.C., 2000, Types of karst and evolution of hydrogeologic setting, in Klimchouk, A.B., Ford, D.C., Palmer, A.N., and Dreybrodt, W., eds., Speleogenesis evolution of karst aquifers: Huntsville, Ala., National

Speleological Society, p. 45-53. Levorsen, A.I., 1967, Geology of petroleum (2nd ed.): San Francisco, W.H. Freeman, 724 p. Libra, R.D., and Hallberg, G.R., 1985, Hydrogeologic observations from multiple core holes and piezometers in the Devonian-carbonate aquifers in Floyd and Mitchell Counties, Iowa: Iowa Department of Natural Resources, Geological Survey Bureau Open-File Report 85-2, p. 1-20. Liesch, B.A., 1973, Groundwater investigation for the Minnesota Highway Department at the Minnehaha Park tunnel: Minneapolis, Minn., State Project no. 2724-78, 131 p. Lindgren, R.J., 1995, Hydrogeology and groundwater flow of the drift and Platteville aquifer system, St. Louis Park, Minnesota: U.S. Geological Survey Water-Resources Investigations Report 94-4204, 79 p. ———2001, Ground-water recharge and flowpaths near the edge of the Decorah–Platteville–Glenwood confining unit, Rochester, Minnesota: U.S. Geological Survey Water-Resources Investigations Report 00-4215, 41 p. Liszkowski, J., 1993, The effects of Pleistocene icesheet loading-deloading cycles on the bedrock structure of Poland, in Zuchiewicz, W., ed., Selected aspects of the neotectonics of Poland: Folia Quaternaria, v. 64, p. 7-23. Lorenz, J.C., and Finley, S.J., 1991, Regional fractures II, fracturing of Mesaverde reservoirs in the Piceance Basin, Colorado: American Association of Petroleum Geologists Bulletin, v. 75, no. 11, p. 1738-1757. Madsen, E.L., and Norvitch, R.F., 1979, Hydraulic properties and tunneling constraints, in Norvitch, R.F., and Walton, M.S., eds., Geology and hydrogeologic aspects of tunneling in the Twin Cities area, Minnesota: U.S. Geological Survey Miscellaneous Investigations Series Map I-1157, 1 pl. McBride, E.F., Picard, M.D., and Folk, R.L., 1994, Oriented concretions, Ionian Coast, Italy; evidence of groundwater flow direction: Journal of Sedimentary Research, sec. A: Sedimentary Petrology and Processes, v. 64, no. 3, p. 535-540. Michalski, A., and Britton, R., 1997, The role of bedding fractures in the hydrogeology of sedimentary bedrock—Evidence from the Newark Basin, New Jersey: Ground Water, v. 35, no. 2, p. 318-327. Miller, R.T., 1984, Anisotropy in the Ironton and Galesville Sandstones near a thermal-energy-storage well, St. Paul, Minnesota: Ground Water, v. 22,

101

no. 5, p. 532-537. Miller, R.T., and Delin, G.N., 1993, Field observations, preliminary model analysis, and aquifer thermal efficiency: U.S. Geological Survey Professional Paper 1530-A, 55 p. Minnegasco Underground Gas Storage Project (MUGSP), 1980, Unpublished reports by the Minnesota Gas Company on the underground gas storage project: On file at the Minnesota Geological Survey. Morin, R.H., Carleton, G.B., and Poirier, S., 1997, Fractured-aquifer hydrogeology from geophysical logs: The Passaic Formation, New Jersey: Ground Water, v. 35, no. 2, p. 328-338. Moerner, N.A., 1978, Faulting, fracturing, and seismicity as functions of glacio-isostasy in Fennoscandia: Geology, v. 6, no. 1, p. 41-45. Mossler, J.H., 1987, Paleozoic lithostratigraphic nomenclature for Minnesota: Minnesota Geological Survey Report of Investigations 36, 36 p., 1 pl. ———1990, Bedrock geology, pl. 2 of Balaban, N.H., and Hobbs, H.C., eds., Geologic atlas of Dakota County, Minnesota: Minnesota Geological Survey County Atlas C-6, 9 pls., scale 1:100,000. ———1992, Sedimentary rocks of Dresbachian age (Late Cambrian), Hollandale Embayment, southeastern Minnesota: Minnesota Geological Survey Report of Investigations 40, 71 p. ———1995a, Bedrock geology, pl. 2 of Mossler, J.H., project manager, Geologic atlas of Fillmore County, Minnesota: Minnesota Geological Survey County Atlas C-8, pt. A, 5 pls., scale 1:100,000. ———1995b, Bedrock geology, pl. 2 of Hobbs, H.C., project manager, Geologic atlas of Rice County, Minnesota: Minnesota Geological Survey County Atlas C-9, pt. A, 6 pls., scale 1:100,000. ———1998, Bedrock geology, pl. 2 of Mossler, J.H., project manager, Geologic atlas of Mower County, Minnesota: Minnesota Geological Survey County Atlas C-11, pt. A, 6 pls., scale 1:100,000. ———2001, Bedrock geology, pl. 2 of Runkel, A.C., project manager, Geologic atlas of Wabasha County, Minnesota: Minnesota Geological Survey County Atlas C-14, pt. A, 7 pls., scale 1:100,000. Mossler, J.H., and Bloomgren, B.A., 1990, Bedrock geology, pl. 2 of Swanson, L., and Meyer, G.N., eds., Geologic atlas of Washington County, Minnesota: Minnesota Geological Survey County Atlas C-5, 7 pls., scale 1:100,000. ———1992, Bedrock geology, pl. 2 of Meyer, G.N., and Swanson, L., eds., Geologic atlas of Ramsey

County, Minnesota: Minnesota Geological Survey County Atlas C-7, pt. A, 10 pls., scale 1:100,000. Mossler, J.H., and Book, P.R., 1984, Bedrock geology, pl. 2 of Balaban, N.H., and Olsen, B.M., eds., Geologic atlas of Winona County, Minnesota: Minnesota Geological Survey County Atlas C-2, 8 pls., scale 1:100,000. Mossler, J.H., and Tipping, R.G., 2000, Bedrock geology and structure of the seven-county Twin Cities Metropolitan area, Minnesota: Minnesota Geological Survey Miscellaneous Map M-104, scale 1:125,000. Narr, W., and Suppe, J., 1991, Joint spacing in sedimentary rocks: Journal of Structural Geology, v. 13, no. 9, p. 1037-1048. Nelson, C.H., 1956, Upper Cambrian Croixan stratigraphy, Upper Mississippi Valley: Geological Society of America Bulletin, v. 67, no. 2, p. 165183. Nicholas, J.R., Sherrill, M.G., and Young, H.L., 1987, Hydrogeology of the Cambrian–Ordovician aquifer system at a test well in northeastern Illinois: U.S. Geological Survey Water-Resources Investigations Report 84-4165, 30 p. Nichols, T.C., Jr., 1980, Rebound, its nature and effect on engineering works: Quarterly Journal of Engineering Geology, London, v. 13, p. 133-152. Norvitch, R.F., Ross, T.G., and Brietkrietz, A., 1973, Water resources outlook for the Minneapolis–St. Paul metropolitan area, Minnesota: U.S. Geological Survey Open-File Report 73-203, 197 p. Norvitch, R.F., and Walton, M.S., eds., 1979, Geologic and hydrologic aspects of tunneling in the Twin Cities area, Minnesota: U.S. Geological Survey Miscellaneous Investigations Series I-1157, 6 pls. Olsen, B.M., 1982, Bedrock geology, pl. 5 of Balaban, N.H., and McSwiggen, P.L., eds., Geologic atlas of Scott County, Minnesota: Minnesota Geological Survey County Atlas C-1, 6 pls., scale 1:100,000. ———1988a, Bedrock geology, pl. 2 of Balaban, N.H., ed., Geologic atlas of Olmsted County, Minnesota: Minnesota Geological Survey County Atlas C-3, 9 pls., scale 1:100,000. ———1988b, Depth to bedrock and bedrock topography, pl. 4 of Balaban, N.H., ed., Geologic atlas of Olmsted County, Minnesota: Minnesota Geological Survey County Atlas C-3, 9 pls., scale 1:100,000. Olsen, B.M., and Bloomgren, B.A., 1989, Bedrock geology, pl. 2 of Balaban, N.H., ed., Geologic atlas of Hennepin County, Minnesota: Minnesota Geological Survey County Atlas C-4, 9 pls., scale

102

1:100,000. Paillet, F.L., Lundy, J., Tipping, R.G., Runkel, A.C., Reeves, L., and Green, J.A., 2000, Hydrogeologic characterization of six sites in southeastern Minnesota using borehole flowmeters and other geophysical logs: U.S. Geological Survey WaterResources Investigation Report 00-4142, 33 p. Price, N.J., and Cosgrove, J.W., 1990, Analysis of geological structures: Cambridge University Press, 502 p. RMT, Inc., 1992, Remedial investigation report: Olmsted County sanitary landfill, Oronoco Township, Minnesota: On file at the Minnesota Pollution Control Agency, 185 p. Rowden, R.D., and Libra, R.D., 1990, Hydrogeologic observations from bedrock monitoring well nests in the Big Spring Basin, Iowa: Iowa Department of Natural Resources, Geological Survey Bureau Open-File Report 90-1, 27 p. Ruhl, J.F., 1995, Relation of fracture orientation to linear terrain features, anisotropic transmissivity, and seepage to streams in the karst Prairie du Chien Group, southeastern Minnesota: U.S. Geological Survey Water-Resources Investigations Report 944146, 42 p. Runkel, A.C., 1996a, Bedrock geology of Houston County, Minnesota: Minnesota Geological Survey Open-File Report 96-4, 13 p., 3 pls., scale 1:100,000. ———1996b, Geologic investigations applicable to ground-water management, Rochester metropolitan area, Minnesota: Minnesota Geological Survey Open-File Report 96-1, 30 p., 7 pls., scale 1:24,000. ———1998, Bedrock geology, pl. 2 of Setterholm, D.R., project manager, Geologic atlas of Goodhue County, Minnesota: Minnesota Geological Survey County Atlas C-12, pt. A, 6 pls., scale 1:100,000. ———1999, Hydrogeologic attributes of Paleozoic strata in southeastern Minnesota: Part 1— Hydrostratigraphic attributes: Minnesota Geological Survey informal report to the Southeast Minnesota Water Resources Board, 48 p. ———2000, Hydrogeologic attributes of Paleozoic strata in southeastern Minnesota: Part 2—Hydraulic attributes: Minnesota Geological Survey informal report to the Southeast Minnesota Water Resources Board, 35 p. Runkel, A.C., Kanivetsky, R., and Tipping, R.G., 1999, Hydraulic and hydrostratigraphic characteristics of the Prairie du Chien–Jordan aquifer system in the Twin Cities Metropolitan area, Minnesota: Informal report submitted to the Minnesota Department of

Health. Runkel, A.C., and Tipping, R.G., 1998, Stratigraphy and hydrogeology of Paleozoic rocks of southeastern Minnesota: Institute on Lake Superior Geology, 44th Annual Meeting, Minneapolis, Minn., Proceedings, v. 44, Field Trip Guidebook, pt. 2, p. 101-130. Ryder, R.T., 1996, Fracture patterns and their origin in the upper Devonian Antrim Shale gas reservoir of the Michigan Basin: A review: U.S. Geological Survey Open-File Report 96-23, 30 p. Sasowsky, I.D., and White, W.B., 1994, The role of stress-release fracturing in the development of cavernous porosity in carbonate aquifers: Water Resources Research, v. 30, no. 12, p. 3523-3530. Schoenberg, M.E., 1990, Effects of present and projected ground-water withdrawals on the Twin Cities aquifer system, Minnesota: U.S. Geological Survey Water-Resources Investigations Report 904001, 165 p. Seaber, P.R., 1988, Hydrostratigraphic units, in Back, W., Rosenshein, J.S., and Seaber, P.R., eds., Hydrogeology: Boulder, Colo., Geological Society of America, The Geology of North America, v. O2, p. 9-14. Setterholm, D.R., Runkel, A.C., Cleland, J.M., Tipping, R.G., Mossler, J.M., Kanivetsky, R., and Hobbs, H.C., 1991, Geologic factors affecting the sensitivity of the Prairie du Chien–Jordan aquifer: Minnesota Geological Survey Open-File Report 915, 29 p., 12 pls. Shevenell, L.A., and Goldstrand, P.M., 1997, Geochemical and depth controls on microporosity and cavity development in the Maynardville Limestone: Implications for groundwater: Cave and Karst Science, v. 24, no. 3, p. 127-136. Sloan, R.E., and Austin, G.S., 1966, Geologic map of Minnesota, St. Paul sheet: Minnesota Geological Survey Geologic Map of Minnesota, scale 1:250,000. Snow, D.T., 1968, Rock fracture spacings, openings, and porosities: Proceedings of the American Society of Civil Engineers, v. 94, p. 73-91. Solomon, D.K., Moore, G.K., Toran, L.E., Dreier, R.B., and McMaster, W.M., 1992, Status report: A hydrologic framework for the Oak Ridge Reservation: Oak Ridge National Lab report ORNL/TM-12026, 116 p. Spong, R.C., 1980, The Crawlway Cave, in Alexander, E.C., Jr., ed., An introduction to caves of Minnesota, Iowa, and Wisconsin: National Speleological

103

Society Convention Guidebook 21, p. 33-42. Squillace, P.J., 1979, The geology of the New Richmond Member of the Shakopee Formation (Lower Ordovician), upper Mississippi Valley: Duluth, Minn., University of Minnesota Duluth, M.S. thesis, 95 p. Stocks, D.L., 1998, Hydrostratigraphy of the Ordovician Sinnipee Group dolomites, eastern Wisconsin: Madison, Wis., University of Wisconsin-Madison, M.S. thesis, 192 p. Subsurface-Water Glossary Working Group, 1989, Subsurface-water flow and solute transport: Federal glossary of selected terms: U.S. Geological Survey, Office of Water Data Coordination, 34 p. Swanson, S.K., 2001, Hydrogeologic controls on springflow near Madison, Wisconsin: Madison, Wis., University of Wisconsin-Madison, Ph.D. dissertation, 436 p. Tipping, R.G., 1992, An isotopic and chemical study of groundwater flow in the Prairie du Chien and Jordan aquifers, northern Dakota County, Minnesota: Minneapolis, University of Minnesota, M.S. thesis, 117 p. ———1997, Mower County ground-water monitoring study: A report to the Mower County Environmental Health Office: Minnesota Geological Survey informal report, 10 p. Tipping, R.G., Green, J.A., and Alexander, E.C., Jr., 2001, Karst features, pl. 5 of Runkel, A.C., project manager, Geologic atlas of Wabasha County, Minnesota: Minnesota Geological Survey County Atlas C-14, pt. A, 7 pls., scale 1:100,000. Tipping, R.G., and Runkel, A.C., 2001, Ground-water flow in the Prairie du Chien carbonate aquifer, southeastern Minnesota: Midwest Ground Water Conference, 46th Annual Meeting, Program and Abstracts, p. 29. Wall, D.B., and Regan, C.P., 1994, Water quality and sensitivity of the Prairie du Chien–Jordan aquifer in western Winona County, Minnesota: Unpublished report by the Minnesota Pollution Control Agency, Division of Water Quality, St. Paul, Minn., 65 p. Walton, M., Eisenreich, S.J., Holm, N.L., Holm, T.R., Kanivetsky, R., Jirsa, M.A., Lee, H.C., Lauer, J.L., Miller, R.T., Norton, J.L., and Runke, H., 1991, The University of Minnesota Aquifer Thermal Energy Storage (ATES) field test facility-system description, aquifer characterization, and results of short-term test cycles: Pacific Northwest Laboratory Report PNL-7220, UC-202, 295 p. Watts, N.L., 1983, Microfractures in chalks of Albuskjell Field, Norwegian Sector, North Sea: Possible origin

and distribution: American Association of Petroleum Geologists, v. 67, no. 2, p. 201-234. Wenck and Associates, Inc., 1997, Phase II: Detailed site investigation report and Phase II workplan for the hydrogeologic investigation of the proposed Red Wing ash disposal facility expansion: 36 p. Wheeler, B.J., 1993, Groundwater tracing in the Duschee Creek karst basin in southeast Minnesota: Plan B paper submitted to the Geology and Geophysics Department, University of Minnesota, 72 p. Williams, J.H., Carswell, L.D., Lloyd, O.B., and Roth, W.C., 1984, Characterization of groundwater in selected fractured rock aquifers using borehole temperature and flow logs: NWWA/EPA conference of surface and borehole geophysical methods in groundwater investigations: San Antonio, Tex., National Water Well Association, Proceedings. Witthuhn, K.M., and Alexander, E.C., Jr., 1996, Sinkholes and sinkhole probability, pl. 8 of Falteisek, J., project supervisor, Geologic atlas of Fillmore County, Minnesota: Minnesota Department of Natural Resources, Division of Waters County Atlas C-8, pt. B, 4 pls., scale 1:100,000. Witzke, B.J., and Bunker, B.J., 1984, Devonian stratigraphy of north-central Iowa: Iowa Department of Natural Resources, Geological Survey Bureau Open-File Report 84-2, p. 107-138. Worthington, S.R.H., 1999, A comprehensive strategy for understanding flow in carbonate aquifers, in Palmer, A.N., Palmer, M.V., and Sasowsky, I.D., eds., Karst modeling: Karst Waters Institute Special Publication 5, p. 30-37. Wyrick, G.G., and Borchers, J.W., 1981, Hydrologic effects of stress-relief fracturing in an Appalachian valley: U.S. Geological Survey Water-Supply Paper 2177, 51 p. Young, H.L., 1992, Hydrogeology of the Cambrian– Ordovician aquifer system in the northern Midwest, United States: U.S. Geological Survey Professional Paper 1405-B, 99 p. Zanini, L., Novakowski, K.S., Lapcevic, P., Bickerton, G.S., Talbot, C., and Voralek, J., 1998, Regional groundwater flow in the fractured dolomite underlying Smithville, Ontario as inferred from combined hydrogeologic and geochemical measurements: Geological Society of America, Annual Meeting, Abstracts with Programs, v. 30, no. 7, p. 224. Zhang, H., and Kanivetsky, R., 1996, Bedrock hydrogeology, pl. 6 of Falteisek, J., project supervisor, Geologic atlas of Fillmore County,

104

Minnesota: Minnesota Department of Natural Resources, Division of Waters County Atlas C-8, pt. B, 4 pls., scale 1:100,000.

105