Loading...
HomeMy WebLinkAboutEffects of Proposed Mine Expansion on Public Water Supply Wells 07-2004 Q Q Q g Q g Q g Q g g Q g Q Q Q Q Q Q Q Q g Q Q g Q Q Q Q Q o g g g Q Q Q Q Q Q g Q Q Q EFFECTS OF PROPOSED MINE EXPANSION ON PUBLIC WATER SUPPL V WELLS City of Carmel Utilities Carmel, Indiana PREPARED By WHPA, INCORPORATED WITTMAN HYDRO PLANNING ASSOCIATES BLOOMINGTON, INDIANA JULY, 2004 Q Q o o Q Q ~ ~ Q Q o o ~ o o o o o o o o o Q o o o o o o o o o Q o o o o o o o o o o o WHPA Assessment of the Effects of Proposed Mine Expansion at the South Mueller Property on Public Water Supply Wells near Carmel, Indiana 27th July 2004 Prepared by Wittman Hydro Planning Associates, Inc. Bloomington, Indiana o Q Q o Q o Q Q Q Q o o o Q o Q Q o Q o o Q Q Q I~ Q Q Q Q Q Q Q Q Q Q Q Q Q Q o Q Q Q Contents 1 Introduction 1.1 Summary of findings 1 6 2 Hydrogeologic setting 2.1 Geology........................ 2.1.1 Outwash along the White River . . . . . . . 2.1.2 Confined sand and gravel intertill aquifers 2.1.3 Bedrock aquifer. . . . . . . . . . . . . . . 2.2 Hydrology...................... 2.2.1 Surface water and groundwater interactions 2.2.2 Potentiometric data . . . . . . . . . . . . . . . 7 7 8 9 10 10 11 11 3 Model development 3.1 Model code selection . . . . . .. ...... 3.2 Aquifer conceptualization . . . 3.3 Regional boundary conditions ......... 3.4 Local boundary conditions . . . . . . . . . . 3.5 Representing mine operations in the model. . . 3.6 Potentiometric head observations. . . . . . . . 11 11 12 13 17 17 19 "Post Mining" scenario . 20 20 23 26 28 28 31 31 4 Results 4.1 Model calibration . . . . . . . . . 4.2 Analysis of parameter uncertainty 4.3 Reduced recharge analysis 4.4 Scenarios and results . . . . . . . 4.4.1 4.4.2 4.4.3 "Current" scenario . . . . . . "South Mueller Active" scenario . 5 Conclusions 36 References 38 Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q List of Figures 1 Area of interest in Hamilton County, Indiana. . . . . . 2 2 Pumping in the City of Carmel wells from 1988-2002. 3 3 Site location map. . . . . . . . . . . . . . . . . . . . . 5 4 Plan view of the relationships between conceptual aquifer domains. 14 5 Conceptual model cross section. . . . . . . . . . . . . . . . . . . . 15 6 Layout of wells and line sinks in the regional model. ........ 16 7 Aquifer domains and local boundary conditions for the current conditions near Carmel Plant 4. ................... 18 8 Results of calibration and best-fit potentiometric surface. 21 9 Summary calibration statistics from GFLOW. . . . . . . 22 10 Determination of a composite capture zone. . . . . . . . 24 11 Former gravel pit ponds that contribute water to the Plant 4 wellfield, with identifying names.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 25 12 Modeled potentiometric surface and composite 5-year capture zone for the current conditions. .............................. 29 13 Current operations for "reduce-recharge" conditions. ............ 30 14 Modeled potentiometric surface an composite 5-year capture zone for the current conditions with Well 22. .... . . . . . . . . . . . . . . . . . .. 32 15 Modeled potentiometric surface and composite 5-year capture zone when the South Mueller property is actively mined. . . . . . . . . . . . . . . .. 33 16 Change in heads between current conditions and those when the South Mueller property is actively mined. . . . . . . . . . . . . . . . . . . . . ., 34 17 Results for "reduced-recharge" conditions when the South Mueller property is actively mined. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 35 List of Tables 1 Annual production well data for the Plant 4 Well Field (All data reported in million gallons per year). . . . . . . . . . . . . . . . 4 2 Aquifer classification scheme for conceptual model. . . . . . . . . . . . ., 14 3 Parameters for "best-fit" model. .... . . . . . . . . . . . . . . . . . .. 22 4 Water levels in Ponds NW and SW (source: Martin Marietta Materials, Inc.). 27 ii ___ ______..__1.______________ Q Q Q Q Q Q o Q o Q Q Q Q Q o Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q o Q Q o o o 1 Introduction Carmel's municipal drinking water is currently supplied by approximately 20 production wells, most of which are located within a mile of the White River in south-central Hamilton County, Indiana (Figure 1). Some of the wells have been in operation since before 1960 but most of the production comes from the most recent additions to the system at Plants 4 and 5. Most of the wells are screened in sand and gravel along the edge of a thick outwash aquifer that roughly coincides with the flood plain of the White River. The wells vary in both depth and yield. In order to satisfy the pace of residential growth in the area the annual production of the four water plants nearly doubled between 1997 and 2002 (Figure 2). Today, Carmel's average annual production (2000-2003) is approximately 8 million gallons per day (MGD) with Plant 4 pumping and average of 0.7 MGD (Table 1). It is expected that water use will continue to increase rapidly in the future; for example, the proposed annexation of land outside the current corporate boundary would greatly expand the demand for drinking water. Martin Marietta Minerals, Inc. has operated gravel pits and aggregate mines along the southern border of Hamilton County for decades. In the area South of 106th street there are several abandoned pits and ponds as well as a large open pit limestone mine. As growth in the area continues there is increasing demand for gravel and aggregate for roads, bridges, and buildings. Until recently, these operations have coexisted near the water wells without any conflict. However, in order to increase production Martin Marietta has proposed an expansion of its current operations. In their proposal they outline a plan to excavate the gravel above the bedrock surface and then to mine into the bedrock surface as they have done in their existing pit. In addition, they are planning to open sand and gravel pits at new locations over a period of years to satisfy the growing demand for aggregate. There are separate applications for expansion at two properties: the North Mueller prop- erty, which includes the area north of 106th street between Hazel Dell Parkway and Gray Road, and the South Mueller property which is located just south of 106th Street between Hazel Dell Parkway and Gray Road (Figure 3). The proposed mining at North Mueller would include some additional removal of sand and gravel as well as new underground limestone aggregate mining. The proposed mining at the South Mueller property would be done in stages. The first stage would be the removal of the sand and gravel while water removal occurs along the edges. The next phase is to expand the boundaries of the existing open pit to the North to serve the North Indianapolis Plant. As Carmel's role as a regional drinking-water supplier grows, it is critical that the city 1 o g (;) Q g g g Q g g Q g Q g g Q g Q g Q g ,Q g Q g g g Q g g g (i) g g Q Q g g Q g g (i) g g HI 'I ill II 'I 'I HJ~i1ton County I~ ,II 'ii II I il II II Area of Interest + 2 0 2 4 6 8 10 Miles 1""'"- , Figure 1: Area of interest in Hamilton County, Indiana. 2 Q g Q g g Q Q Q Q Q Q Q Q Q g Q Q Q g Q Q Q Q Q g g g Q Q Q g Q Q Q Q Q Q Q Q Q Q Q g g 3000 (j) 2000 c .2 ell G c o 2 '--" 0> c "0.. E ::J 0... 1000 3 2500 1500 500 o * ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ * ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ Figure 2: Pumping in the City of Carmel wells from] 988-2002. ~c~~~~~vcv~~vvv~~~~cv~~~v~ccc~cc~~c~ccc~~ccv Well 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Number 10 102.94 206.23 226.40 223.85 135.71 181.69 190.12 61.01 24.59 182.31 175.85 273.57 217.26 212.47 124.82 11 88.35 206.23 154.67 63.36 64.65 65.24 95.98 42.32 40.25 26.06 142.73 97.95 11.79 29.97 30.86 12 93.88 104.15 72.32 71.87 83.96 65.05 54.70 31.93 57.49 23.93 69.13 37.51 5.79 59.71 64.02 Annual Total 285.17 516.61 453.39 359.08 284.32 311.98 340.80 135.26 122.33 232.30 387.71 409.03 234.84 302.15 219.70 Average Annual Production 306.31 Table 1: Annual production well data for the Plant 4 Well Field (All data reported in million gallons per year). ~ Q Q Q g g Q Q Q Q g Q Q Q Q g Q Q g Q g g g g Q Q Q g Q Q Q ,Q g Q Q g g g g g g Q g g Q 5 + 0.1 0 0.1 0.2 Miles !"""'-_ I LEGEND Wittman Hydro Planning Associates \"~.~o..,..~Y\;I""J1rlllC:;m"b-~ N Roads D Mining Gravel Pits _ Hydrology Figure 3: Site location map. Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q ~ ~ Q ~ Q Q Q Q ~ Q ~ Q Q protects the aquifers from any deterioration in yield. This report was written to answer an important question raised by the recent growth in the area: Will the proposed new mining activity affect Carmel's drinking-water supply? How much do these two activities conflict with one another and how compatible are they? We have used groundwater models to investigate these issues. The models that include regional boundary conditions and flow in the unconsolidated aquifers, with local refinement in the vicinity of the proposed mining works. Our analysis provided information about the marginal change of the new mines that could alter conditions at the closest drinking-water treatment plant (Plant 4). This analysis was done with data provided by the City about the configuration of the gravel deposits in the subsurface, information provided by the Indi- ana Department of Natural Resources from domestic well logs, and standard groundwater modeling tools. The effects of mining operations and reduced recharge were considered together to better understand unknown future conditions. This study is part of a larger ongoing water-supply planning effort being conducted by the Utility that includes subsurface data collection and organization, wellhead protection planning, and water-supply expansion planning. 1.1 Summary of findings There is no single index of the "effects of mining" that captures all of the changes that could take place once the proposed excavations are underway. Instead, we considered two types of impacts: changes in water levels in the aquifer near the Plant 4 well field, and changes in the source water pumped by the well field. The first of these is an indicator of the reduction in yield that results from the competition between the drinking-water wells and around the mine pits. We examined impacts both during and after mining. . Additional mining could reduce groundwater levels at the Plant 4 well field by about 2 - 3 ft while the mines are active. The amount of water pumped by a well is roughly proportional to the draw down in the well. Since the original pumping tests showed about 16ft of drawdown when the wells were pumped at capacity and the modeling predicts an additional 2 - 3 ft of decline, this translates to roughly 15% reduction in total wellfield capacity during mining. . In the 20 years that the proposed mines are active there would be an increase in the amount of drinking water that comes from abandoned gravel pits and ponds. Cur- rently we estimate that approximately one-half of the water pumped at the Plant 4 6 o Q o Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q well field had been in a pit or pond within the last five years. While the mines are active this will increase to more than 70% of all the water in the well field. . With conditions of reduced recharge, the effects of the proposed mining will be to decrease water levels by 3 it compared to reduced recharge under current conditions. With reduced recharge there are changes in water use and stage along the river that may alter the actual conditions at the well field. This is a result of reductions in regional flow from the till aquifers west of Plant 4 and the increasing importance of the White River. 2 Hydrogeologic setting According to the Hydrogeologic Atlas of Aquifers in Indiana [Fenelon and Bobay, 1994] aquifers in the West Fork White River basin are categorized as follows: 1. thick outwash adjacent to major rivers 2. thin outwash aquifers in thick glacial till 3. discontinuous sand and gravel in thin glacial till 4. limestone in bedrock covered by variable layers of till (northwest of Carmel, Indiana) The well fields developed by Carmel Utilities in Hamilton County are typical of commu- nity well fields in central Indiana. Carmel's well fields are located in thicker portions of the sand-and-gravel aquifer west of the West Fork (White River). The wide outwash plains along the White River have relatively large recharge rates due to the highly-permeable sur- ficial deposits and the low, flat-lying landscape. Under natural conditions, recharge into the thick outwash aquifer leaves as "baseflow" in the perennial streams. Part of the water, how- ever, is intercepted by the Carmel wells and is used for drinking-water supplies. In some cases, wells may even induce aquifer recharge from the White River or its tributaries. The interactions between surface water and groundwater is affected by the pumping rates of the wells, the hydraulic properties of the river bed, the distance between the wells and the river, and other factors, including past and present mining activities. 2.1 Geology Hamilton County is in the central part of the glaciated portion of Indiana, near the center of the White River basin. The thickness of unconsolidated deposits varies from less than 20 it 7 Q Q Q Q Q Q Q (;,) Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q o Q Q Q Q Q Q Q to over 500ft [Gray, 1983]. Beneath the unconsolidated section, the bedrock dips towards the Illinois Basin. The bedrock underlying most of the unconsolidated section in the study area is Silurian-Devonian aged carbonate rock. The thickness of the bedrock increases to the west. As described above, the two water resource aquifers tapped by high capacity wells in Hamilton County are; 1) outwash deposits along the larger streams, and 2) thin outwash layers in the till. 2.1.1 Outwash along the White River The White River outwash was deposited by rapidly moving meltwater during the retreat of continental glaciers at the end of the last Ice Age. In Hamilton County the deposits are in a 1 - 3 mi wide channel of coarse sand and gravel in an irregularly-shaped valley, roughly coinciding with the West Fork White River flood plain. Typically, the thick outwash aquifer is overlain by 5 - 10 ft of clayey material and may extend to depths of 110 ft below grade. Outwash aquifer deposits range from fine sands to coarse gravel to cobbles but primar- ily consist of well sorted, coarse-grained sand and gravel [Jones and Henry, 1994]. Thick outwash aquifer deposits also exist west of the river near Martin Marietta's operations. The deposits used by the city wells are ideal as a water supply for two reasons: 1) the gravel and sand have a high transmissivity, allowing high flows towards the wells, and 2) water-supply wells situated near a perennial stream can "induce" recharge from the stream into the aquifer, assuring a more reliable water supply during times of drought and reduced recharge. The aquifer varies in thickness and extent within the Study Area. Based on Carmel Utilities test hole and production well data, the thickness of the sand and gravel ranges from 5 - 100 ft. Within the model area of detail, the average thickness is 80 ft. The gravel thins both east and west of the pre-glacial bedrock valley trough and may be absent, particularly in areas where subsurface bedrock highs exist. While unconfined conditions prevail in the thick outwash aquifer, semi-confined to confined conditions do occur where surficial clay layers and/or significant subsurface clay layers are present. The hydraulic conductivity (K) of the outwash along the West Fork has been estimated at over 340ft/d in some areas based on pump test results [Bailey and Imbrigiotta, 1982]. The saturated thickness of the outwash varies between 50 - 120 ft in the region. The trans- missivity of the aquifer (defined as the product of hydraulic conductivity and saturated thickness) is highest where the aquifer is thickest west of the river along 106th Ave. in southern Hamilton County [Cable et al., 1971]. Previous analyses by the USGS have sug- gested that in Hamilton County this aquifer is capable of producing 30MGD of water with 8 o o o Q ~ o Q Q Q Q Q Q ~ Q Q Q o Q o Q o o Q Q ~ o ~ ~ o o Q ~ o o Q Q o Q ~ ~ Q o o o limited effect on regional water levels. Arihood [1982] reports a range of K values for the thick outwash aquifer (based on aquifer material) to be from 40 - 415ft/d with corresponding values for T ranging from 1,000 - 28,000 I t2 / d. Based on aquifer test pumping results from those Carmel production wells that are completed in the thick outwash aquifer, estimated K values range from 191 - 8651t/d corresponding to T values from 17, 500-38,000It2 /d [WHPA, 2003]. While the range for K and T in this aquifer can be highly variable, typical values for K and T range from 200 - 3001t/d and 16,000- 24,000ft2 /d, respectively. This aquifer is recharged by precipitation and by inflow from intertill aquifers east and west along the perimeter. Water budget studies done by the USGS suggest that infiltration rates into the outwash system along the White River may be up to 12in/yr [Meyer, 1979]. In addition to direct areal recharge into the outwash, the aquifer receives water along the perimeter from the confined intertill aquifers discharging towards the White River. 2.1.2 Confined sand and gravel intertill aquifers Further away from the river, wells are often situated in sand and gravel deposits buried within the glacial till. These intertill aquifers are often discontinuous and, beyond the City of Carmel geodata archive, they have not been mapped at a useful scale. The hydrogeologic investigations done by the city, along with a more recent mapping effort conducted by the Indiana Geological Survey, have provided substantial information about these systems [Brown et al., 1995]. The thicker unconsolidated deposits almost always include poorly connected layers of sand and gravel, which together may form a series of partially connected till aquifers. The distribution of these intertill aquifers were further evaluated by direct analysis of the well log records from the area. Where present, thin outwash aquifers within the till are confined and are commonly relatively thin (average thicknesses range from 5- 20 It). In some parts of the study area there is no significant transmissivity above bedrock. However, the data indicates that in the thicker unconsolidated section the till often includes fairly continuous sand and gravel layers. In general, K values reported for glacial till range from 10-7 - 25 ft / d [Driscoll, 1986] while the transmissivity of the interbedded thin outwash aquifers that occur within the till are typically much higher. It is important to remember that these more permeable sand and gravel layers may be regionally continuous and locally variable. Previous work in the area indicates that there are several "zones," including the average or "regional" value of K for the intertill aquifer, that account for transmissivity of both glacial till and discontinuous, interbedded sand and gravel. 9 o o Q o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o Arihood [1982] reports a range of K values for the confined interbedded thin out- wash aquifers in Hamilton County to be from 200 - 433 ft j d . Assuming average aquifer thicknesses ranging from 5 - 20 ft, corresponding values for transmissivity (T) is between 1000 - 8660 ft2 j d. Where multiple interbedded outwash aquifers are present within the till at any given location, the combined T over the entire section will be larger. 2.1.3 Bedrock aquifer In comparison to the thick outwash and intertill aquifers, the bedrock aquifer is present throughout the model area but infrequently used as a groundwater resource because suf- ficient groundwater supplies are usually encountered within the unconsolidated deposits. Based on well log data, the bedrock is occasionally in direct contact with overlying out- wash deposits but may also be separated from the overlying outwash by relatively thin clay layers (till). In isolated areas, the thickness of the overlying till approaches 50ft. Within the bedrock, well drillers report some variability of hardness, fracturing, and solution features but overall, it is relatively impermeable and does not appear to be an im- portant water resource aquifer [Gillis, 1976]. Martin Marietta's rock quarry near 96thStreet and Hazel Dell has been excavated approximately 140ft into the limestone and beneath approximately 30 - 35 ft of overlying outwash sands and gravel. Quarry operators report little if any seepage of groundwater from the limestone [Gillis, 1976]. For the purpose of this study, groundwater flow in the bedrock formations is negligible. 2.2 Hydrology Of the 37 in of precipitation that that fall each year in Hamilton County, about 12injyr (35%) flow directly into surface waters. The remaining water is returned to the atmosphere as evaporation and transpiration or infiltrates the ground and becomes groundwater. The perennial streams near the well field have their highest average flows in the spring, and seasonal low flow often occurs in the late fall or early winter. USGS stream flow records indicate that in all of the streams, spring low flow (from March through May) is about two times the low flows observed between December and February [Fenelon and Bobay, 1994]. Stream flow is composed of groundwater that is discharging from aquifers and more recent water that is draining off the land surface during and immediately following precipitation events. It has been estimated that, over the entire basin, groundwater discharge into the White River (also referred to as base flow) is about 4inchesjyear or 1 of the total stream flow 10 Q Q Q Q Q Q Q Q ~ o Q o o r;; ~ Q ~ ~ o o ~ ~ o o ~ o ~ o ~ ~ o Q ~ Q o o r;; o o o r;; Q o o [Wittman and Haitjema, 1995]. In the study region, the recharge rates in the outwash will be higher than the basin average, while recharge rates in upland tills will be lower. 2.2.1 Surface water and groundwater interactions The degree of hydraulic connection between surface streams and groundwater can have an important effect on the source areas for wells and the well yields. There have been several studies of the hydraulic connection between the thin outwash aquifers and the streams south of the study area in Marion County [Meyer, 1979, Saul and Robinson, 1989]. The general conclusion of these investigations is that the connection may vary locally but where the aquifer is composed of coarse sand and gravel, there is little resistance between the White River and the alluvial aquifer. 2.2.2 Potentiometric data The water levels in the aquifers result from the hydraulic properties of each aquifer, the pumping rates of the wells, recharge rates (or leakage rates) into or out of the system, and the location of boundary conditions within the system. The water levels in each of the aquifers also varies with time. During prolonged drought (such as the drought experienced in early 1960s) annual maximum water levels declined in observation wells throughout Hamilton County [Meyer et al., 1975]. According to the rating curve for the White River at Nova, the stage during a major drought is 7 - 9 ft lower than the average stage. No water-level measurements were made specifically for this report. Instead, the In- diana Department of Natural Resources water well records and water levels reported by Martin Marietta were used as the calibration targets for flow modeling. 3 Model development This section describes the development of the regional and local model, the conceptual model that was used, the code used, and the calibration data. 3.1 Model code selection The analytic element model (AEM) GFLOW was used because: 1. It is suitable for modeling large domains 2. It can model base flow in surface water 11 Q Q o Q o o o Q Q Q Q o o o o Q Q Q Q Q Q o Q o o o o o o Q o o o o o Q o Q o Q 10 Q o o 3. It can model large transmissivity contrasts accurately, e.g. at gravel pits. A complete discussion of the theory of the analytic element method can be found in Strack [1989] and a discussion of the practical aspects of modeling with the AEM technique is found in Haitjema [1995]. Analytic element models do not discretize the aquifer into a grid or mesh. Instead the hydrologist discretizes the surface-water features and defines domains of the aquifer with similar hydraulic properties. The modeling technique forces the hydrologist to use distant boundary conditions to generate flow in the near-field; this makes the AEM uniquely well suited to modeling flow at large scales [de Lange, 1991, de Lange, 1996]. Unlike numerical models the analytic element method allows the user to analytically calculate potentiometric heads and groundwater fluxes at any point in the domain. In AEM models, common practice is to embed regions with highly-detailed elements in a less-detailed regional model. We have followed this approach. 3.2 Aquifer conceptualization The conceptual model used for the groundwater flow modeling assumes a single aquifer with an impermeable horizontal base, representing the regional unconsolidated aquifers. In the conceptual model, the aquifers are specified in 5 categories, as shown in Figures 4 and 5, and described in Table 2. When the model was constructed in GFLOW, the various domains were given transmissivity values such that the model matches the calibration data. The classification scheme is described in detail below: Regional intertill aquifer Variably connected sand and gravel deposits within thicker sur- ficial clay till formations. Typically, the overall transmissivity and recharge rates are low. Thin outwash aquifer Sand and gravel in the floodplain of the White River that is buried by 10 it or more of clay till. Intermediate values of transmissivity and recharge rate are expected. Thick outwash aquifer Localized, thick gravel formations in the floodplain of the White River; may be coincident with the location of the river. High values of transmissivity and recharge rate are expected. Bedrock hills Localized regions within the floodplain of the White River where the bedrock elevation is much higher than elsewhere in the floodplain. These bedrock hills can cut 12 Q Q Q Q Q Q Q Q Q Q Q Q Q o Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q o Q Q Q Q o Q Q off much of the permeable unconsolidated deposits. These are simulated as domains of very low transmissivity and low recharge. Mine pit ponds Inactive gravel operations that have filled with water. These have effec- tively infinite transmissivity and high recharge. For the purpose of the model, a trans- missivity value that is two orders of magnitude (lOOx) the outwash transmissivity was used. In the model, domains of varying transmissivity and recharge were created to simulate the various aquifer types (Figure 7). 3.3 Regional boundary conditions The approach used in this investigation was to first model regional flow and then to add detail near the Martin Marietta facilities. The local model used to evaluate impact was constructed by adding local detail to the regional flow model. Consequently, the local model contained the information obtained while modeling the regional system. The regional model represented boundary conditions such as rivers and streams in a coarse manner, while the aquifer system was modeled as a single, hydraulically conducting layer with locally varying transmissivity (Figure 6). The conceptual geologic model used for the regional analysis was based on our earlier modeling work done in central Indiana and water re- source investigations conducted by the U.S. Geological Survey. This was all supplemented by our own analyses of well log information obtained through the Indiana Department of Natural Resources [Bailey and Imbrigiotta, 1982, Gray and Hartke, 1989, IDNR, 1976, Meyer et al., 1975, Meyer, 1979, Smith, 1983]. The regional model includes all or parts of four counties in central Indiana: Hamil- ton, Marion, Hendricks, and Hancock counties. Also included were the significant aquifer properties, major rivers and streams, and all of the high capacity wells registered in the area The boundary conditions in an analytic element model are not perimeter conditions imposed by the modeler around a finite grid domain. Instead, the boundary conditions are natural hydrologic features, such as streams and lakes, that control flow in the aquifer. These surface-water features are provided as line-sink elements with specified heads (Figure 6). The far-field features were modeled with few, long line sinks. The line sinks used to model distant rivers do not include the effects of entry resistance along streams. It is not appropriate to perform detailed analyses in the far field; these elements are used only to ensure appropriate flow conditions at the perimeter of the study domain. 13 o Q Q g Q Q g g g g g (i) g Q g Q Q g g Q Q Q Q Q IQ Q Q g Q g Q Q Q g g g Q Q g g Q g Q Q 14 intertlll Q1;t\lifor Figure 4: Plan view of the relationships between conceptual aquifer domains. Aquifer type Transmissivity Recharge Regional intertill Low Low Thin outwash Medium Medium Thick outwash High High Bedrock hills Low Low Gravel pit ponds 00 High Table 2: Aquifer classification scheme for conceptual model. Q Q Q Q g Q g g g g Q g Q Q g Q Q g g g g g Q Q Q g g g g g g Q g Q g Q IQ Q g Q Q g Q Q ~ fl'ans!otlT1fJ ~ptlA1l ~I ~ ~~d1lilll (I'ow 1"lI:t6M'9") ~ 1rQI!l'llIi$"iyi,ty. c(lft,1,".xl inkl"tiD tlq,uif~ H m~rqt4 l' row' T Piigh T r- (;qnfin,dl "",'tw4P' O4I~tf"..., Figure 5: Conceptual model cross section. ~ -1 15 Q Q Q Q g Q g g Q g Q Q g Q Q Q Q g g tiil g g ,g Q ~ ,Q Q g Q g Q Q Q Q g g g Q Q Q Q g g Q 16 -~ / ~'c~n'~~ - ?'~ \ I 5 I o 5 Miles I LEGEND WiHman Hydro Planning Associates W.ltIl-.P..W..... P\;)--o"..C;M"F..J;:;T~ Ii) Model Wells N Model Linesinks N Hydrology D County boundary ..III...... Figure 6: Layout of wells and line sinks in the regional model. ~ IQ I iQ Q Q Q Q Q Q Q Q Q Q Q o Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q o Q Q Q o Q Q Q Q Q Q Q 3.4 Local boundary conditions The local model was used to assess both existing and future hydraulic conditions associ- ated with groundwater pumping in the Carmel area. Specifically, the area of detail includes Carmel Utilities Plant 4 Well Field and Martin Marietta's existing and proposed sand and gravel and limestone mining operations in Carmel. The local model contains the informa- tion obtained from modeling the regional system but also provides details that incorporate calibration head points, pumping, mine operation, variable recharge rates and transmissiv- ity, and more detailed surface-water features that include streams with variable resistance and quarry lakes. Since 1988, the IDNR Division of Water has required all surface water and groundwater withdrawal facilities within the state that have a pumping capacity of 70 gpm or greater to record and submit annual production data. This includes municipal, commercial, institu- tional, industrial, geothermal, and irrigation facilities as well as mining operations. Since significant pumping of an aquifer may influence both local and regional groundwater flow, a review of registered high-capacity groundwater withdrawal facilities located within Hamil- ton County was performed. The hydraulic significance of these pumping facilities, which includes Carmel Utilities' twenty production wells, were evaluated in the computer model using well elements. For the pumping rate, an average of Carmel's pumping from 1990- 2002 was calculated. Martin Marietta reports that it currently pumps approximately 6.4 million gallons per day (MGD) at the Indianapolis North Plant. This total includes inflow from the adjacent thin outwash aquifer, surface runoff and any seepage from the bedrock aquifer. Of this total, approximately 4.8MGD is estimated to be from groundwater inflow. Removal of water at the mining operations of the aquifer will be simulated in the computer model using a high conductivity inhomogeneity and a single line-sink with a specified head similar to the bedrock mine floor elevation. By using this strategy, the line-sink element will tell how much water is removed at the mine. In the analysis of uncertainty, hydraulic parameters (recharge and hydraulic conductiv- ity) were varied to achieve a calibrated model (sensitivity analysis). In all cases, aquifer porosity was modeled at 20%, and the aquifer thickness was assumed to be 80 feet. 3.5 Representing mine operations in the model In the local model, it is necessary to represent two hydrologic features that result from mining operation: (1) active mine pits; and (2) inactive pits that have filled with water. 17 Q Q Q g g g g g Q Q Q g g Q g g Q g (i) g g g o g Q g Q Q Q Q Q Q Q g (i) g o Q g g Q Q g g 18 -+ LEGEND 0.3 0 0.3 Miles Wittman Hydro Planning Associates .~FtR"SC.Ir:;er'1:.<rnirl!lCcf1s,.'ltJfrn . ~ Carmel wells Linesinks Hydrology Roads Inhomogeneities _ Mining ponds D Thick outwash _ Bedrock high I!III Thin outwash D Intertill Figure 7: Aquifer domains and local boundary conditions for the current conditions near Carmel Plant 4. Q Q Q Q ~ ~ Q ~ Q ~ ~ Q ~ ~ Q Q Q Q Q Q Q Q Q Q Q Q ~ Q Q Q Q Q ~ Q ~ Q Q ~ Q Q Q Q Q ~ These were simulated as follows: Active mining operations It was assumed that the potentiometric surface at the edge of the mine is held at the top of the bedrock by perimeter drains (as it is at the existing Martin Marietta quarry pit). In this case, the pit is modeled using line-sink elements with their head specified at the bedrock elevation. Since the model is based on con- stant transmissivity in each domain, it is necessary to specify an entry resistance to account for the thinning of the potentiometric surface where flow is unconfined near the perimeter drains, or for a seepage face. The resistance was determined as part of model calibration. Inactive gravel pits These are modeled as open water, using inhomogeneity domains with very high transmissivities (see above). In the post-mining model runs, the current quarry pit and the proposed pit at the South Mueller property were modeled in this manner. In addition, the inactive gravel pit that receives water pumped from the Martin Ma- rietta quarry pit perimeter drain will lose some water to the aquifer and the remain- der will overflow into Blue Woods Creek. This is modeled by adding a quantity of recharge to that pit in the model. 3.6 Potentiometric head observations Static water levels utilized for calibrating the model were selected from existing measure- ments reported on IDNR well logs and a number of Carmel Utilities' test well logs. Vari- ations in the reported static water levels at any location can be attributed to one or any combination of the following: seasonal fluctuations, long term climatic variations, different aquifers, transient pumping conditions and the reporting of erroneous well locations and/or static water levels by drillers. To ensure that appropriate data were used for calibration, we used these rules: . Only the Indiana Geological Survey (IGS) filtered IDNR database was used for IDNR well locations. IGS has provided quality control checks on their filtered database to reduce errors. . Only wells completed since January 1, 1998. . Only wells completed with depths in the range of 40 - 100 ft. . Static water levels from wells completed in the bedrock aquifer were removed. 19 Q Q Q Q Q Q Q Q Q Q Q Q Q Q ~ Q Q Q Q Q Q Q Q ~ Q Q ~ Q o ~ Q Q Q ~ ~ Q Q Q ~ Q Q ~ Q Q 4 Results This section describes the approaches used to determine the effects of the proposed mine expansion. In particular, we discuss the calibration of the model to potentiometric head data, the scenarios used to evaluate effects, and the examination of uncertainty related to model parameters. 4.1 Model calibration In the model, there were several parameters that could be adjusted as part of model calibra- tion: . Transmissivity and recharge rates in the regional intertill aquifer, buried sand-and- gravel aquifer, thick outwash aquifer, and bedrock hills . Entry resistance in surface waters . Entry resistance for the perimeter drain at the current Martin Marietta quarry pit . Amount of water infiltrated into the inactive gravel pit where water pumped from the Martin Marietta perimeter drain is disposed. For calibration purposes, the Carmel Plant 4 wells were given pumping rates consistent with their operation in 1993 (a representative year in the time period of available water- level data). Because the modeling analysis is intended to show differences arising from the proposed mine operation, these pumping rates were used throughout the modeling analysis, As described above, the regional model was calibrated to match well with observed heads from IDNR water well data that have been filtered for quality by the Indiana Geolog- ical Survey. In addition, local head observations were derived from the potentiometric map provided by Martin Marietta that was developed from their own monitoring wells. One flux calibration value was also available: the estimate that 4.8MGD of groundwater is pumped from the perimeter drain at the existing quarry pit. The parameter values that were determined during calibration of the model (matching both regional and local observations) are shown in Table 3. The spatial distribution of calibration error is shown in Figure 8. Summary statistics for the calibration results are shown in Figure 9. The "best-fit" modeled inflow into the existing perimeter drain was 4.78MGD. 20 Q Q Q Q Q Q Q g Q g g g Q g g Q g g g g Q g IQ 'g Q Q g g Q Q Q Q Q g Q g g g g g Q g Q g 21 + LEGEND 0.3 , o 0.3 0.6 Miles Calibration error (model head-obs. head) . <-10 .6. -10--5 '" -5 - 0 '" 0 - 5 .6. 5-10 . >10 . Carmel wells N Potentiometric contours '\/ Roads 1\/ Rivers D Hamilton County Wittman Hydro Planning Associates WaLB- R€-~o.m:.;.;t PkFllil-";; ConslA:ants Figure 8: Results of calibration and best-fit potentiometric surface. o g Q Q g Q g g g Q Q Q g Q :g ,g Q g Q Q g Q g ,Q g g g g g Q g g g g Q g g g Q Q g Q Q Q Parameter Value(s) Intertill aquifer Transmissivity 1600ft2 jd Intertill aquifer recharge rate 2.5injyr Buried sand-and-gravel transmissivity 5600ft2 jd Buried sand-and-gravel recharge rate 7.5injyr Outwash transmissivity 24000 ft2 j d Outwash recharge rate 10injyr Resistance of surface waters Ranged from I - 100d Infiltration at disposal pit 2.2M CD Resistance of perimeter drain 2d Table 3: Parameters for "best-fit" model. tt,'tieiid 0Jllbr~ltIC'l SJdstlc-s .rl""flJr t\!lp~ rom fXljt -SI<<i::tic: Greph: Cv.....: (~ ~le( PIoI liI,~oo.rcnc:. I fji5- r I 1121' ! [1"'--1 !-1 a I _I f14 f ~_l :41 I , ,--""- 1'51' 1m ~i'5I I 7sa o b:awkn: ('" ~.I't't'PJo:ib.!biI4lJ' M 111 liIninLn [l ifelel1l:a s..! 7,;' ~,DitMrcnc:1l lei liIedilll1 [J ii'c1el1Ql li\.,an .tJ:ncU. [l~ A DOI-li!olll1-Squ... [llt~ ,--- 11126.& 74~ s..rn. III SqlJ",ed [l iN eloet~~ in 720 l j 71~ 710 721 731 140 760 701 771t ?OO 700 800 Sll ClbUlYed KUltI Ic=-~ _.JI Figure 9: Summary calibration statistics from GFLOW. 22 EJ Q Q Q Q g Q Q g ,Q Q Q Q g Q Q g g g g Q g g Q g Q g g g Q Q g Q Q Q g g Q g g g Q g g g 4.2 Analysis of parameter uncertainty For each scenario in the analysis (see following sections), a sensitivity analysis for critical parameters was performed. The following parameters were adjusted: . Transmissivity of the thin and thick outwash aquifers was raised and lowered by 20% . Resistance of the White River was reduced by 50% . Resistance of the modified reach of Blue Woods Creek was reduced by a factor of lOxin the "South Mueller active" scenario only, to examine the possible impact should the construction of the new creek channel increase communication with the aquifer. For each scenario, a range of predicted values is provided for each of the indices of impact (see below). In addition, a composite capture zone analysis for the Plant 4 wells was performed for each scenario. The approach used to determine the 5-year composite capture zone is shown in Figure 10. Streamlines were traced back in time from each well for 5 years. The streamlines for all simulations were superimposed using GIS software and the composite capture zone was digitized manually as the polygon that enclosed all of the streamlines. Including ponds in capture zone delineations Our model predicts that a large fraction of the water pumped by the Plant 4 wells comes from the northernmost of three ponds (for this discussion, we refer to the ponds as "NW", "SE", and "SW" as noted in Figure 11) just west of the current quarry operation. Since the ponds are hydraulically connected to varying degrees, this introduces several complex issues for capture zone delineation. We have taken the conservative approach that if there is a possibility that one of the gravel pits contributes to the wellfield, it should be included in the wellfield capture zone. Further- more, we describe the ultimate sources of waters into the three ponds, for consideration in wellhead protection planning. Ponds SW and SE Ponds SW and SE are well connected by culverts, and according to the data collected by Martin Marietta, the water levels in both ponds are essentially the same. Water pumped from the collection ditches and other facilities at the quarry operation is pumped into Pond SW. Particulate matter that is entrained in this water settles into Pond SW, and the water then moves primarily into Pond SE, from which it then exits via a culvert under Gray Road into Blue Woods Creek. Although most of 23 Q Q g Q Q Q (i) Q g Q g g g g Q g g g g g Q g g Q g Q Q g Q g Q g Q g Q Q Q Q g g Q g g Q 24 , , -~~\~ \ "', r -~ \. ~~ \/ i /\ ,- -~"~ ~~~-r~ \ -~~ //--j I I ) ,/ /\ \- /11 ,A ( Wittman Hydro pfenning Associates \',l;]W.RE,~o_~~-n"llC:lo5...r.r15 . Carmel wells N Linesinks N Hydrology N Roads D Composite capture zone LEGEND Pathlines N Calibration N Low conductivity N Low transmissivity N High transmissivity + 0.1 , o 0.1 Miles Figure 10: Dete~mination of a composite capture zone. o Q o Q Q Q Q g g Q Q Q Q g g g Q g g Q Q Q g Q Q g g 'g g g Q g g ~ Q g g Q g g g Q Q Q 25 + 0.2 o 0.2 Miles LEGEND . Carmel wells N Roads _ Hydrology Figure 11: Former gravel pit ponds that contribute water to the Plant 4 wellfield, with identifying names. o o o ~ Q Q o o o o o o Q o Q Q Q o ~ 10 o o Q Q o o ~ Q o Q Q Q ~ Q ~ Q Q Q Q Q ~ Q o o the water moves from Pond SW into Pond SE, some also moves into Pond NW (see below). Pond NW Pond NW has two sources of water: (1) inflow from Pond SW through culverts that connect the ponds; and (2) inflow from Blue Woods Creek. Water-level data for Pond NW and Pond SW (Table 4) indicate that the water level in Pond NW is usually higher than the level in Pond SW, although in about 10% of the data, the level in Pond SW is higher. This indicates that water removed from the mining operation and runoff water in Blue Woods Creek are potential sources of contamination for the Plant 4 wells. From a wellhead protection standpoint, all three ponds (NW, SW, and SE) contribute water to the Plant 4 wells. All potential sources of contamination into the ponds should be con- sidered as part of the Plant 4 wellhead protection plan, including runoff of lawn chemicals and spills into Blue Woods Creek, and also potential contaminants entering the pond from the mining operation. 4.3 Reduced recharge analysis Since little is known about the stage-discharge relationships in the surface streams near the mine, it is not presently possible to simulate all features of the hydraulics of this system during a major drought. Indices of impact in this report are derived from the change in water levels or water leaving the inactive gravel pit and entering a Plant 4 well. Therefore, we have performed one "reduced recharge" run for each scenario. In each case, the potentiometric surface and streamlines from the Plant 4 wells are shown for the case that the recharge rates are reduced to 1/3 of the "best-fit" calibration values. The reduced-recharge scenarios suggest the effects on the capacity of the Plant 4 well- field that result from reductions in recharge. Two important factors that can reduce recharge are natural drought events and urbanization. The model does not include the effects of drought on major surface-water levels (e.g. in the White River), so it under-predicts the ef- fects oflong-term drought on Plant 4. However, it is likely that short-term (e.g. 1-2 month) dry periods in summer could reduce recharge (and therefore plant capacity) significantly for a short time. The issue of urbanization is more long-lasting. As development contin- ues near the Plant 4 wellfield, the long-term recharge rate may be significantly reduced by the presence of homes, streets and parking facilities, with surface-water structures handling runoff and reducing recharge. 26 Q Q Q Q Q o Q IQ IQ IQ Q Q Q o Q Q Q o o o r;;;; r;;;; Q Q Q Q Q Q Q Q o Q o Q o ~ Q Q Q Q Q o Q Q I Date I Pond NW I Pond SW I 5/22/02 NA 737.22 6/12/02 NA 736.92 8/13/02 NA 736.40 12/16/02 735.24 736.46 2/10/03 736.94 736.40 4/8/03 739.13 736.53 6/16/03 737.79 736.88 9/15/03 737.59 737.61 10/15/03 736.99 736.33 11/17/03 736.59 736.53 1/15/04 736.79 736.53 2/17/04 736.79 736.53 3/17/04 736.50 736.50 4/17/04 736.39 736.33 Table 4: Water levels in Ponds NW and SW (source: Martin Marietta Materials, Inc.). 27 IQ .~ o Q Q Q Q o o o Q Q o o o Q Q o Q o o o Q Q Q Q Q Q Q o r;; Q o o Q Q Q Q (;;; Q ~ Q Q ~ 4.4 Scenarios and results Three scenarios were perfonned, corresponding to the following: Current The current situation South Mueller Active The South Mueller gravel pit is fully built-out and the water table drawn down to bedrock. After Mining The South Mueller pit and the current quarry have been closed; they have filled with water and become lakes. The scenarios are discussed below. For the "South Mueller Active" and "After Mining" scenarios, three indices of impact are provided, 1. Change (in ft) for water levels in the currently-existing inactive gravel pit ponds, 2. Change in water level at a point located between the Plant 4 wells (this is related to the capacity of the wells), 3. Fraction of the total water pumped by the Plant 4 wells that travels from the inactive gravel pit south of the wellfield in less than 5 years. 4.4.1 "Current" scenario Results for the "Current" scenario are shown in Figure 12. The illustration shows the mod- eled potentiometric surface based on best-fit model parameters and the composite 5-year time-of-trave1 capture zone for the Plant 4 wells. Based on the sensitivity analysis, the Plant 4 wells receive about I-half (51-55%) of their water from the inactive gravel pit south of the wellfield. Although time-of-travel is not shown on the figure, it should be noted that the travel time from the gravel pit to well #12 is less than 1 year. The reduced recharge run for the current scenario is shown in Figure 13. The model predicts that the Plant 4 wells receive 68% of their water from the inactive gravel pit south of the wellfield. Consideration of the new Well #22 All of the analyses presented here are based on the 1993 "representative year" pumping rates for the Plant 4 wells. During the summer of 2004, the new Well 22 will be corning on-line at Plant 4. We have perfonned an additional set of model runs based on current 28 Q Q Q Q Q Q Q g Q g g g Q Q Q Q g Q g ~ g Q Q Q 'i) g g Q Q Q Q IQ ~. g g g g g Q Q g Q Q Q 29 -+ 0.2 , o 0.2 Miles LEGEND l~"f.'~ Wittman Hydro Planning Associates .:;~;~\, \'oI;Jb;>-RHQ.r:;.; ",",-nnq~...b-ll; ........ ........ .,...... . Carmel wells N Linesinks N Roads Potentiometric contours D 5-Year capture zone _ Hydrology Figure 12: Modeled potentiometric surface and composite S-year capture zone for the cur- rent conditions. Q Q o Q Q Q Q Q Q Q Q Q Q Q Q g g Q Q ~ g Q Q Q Q g Q g g g Q g Q Q Q Q g g Q Q Q Q Q g 30 + 0.2 o 0.2 Miles LEGEND Wittman Hydro Planning Associates ".Qt;;-" Roo,o...~ Pb'T"'l!l (;;Il'1S.J;;rts . Carmel wells N Linesinks N Roads Potentiometric contours N Reduced recharge pathlines _ Hydrology Figure 13: Current operations for "reduce-recharge" conditions. Q Q Q g g Q g g g Q g Q g Q g Q g Q g g Q Q Q Q g Q g g g g Q Q Q g g g Q g Q Q Q g g Q conditions, but adding Well 22. Since no actual pumping data for Well 22 are available, we assumed that in the "representative" scenario, it would pump the same amount as the largest well in the previous runs. The composite capture zone and potentiometric head field is shown in Figure 14. 4.4.2 "South Mueller Active" scenario Results for the "South Mueller Active" scenario are shown in Figure 15. The illustration shows the modeled potentiometric surface based on best-fit model parameters and the com- posite 5-year time-of-travel capture zone for the Plant 4 wells. Based on the sensitivity analysis, the Plant 4 wells receive about 70% (70-71 %) of their water from the inactive gravel pit south of the well field. Although time-of-travel is not shown on the figure, it should be noted that the travel time from the gravel pit to well #12 is less than 1 year. The total amount of water pumped from the quarry pit and the South Mueller pit is about 7.2MGD. Of this, we reinfiltrated the same fraction as found in the calibrated model at the gravel pit (3.3MGD), leaving about 3.9MGD (6cfs) in Blue Woods Creek. The changes between current conditions and those when the South Mueller property is being actively mined are shown in Figure 16. The most important value is the change in the water levels in the wellfield: the model predicts a decline of 2.0 - 2.8 ft. This corresponds to about 10-15% of the capacity reported in the well logs for Plant 4. When this model is run with a very low resistance in the modified reach of Blue Woods Creek, the results are quite different. An increased head of 0.3 ft is predicted at the well- field, and the Plant 4 wells receive 77% of their water from the inactive gravel pit. The reduced recharge run for the current scenario is shown in Figure 17. The model predicts that the Plant 4 wells receive 83% of their water from the inactive gravel pit south of the wellfield. 4.4.3 "Post Mining" scenario After surface operations at the South Mueller property are complete, a dry reclamation plan has been proposed. A network of drains will be incised into the bedrock surface to route water into the ditches near the current quarry operation. The water will then be pumped into the ponds west of the quarry. From a modeling perspective, the "post-mining" conditions do not differ from the conditions when the South Mueller operations are active. Thus, the "South Mueller active" scenarios are representative of the post-mining conditions. 31 {J Q Q Q g Q g g Q ~ g Q g Q Q Q Q Q Q ~ Q Q Q Q g Q g Q Q Q Q g Q Q Q Q g g Q Q Q g Q Q 32 + 0.2 , o 0.2 Miles LEGEND l~~"" Wiffman Hydro ptanning As.sodates .:;.;..'\. \'~.Rm)"""-~i>\.:I""fl",,G~..t;..-b ........ ........ ..".... . Carmel wells N Linesinks N Roads Potentiometric contours D 5-Year capture zone _ Hydrology Figure 14: Modeled potentiometric surface an composite 5-year capture zone for the current conditions with Well 22. o o Q Q Q o Q Q g Q Q Q Q Q g g Q Q Q g Q Q Q g Q g g Q g Q Q g Q g Q g g g g Q g g Q g 33 +- 0.2 o 0.2 Miles LEGEND Wittman Hydro Planning Associates \'.QlIl'."'....o.,..-~"'I.;J...,....~.I:7ll; . Carmel wells N Linesinks N Roads Potentiometric contours D 5-Year capture zone _ Hydrology Figure 15: Modeled potentiometric surface and composite 5-year capture zone when the South Mueller property is actively mined. g Q Q Q g Q g g Q g g Q Q Q g Q g Q Q g g Q Q Q Q g Q Q Q Q Q g Q g Q g g g Q Q Q g Q g 34 + 0.2 , o 0.2 Miles LEGEND . Carmel wells N Linesinks N Roads Range of water level a change from current conditions (ft) _ Hydrology Figure 16: Change in heads between current conditions and those when the South Mueller property is actively mined. o o Q Q Q Q <Ii) g Q Q Q Q Q Q Q <Ii) g Q g g Q Q Q <Ii) Q g Q Q g <.;> Q Q Q Q g g Q g Q Q Q g Q Q 35 + 0.2 o 0.2 Miles LEGEND Wiffman Hydro Planning Associates Wil~-IW.o;>..= "1.:1-.. nq <;~b""ls . Carmel wells N Linesinks N Roads Potentiometric contours N Reduced recharge pathlines _ Hydrology Figure 17: Results for "reduced-recharge" conditions when the South Mueller property is actively mined. o o Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q o Q o Q o Q o Q Q Q o Q Q Q o Q Q Q o Q Q Q Q Q Q 5 Conclusions There is no single index of the "effects of mining" that captures all of the changes that could take place once the proposed excavations are underway. Instead, we considered two types of impacts: changes in water levels in the aquifer near the Plant 4 well field, and changes in the source water pumped by the well field. The first of these is a measure of the reduction in yield that results from the competition between the drinking-water wells and the mining operations around the active pits. We examined impacts both during and after mining. . The wellfield is significantly affected by the reinfiltration of water pumped from the existing Martin Marietta quarry pit. Currently, about 2.2MGD of the 4.8MGD that is pumped from the drains at the quarry pit is reinfiltrated after it is disposed at the inactive gravel pond. The remaining 2.6MGD (4cfs) is lost from the gravel pit to Blue Woods Creek. . The addition of Well 22 to the current conditions increases the size of the Plant 4 wellhead protection area, particularly north and east of the wellfield. . Additional mining could reduce groundwater levels at the well field by about 2- 3 ft while the mines are active. The amount of water pumped by a well is roughly proportional to the drawdown in the well. Since the original pumping tests showed about 16ft of drawdown when the wells were pumped at capacity and the modeling predicts an additional 2 - 3 ft of decline, this translates to roughly 15% reduction in total wellfield capacity during mining. . Additional mining at the South Mueller property will increase the amount of water pumped from the mine by about 50%. . In the 20 years that the proposed mines are active there would be an increase in the amount of drinking water that comes from the pits and ponds. Currently we estimate that approximately one-half of the water pumped at the Plant 4 well field had been in a pit or pond within the last five years. While the mines are active this will increase to more than 70% of all the water in the well field. . With reduced recharge conditions, the effects of the proposed mining will be to de- crease water levels by 3 ft compared to reduced recharge under current conditions. With reduced recharge there are changes in water use and stage along the river that 36 Q o o Q (;) o o (;) o o o o o o o o o o o (,,) o o Q o o o o o o Q o o o o Q o o o o o o o o o may alter the actual conditions at the well field. The effects that we predicted are a re- sult of reductions in regional flow from the till aquifers to the West and the increasing importance of the White River to the East of the wells. . If the reconstructed section of Blue Woods creek has a very good connection with the aquifer, it may be able to supply a significant amount of water to the Plant 4 wells, even increasing the heads in the aquifer near the wellfield. . After mining ceases, the water levels near the wellfield will decline compared to current conditions as a result of the loss of additional aquifer recharge at the inactive gravel pit just south of the wellfield. The amount of water moving from the gravel pit to the wellfield will be about one-half what it is currently. . We recommend that a monitoring well program be initiated to evaluate the interaction between the inactive gravel pits south of the Plant 4 wellfield and the wells. A detailed monitoring plan is in development and will be submitted in mid-July. 37 Q ~ Q o Q o Q Q o o Q o o Q o o o o o Q o o o ~ o Q o o o ~ o o o Q Q o o o o Q o o o o References [Arihood, 1982] Arihood, L. D. (1982). Ground-Water Resources of the Upper White River Basin, Hamilton and Tipton Counties, Indiana. Water Resources Investigation Report 82-48, U.S. Geological Survey. [Bailey and Imbrigiotta, 1982] Bailey, Z. and Imbrigiotta, T. (1982). Ground-Water Re- sources of the Glacial Outwash along the White River, Johnson and Morgan Counties, Indiana. Water Resource Investigation 82-4016, U.S. Geological Survey. [Brown et al., 1995] Brown, S., Ferguson, V., and Flemming, A. (1995). Hydrologic Framework of Marion County, Indiana. Open File Report 93-5, Indiana Geological Survey, Environmental Geology Section. [Cable et al., 1971] Cable, L. W., Damiel, J., Wolf, R. J., and Tate, C. H. (1971). Water Resources of the Upper White River Basin, East Central Indiana. Water Supply Paper 1999-C, U.S. Geological Survey. [de Lange, 1991] de Lange, W. (1991). A Groundwater Model ofthe Netherlands. Techni- cal Report Note 90.066, The National Institute for Inland Water Management and Waste Water Treatment. [de Lange, 1996] de Lange, W. (1996). Groundwater Modeling of Large Domains with Analytic Elements. Master's thesis. [Driscoll, 1986] Driscoll, F. G. (1986). Groundwater and Wells. Johnson Screens, second edition. [Fenelon and Bobay, 1994] Fenelon, J. and Bobay, K. (1994). Hydrologic Atlas of Aquifers in Indiana. Water Resource Investigation 92-4142, U.S. Geological Survey. [Gillis, 1976] Gillis, D. C. (1976). A Model Analysis of Ground-Water Availability Near Carmel, Indiana. Water Resource Investigation 76-46, U.S. Geological Survey. [Gray and Hartke, 1989] Gray, H. and Hartke, E. (1989). Geology for Environmental Plan- ning in Marion County, Indiana. Geological Survey Special Report 47, Indiana Depart- ment of Natural Resources. [Gray, 1983] Gray, H. H. (1983). Map of Indiana Showing Thickness of Unconsolidated Deposits. Misc. Map. 38 (;) Q o o (,;) (,;) (,;) (,;) o (,;) Q r;,;; (,;) o o o o (;;; o Q o o o o o o o o o r;,;; o o o r.,;;; o o o (,) o o o o o o [Haitjema,1995] Haitjema, H. (1995). Analytic Element Modeling of Groundwater Flow. Academic Press. [IDNR,1976] IDNR (1976). Technical Atlas of the Ground-Water Resources of Marion County, Indiana. Water resource assessment, Indiana Department of Natural Resources, Division of Water. [Jones and Henry, 1994] Jones and Henry (1994). Wellhead Protection Area Delineation Report. Technical report, Carmel Utilities, Carmel, Indiana. [Lapham, 1981] Lapham, W. W. (1981). Groundwater Resources of the White River Basin, Madison County, Indiana. Water Resource Investigation Report 81-35, U.S. Geological Survey. [Meyer, 1979] Meyer, W. (1979). Geohydrologic Setting of and Seepage from a Water- Supply Canal, Indianapolis, Marion County, Indiana. Water Resource Investigation 79- 115, U.S. Geological Survey. [Meyer et al., 1975] Meyer, W., Reussow, J. P., and Gillis, D. C. (1975). Availability of Ground Water in Marion County, Indiana. Open File Report 75-312, U.S. Geological Survey. [Saul and Robinson, 1989] Saul, M. and Robinson, B. (1989). Streambed Permeability and Seepage in White River, Marion County, Indiana. Special Project 4, Indiana Department of Natural Resources, Division of Water. [Smith, 1983] Smith, B. (1983). Availability of Groundwater from the Outwash Aquifer, Marion County, Indiana. Water Resource Investigation 83-4144, U.S. Geological Survey. [Strack, 1989] Strack, O. (1989). Groundwater Mechanics. Prentice Hall. [WHPA, 2003] WHPA (2003). Phase I wellhead protection plan. Technical report, Carmel Utilites, Carmel, Indiana. [Wittman and Haitjema, 1995] Wittman, J. F. and Haitjema, H. (1995). Delineation of Wellhead Protection Zones in Marion County - A Report to the Marion County Well Field Protection Technical Committee. Technical Report 97-E03, IUPUI Center for Ur- ban Policy and the Environment. 39