The effects of the Penry Wellfield (Delaware, Ohio) on well-water quality.

Mann, Keith O.

ABSTRACT. Water samples collected during three phases (Background, Pumping and Recovery) of a year-long study in Delaware County, Ohio, document groundwater quality and the effects of pumping up to 1,500 gpm from the City of Delaware's Penry Wellfield on nearby well-water quality. Study-phase means of pH ranged from 6.84--6.99 and alkalinity means varied from 352--371 mg/L (as CaC[O.sub.3]), while specific conductance means exceeded 825[micro]S/cm, disolved solids means exceeded 499 mg/L, hardness means exceeded 525 mg/L (as CaC[O.sub.3]), and iron means exceeded 2.45 mg/L. The turbidity medians for all study phases exceeded 0.75 NTU. No significant links, either in a predictive or general sense, between water-quality parameters and well depth, flow path, time, or combinations of these variables existed. The average pH of the Background Phase (6.99) differed significantly ([alpha]=0.05) from both the Pumping Phase (6.89) and the Recovery Phase (6.84) and the mean alkalinity concentration of the Pumping Phase (352 mg/L) differed significantly from the Background Phase (371 mg/L). The Pumping Phase experienced significantly larger dissolved solids concentrations (639 mg/L) than either the Background Phase (499 mg/L) or the Recovery Phase (521 mg/L); however, no significant differences were detected with respect to specific conductance, hardness, or iron concentrations. The turbidity median of the Background Phase (0.75 NTU) differed significantly from both the Pumping Phase (1.48 NTU) and the Recovery Phase (1.36 NTU) medians and turbidity values routinely (52%) exceeded 1.0 NTU. Finally, pumping did not cause [H.sub.2]S concentrations to rise above detectable levels.

INTRODUCTION

The City of Delaware, Ohio, faces problems that a number of municipalities in Ohio and the Midwest face: providing ample, safe drinking water to a growing population. The City of Delaware used the Olentangy River as the sole source for drinking water until 1994 when the city brought into operation three wells, providing a total of 1,000 gpm, on the grounds of their water treatment plant. Soon thereafter, the city realized a need for an additional source of groundwater to blend with Olentangy River water when this surface water contains undesirable levels of agricultural contaminants, which occurs intermittently from April through September. In 1998, based on the findings of earlier studies (Ground Water Associates 1991, Ground Water Associates 1992), the city began to explore areas close to the water-treatment plant to establish an additional wellfield. During the early stages of that investigation, conducted by Collector Wells International Inc., the city purchased land on Penry Road, 5.6 km north of Delaware. After three 30.38 cm diameter production wells (TW-4, 65.5 m deep; TW-5,62.8 m deep; and TW-6a, 71.6 m deep) were drilled (Fig. 1), Collector Wells International Inc. performed several pump tests and determined the Penry Well field could supply 1,500 gpm. The study also noted that wells less than 15.2 m (50 ft.) deep within a 3.2 km (two mile) radius of the wellfield could be at a &watering risk during well field operation. In fact, during the 1998 pump tests, several wells near the wellfield experienced significant water-level declines.

Sometime after the completion of the well tests, a number of residents living near the wellfield became concerned about the potential impact of the wellfield on groundwater supply and quality and voiced their concerns to the local press and the City of Delaware. Because of these worries, the City of Delaware and its consultant, Malcolm Pirnie Inc., initiated an investigation designed to document the effects (both on groundwater quantity and quality) of the Penry Wellfield on residential wells within a 3.2 km radius of the wellfield. Collector Wells International Inc. was primarily responsible for supervising the study and collecting water-level measurements, while faculty and students at Ohio Wesleyan University concentrated on the water-quality portion of the study. Other agencies and people also contributed to this study, including the Ohio Department of Natural Resources, the Delaware Department of Health, and many residents of Troy Township. This report focuses on the water quality within a 3.2 km radius of the Penry Wellfield and so presents only a portion of the data collected and included in the comprehensive report of Collector Wells International Inc. (2001). Whereas earlier studies (Ground Water Associates 1991, Ground Water Associates 1992, Collector Wells International Inc. 1998), concentrated on locating and establishing a wellfield that could provide sufficient quantities of safe groundwater to fulfill the needs of the city, this present report, building on that earlier work, addresses solely the potential impact on water quality caused by the wellfield.

[FIGURE 1 OMITTED]

Geologic and Hydrologic Setting

Physiographically, Delaware County lies within the Till Plains of the Central Lowlands, immediately west of the Allegheny Plateau. The topography of Delaware County, led Westgate (1926) to remark "... the most striking scenic feature of the county is its flatness." Regionally, the study area lies upon the eastern flank of the Findlay Arch with Paleozoic bedrock units dipping gently (~4 m/km) eastward, toward the Appalachian Basin (Fig. 1), with tills of the Scioto Lobe, deposited during the Wisconsinan, covering the Paleozoic bedrock. The northern border of the Broadway Moraine lies immediately south of Penry Road and the St. Johns Moraine rests 21 km to the north, outside the study area.

The near-surface Silurian and Devonian units relevant to the study include, in ascending order, Salina undifferentiated (dolomite), Columbus Limestone (the lowest portion, the Bellepoint Member, is primarily dolomite with the remainder of the Columbus Limestone composed of limestone), and finally the Delaware Limestone (Fig. 1). Although unconformities separate these carbonate units (the Walberg Unconformity separates the Columbus Limestone from the underlying Salina undifferentiated and the australis conodont zone is unrecorded in Central Ohio [Sparling 1983, 1985] between the Columbus Limestone and the overlying Delaware Limestone), the Silurian and Devonian carbonates act as a single hydrologic unit, confined below by Ordovician shale-rich units. The structural-contour maps of Larsen and others (1992a, 1992b, 1992c) suggest that the total thickness of the Silurian-Devonian carbonates in Delaware County is roughly 230 m.

Delaware County is within the Midwestern Basins and Arches Aquifer System (as defined by Shaver 1985), which lies along the axes of the Cincinnati, Findlay, and Kankakee Arches, with the adjacent flanks of the Appalachian, Michigan, and the Illinois basins comprising the remainder of the system. Groundwater occurs primarily in fractures, bedding joints, and secondary porosity within the Midwestern Basins and Arches Aquifer System (Bugliosi 1989, 1990). Casey (1994) noted that the bedrock within 76-92 m of the surface contains the greatest number of fractures and solution-enlarged openings, created by weathering, unloading (both through erosion of Paleozoic sediments and repeated glaciations), and groundwater flow. Arihood (1994), working in the Devonian and Silurian carbonates covered by Pleistocene sediments in northwestern Indiana, determined that groundwater primarily followed horizontal fractures. The Devonian carbonates exposed within the Penry Quarry, 6 km west of the Penry Well field, display typical horizontal and vertical fractures (strike N20 [degrees] E and N70 [degrees] W) found throughout the Devonian Carbonate Aquifer in cental Ohio and so the area surrounding the wellfield area most likely possesses flow characteristics typical of the Midwestern Basin and Arches Aquifer System.

The geology of the Pleistocene sediments contributes significantly to groundwater in the Midwest. In some settings throughout the Midwestern Basins and Arches Aquifer System, Pleistocene sediments serve as aquifers. In northern Delaware County near the Penry Wellfield local lenses of sand and gravel within glacial tills supply a few shallow, private wells. In addition to serving as aquifers, the fabric of Pleistocene sediments, especially the clay-rich tills, in central Ohio can play an important and often an underappreciated role in groundwater studies. A number of studies (e.g. Grisak and others 1976, 1980; Prudic 1986; Keller and others 1988; McKay and others 1993; Strobel 1993) have shown that fractures (as well as conduits created by the dissolution of carbonates) within glacial sediments significantly increase conductivity and so promote surface recharge and enhance the connection between the overlying glacial aquifers and the underlying carbonate aquifers. Allred (1999, 2000) noted that the hydraulic conductivity of fractured glacial till is often two, or more, orders of magnitude greater than that of the matrix. Fausey and others (2000) examined tills in Madison County, Ohio and found that the fractured till they sampled had saturated hydraulic conductivity values one or more orders of magnitude greater than the non-fractured portions. These results support the contention that workers should not automatically view till deposits as impermeable. Fausey and others (2000) felt their results probably applied to other tills in central and western Ohio as well. In their review the literature and their own examination of sites in Ohio (one fairly close to the Penry Wellfield), Brockman and Szabo (2000) found that Quaternary deposits are commonly fractured in all glaciated regions of Ohio. Weatherington-Rice and others (2006) remarked that because this ubiquitous fracturing affects hydraulic conductivity so significantly, geologists in Ohio have needed to alter the DRASTIC (Depth to water, Recharge, Aquifer, Soils, Topography, Impact to vadose zone Conductivity) methodology. Fractured glacial tills now possess their own vadose zone ratings, thus increasing both Recharge and Impact values in the DRASTIC calculations.

A number of years ago, while investigating the hydrogeology of an area to the north of the Penry Wellfield for the potential siting of a Superconducting Super Collider, Raymondi (1997) noted that a high degree of interconnection existed between the carbonate aquifer and the overlying glacial aquifer and that these two aquifers responded to pumping as a single hydrologic unit. The stratigraphy of the sites Raymondi (1997) studied is similar to the Penry Wellfield; however, the Waldo, Ohio locality (30 km north of the Penry Wellfield) also possessed 5.5 m of Olentangy Shale between the Delaware Limestone and the overlying glacial sediments.

Regional Flow and Chemistry

Intermittently, investigations have addressed both the regional flow and the chemistry of the Silurian and Devonian Aquifer in central Ohio. A number of observations, including the lack of a groundwater age increase along dominant regional flow-paths, led Eberts and George (2000) to conclude that the Midwestern Basin and Arches Aquifer System possesses alternating recharge and discharge areas on a scale of less than 16 km. In an earlier study, Hanover (1994) determined that 84% of the groundwater that travels along a flow path from the Bellefontaine Outlier in central Ohio to Sandusky Bay on Lake Erie flows less than 8 km in the subsurface. Both Norris and Fidler (1973) and Eberts and George (2000) presented maps that showed the potentiometric surface near the Penry Wellfield was slightly above 274 m with the Olentangy River to the east and Scioto River to the west acting as areas of discharge. On a local scale, Raymondi (1997) measured the potentiometric surface in portions of Delaware, Marion, and Union counties and found an area slightly above 286 m between the Olentangy and Scioto rivers that had nearly a flat gradient and functioned as a groundwater divide, with groundwater flowing toward the rivers on either side of this divide. The Penry Wellfield rests on the eastern flank of this divide, with the potentiometric surface sloping eastward (Collector Wells International Inc. 2001) toward the Olentangy River. The regional flow-model of Eberts and George (2000) also indicated the presence of a drainage divide running east-west just to the north of Delaware County, in Marion County. With this drainage divide just north of Delaware County, a groundwater divide directly to the west of the Penry Wellfield, and the Olentangy River immediately to the east serving as a local discharge area, it appears that the Penry Wellfield lies in an area that possesses typical flow characteristics found in the Midwestern Basin and Arches Aquifer System.

Recent investigations on recharge rates near the Penry Wellfield have produced nearly identical results. The map produced by Eberts and George (2000), using a model to simulate regional groundwater flow, indicates that the area surrounding the Penry Wellfield has a recharge rate of just over 10.2 cm/yr. In a more recent report, which used a number of computer modeling programs to estimate recharge rates for 103 drainage basins in Ohio, Dumouchelle and Schiefer (2002) estimated that the Olentangy River Basin near the City of Delaware experienced a recharge rate of 12.7 cm/yr. It should also be noted here that the map (the 50-yr. average annual precipitation for Ohio) of Harstine (1991) shows that western Delaware County receives about 91.5 cm/yr.

Fairly recently, Eberts and George (2000) delineated and mapped seven hydrogeochemical facies within the Midwestern Basin and Arches Aquifer System. Their map indicates that the carbonate aquifer in north-central Delaware County belongs to a Ca-Mg-S[O.sub.4] facies with an isolated area comprised of a Ca-Mg-HC[O.sub.3] facies lying directly to the northwest and north of Delaware County. They also concluded, based on concentrations of dissolved solids and sulfate, that groundwater chemistry did not change along regional flow-paths.

METHODS

Using results from an earlier investigation (Collector Wells International Inc. 1998), Malcolm Pirnie Inc., The City of Delaware, and Collector Wells International Inc. established a well-monitoring network within 3.2 km radius of the Penry Wellfield (Fig. 1) for this study. Although the monitoring wells were selected to ensure a wide distribution with respect to depth and geography, well selection was contingent upon landowner permission, known well construction, and plumbing systems with taps available before water-treatment systems. The original water-quality monitoring network consisted of 29 private wells, but this paper excludes five of those wells: one well was in very poor condition, finished in glacial sediments, and sampled only once; the owner of another well wintered in the south and so that well was sampled only three times; and three wells were abandoned during the study. Because most of the land in the study area that lies to the east of US Route 23 is within the Delaware Reservoir Wildlife Area, the aerial coverage of the study was reduced by about 25%, from 32.5 [km.sup.2] to 24.6 [km.sup.2]. Collectively the 24 private wells used in the present report span the uppermost 36 m of the carbonate aquifer and so the wells provide a sampling of an aquifer volume of about 0.83 [km.sup.3].

Sampling Rounds and Study Phases

After establishing the well-monitoring network, the investigation basically consisted of three phases: a Background Phase, a Pumping Phase, and a Recovery Phase. The Background Phase monitored water quality without pumping the wellfield. Initially, this phase was scheduled to begin in July 2000 and end in December 2000 when the Pumping Phase would begin; however, due to an unexpectedly long delay in bringing electrical service to the wellfield, the Pumping Phase did not begin until 18 April 2001. Samples collected during the Background Phase were collected in July (3-7), August (22- 24), October (20-24), and December (8-11). The Pumping Phase lasted for four weeks with the following pumping rates used: 18-26 April, well TW-5 pumping at 590 gpm; 26 April-7 May, wells TW-5 and TW-4 pumping at a combined rate of 1,000 gpm; and 7-16 May, wells TW-5, TW-4, and TW-6a pumping at a combined rate of 1,490 gpm. Water-quality samples were collected during all three of the pumping steps on the following dates: 20-23 April, 27-29 April, and 11-13 May. The final two water-quality sampling rounds (29- 31 May and 13-15 June) were conducted during the Recovery Phase of the study (17 May-15 June).

Standard sampling and field measurement procedures were employed at each residence with field measurements (water level, temperature, pH, specific conductance, and turbidity) performed prior to collecting samples for laboratory analyses. Water samples were collected using standard EPA approved procedures (for analysis of hydrogen sulfide, total alkalinity, total iron, total hardness, and total dissolved solids) and then submitted to MASI Environmental Laboratories (Dublin, Ohio), a commercial laboratory certified by the Ohio EPA to perform chemical analyses on public drinking water. Field duplicates, about one for every 10 samples, were collected and submitted (using fabricated identification numbers and collection times) to MASI Environmental Laboratories as well.

RESULTS AND ANALYSES

Table 1 documents the laboratory precision of the duplicate samples sent to MAS I Environmental Laboratories during the study. Table 2 contains both field and laboratory data for all individual sampling rounds: data are also grouped by study phase (Background Phase: sampling rounds 1-4; Pumping Phase: sampling rounds 5-7; and Recovery Phase: sampling rounds 8 and 9). When laboratory duplicates were collected, the average value was used for statistical analysis. Figure 2 presents graphs of the water-quality data plotted with respect to time (sampling round). Although 223 analyses were conducted for hydrogen sulfide, only three samples were above detectable limits (<0.05 mg/L) and all three of these samples (with concentrations of 0.05 mg/L, 0.07 mg/L, and 0.09 mg/L) were collected during the Background Phase.

Background Phase

Figures 3 and 4 show the data collected during sampling round 1 plotted with respect to geography (from west to east) and well depth. The geographic distribution from west to east was used because the potentiometric surface slopes to the east, toward the Olentangy River. The graphs for the remaining eight sampling rounds look very similar to those of sampling round 1. Although, visually, neither geography nor well depth appears to predict water chemistry, a number of regression analyses were conducted to document whether geography (with a flow-path from west to east), well depth, time (sampling round), or combinations of these independent variables could be used to predict water chemistry. Linear regression was applied first to the data from each of the sampling rounds (1-4) of the Background Phase. Table 3 displays the results of linear regression analyses performed on the measured parameters (pH, alkalinity, specific conductance, dissolved solids, hardness, turbidity, and iron). Note that the largest adjusted [R.sup.2]value for sampling round 1 is 0.29 with most (11 of 14) of the values less than 0.10, indicating that both well depth and geography cannot be used as predictive indicators of water chemistry.

Multiple regression analysis was applied next to sampling round 1 data using two different models: the first-order model

[Y.sub.[??]] = [[beta].sub.0] + [[beta].sub.i], (well depth) + [[beta].sub.2] (geography) + [[epsilon].sub.[??]] Equation 1

[Y.sub.[??]] = response

[[beta].sub.0] = Y intercept

incorporated two predictor variables (well depth and geography) with model parameters ([[beta].sub.1] and [[beta].sub.2]) and the error term ([[epsilon].sub.[??]],); and the second model

Y = [[beta].sub.0] + [[beta].sub.1] (well depth) + [[beta].sub.2](geography) + [[beta].sub.3] + [[epsilon].sub.[??]]

Equation 2

contained the interactingeffect, [[beta].sub.3] (well depth x geography), among the predictor variables.

The adjusted [R.sup.2] values of the multiple regression analyses for sampling round 1 data (Table 3) demonstrate that the data poorly fit the various regression models and that the majority of the regression models (nine of 14) explain less then 15% of the variation, with the best model responsible for only 27% of the measured variation. Such findings indicate that well depth and geography have insignificant controlling effects upon water chemistry within the study area. The results for the other sampling rounds (2, 3, and 4) of the Background Phase are very similar to, and in fact poorer than, the results for sampling round 1.

Pooling the data from the first four sampling rounds into a single Background-phase data set allowed the impact of time (sampling round), geography, and well depth to be tested over an extended period (July-December 2000). A similar statistical approach was used on these pooled-background data as was used for individual sampling rounds. Linear regression, including the variable time (sampling round), was applied to the data followed by several multiple-regression analyses using the following independent variables and groupings: geography and well depth; well depth and an interaction variable (geography x well depth); and finally geography, well depth, and time (sampling round). Table 3 shows the adjusted [R.sup.2] values for both linear and multiple regression models. These regression models perform no better than those models built solely on sampling round 1 data and in fact most of these models perform worse than those of sampling round 1. Note that the temporal variable accounts for less than 5% of the variation for six of the seven parameters measured and it explains only 19% of the variation for dissolved solids.

Pumping Phase

Only the results for the regression analyses of sampling-round 7 data (Table 4) are presented here, because it was the last sampling round of the Pumping Phase and so any effects caused by pumping would have a higher probability of being detected during this sampling round than sampling rounds 5 and 6, which, incidentally, had very similar results to those of sampling round 7. The same statistical approach used previously was also applied to sampling-round 7 data, with one important difference: instead of using the independent variable geography (west to east) a new independent variable, distance (measured radially outward from production well TW-5), was used. The statistical results for sampling round 7 are nearly equivalent to the regression results obtained from the Background-Phase data. The adjusted [R.sup.2] values of the linear-regression models do not exceed 0.25 (Table 4). Only alkalinity experiences a correlation greater than 0.20 and 11 of the remaining 13 adjusted Rivalues are less than 0.15. The multivariate-regression models perform similar to the previous sampling rounds with all adjusted [R.sup.2] values below 0.30. Taken as a whole, combining the data of sampling rounds 5-7 (Table 4) produces similar correlation patterns and the linear models function no better than the models based on sampling round 7 or the Background Phase, with only two of the water-quality parameters possessing adjusted [R.sup.2] values above 0.10, but below 0.21. Likewise, multivariate regression models of the pooled data (sampling rounds 5-7) behaved similarly to those performed solely on sampling round 7, with only five of 21 values rising above 0.20, yet below 0.31.

[FIGURE 2 OMITTED]

[FIGURE 3 OMITTED]

[FIGURE 4 OMITTED]

In order to explore further the effects of pumping upon water-quality parameters, a number of additional regression analyses were conducted using a slightly modified data set. Instead of using the measured water-quality parameters, the change in value (determined by calculating the difference between sampling round 7 and 4) was used. The independent variables also differed for this portion of the analysis: the effects of water-level elevation, change in water-level, length of open-bore exposure, and distance to Well TW-5, were investigated. Subtracting the water level measured during sampling round 7 from the water level measured just prior to the Pumping Phase produced the change-in-water-level variable and the length of open-bore exposure was calculated by subtracting the elevation of the water level in the well during sampling round 7 from the elevation of the casing bottom (only eight wells had water levels that fell below the top of the casing). Table 5 contains the results of linear regression analyses performed on these new independent variables and the changes in parameter values. The results indicate that only one of the 28 models possessed an adjusted [R.sup.2] value above 0.20, while 25 of the remaining 27 models produced adjusted R2 values below or equal to 0.15.

Recovery Phase

Table 6 contains the results of the regression analyses performed on data from sampling round 9, the last sampling round. The effects of both geographic variables (west-to-east and distance-from-well TW-5) were investigated during this portion of the analyses. Again, the results are very similar to those of sampling round 7, the last sampling round of the Pumping Phase, and show that the models usually produced the highest adjusted [R.sup.2] values for alkalinity. Additionally, hardness and specific conductance often display higher adjusted [R.sup.2] values than were documented during the Background or Pumping Phases; however, all adjusted [R.sup.2] values remain low (below 0.31), with 17 of 21 values below or equal to 0.20. Multivariate analyses showed better correlation between water quality measurements and the independent variables, as one would expect. Again alkalinity, specific conductance, and dissolved solids displayed the highest correlations, with only four of 28 values larger than 0.30, yet less than 0.42. Table 6 also displays the results of both linear and multivariate regression models run on the pooled data set (sampling rounds 8 and 9) of the Recovery Phase. These results essentially repeat the results obtained for sampling round 9.

The statistical results on hardness concentration, for both sampling round 9 and the pooled data of sampling rounds 8 and 9, are peculiar: models that contain either geography or distance-from-well TW-5, show similar correlations with some of the highest adjusted [R.sup.2] values recorded during the study. This is surprising because these independent variables order the spatial distribution of the data quite differently (and in some instances oppositely): the geography variable orders the data from west to east while the distance variable arranges it radially outward from well TW-5. This curious result clearly demonstrates that regression models with [R.sup.2] values of 0.45 are essentially meaningless in a practical setting.

Water-Quality Trends

Although the independent variables of geography (measured either from west-to-east or as distance-from-well TW-5) and well depth cannot be used as predictors of water chemistry, general geographic trends in water quality may exist within the study area. In order to address this possibility, ANOVA was conducted for each water quality parameter with respect to geography during each sampling round (Table 7). In this portion of the analyses, the Bonferroni Correction, based on the Bonferroni Inequality, was used to help guard against the accumulation of Type-I errors (a false rejection of the null hypothesis) when conducting multiple comparisons. The Bonferroni Inequality

p [greater than or equal to] [(1-[alpha]).sup.K] Equation 3

P = probability

[alpha] = level of significance

K = number of simultaneous comparisons

reveals that as the number of simultaneous comparisons rises, the probability of committing a Type-I error increases dramatically, if one uses the same confidence interval for each test as used for the entire study. For example, if 20 simultaneous t-tests are conducted (K = 20), using [alpha] = 0.05, the Bonferroni Inequality indicates that the probability of committing a Typed error would not be 5%, but would inflate to 64%: clearly an unacceptable level. The Bonferroni Correction,

[[alpha].sub.i] = [[alpha].sub.s/K, Equation 4

adjusts the confidence interval to reduce the chance of committing a systematic increase of Type-I errors when performing multiple-comparison statistical tests. The Bonferroni Correction enables the researcher to determine the level of significance ([[alpha].sub.i])for a given number of individual tests (K), once the study-wide level of significance ([[alpha].sub.s]) has been chosen. For this study, using [[alpha].sub.s] = 0.05 and performing seven individual comparisons (K= 7) for each round, the Bonferroni Correction indicates that [[alpha].sub.c], should be set at 0.007. Of the 63 slopes tested for significance, four significant trends during the entire study were identified: pH showed a significant trend only during sampling round 8 (declining along the flow path), specific conductance experienced a significant trend only during the last sampling round, and hardness displayed a positive trend (increasing along flow path) during both the first (see also Fig. 3) and last sampling round. Interestingly, the slope (both significant and non-significant) of 30 of the 63 comparisons were in fact opposite to what one would expect given the flow-paths of the study area. Clearly no significant trends persisted during any study phase and thus no strong and well-defined general trend of water quality exists within the study area either during non-pumping or pumping conditions. Such few and sporadic findings of significance may simply result from random chance (with four of 63, 6%, comparisons documenting significant findings).

Among Phase Analyses

Since no spatial or temporal trends appear to exist within each study phase, the data of each study phase can now be pooled and water-quality differences among the study phases can be explored. Although the independent variables could not be used to predict water quality within each of the three study phases, comparisons of the water-quality parameters among the study phases may reveal effects of pumping on water chemistry. Both parametric and nonparametric methods were used to explore this possibility. Parametric testing began by first applying ANOVA and if it exposed significant differences, multiple t-tests were then applied. When appropriate, the non-parametric Kruskal-Wallis Test was used, followed by the Wilcoxon Rank Sums Tests for pair-wise comparisons.

Table 8 contains ANOVA results that show only three parameters, pH, alkalinity, and dissolved solids, exhibited significant differences among the study phases. Since the F-test identified these significant differences, naturally the next step sought to identify which study phases (Background, Pumping, and Recovery) differed from each other. Although conducting an F-test first helps protect against Type-I errors, the Bonferroni Correction was again employed to help guard against the accumulation of Type-I errors when conducting multiple pair-wise comparisons. For this study, using a[[alpha].sub.s] = 0.05 and performing three individual t-tests (Background verses Pumping, Pumping verses Recovery, and Background verses Recovery), the Bonferroni Correction indicates that [[alpha].sub.[??]], should be set to 0.016. Table 9 contains the results oft-tests performed on pH, alkalinity, and dissolved solids, grouped by study phase (Background, Pumping, and Recovery). Combining the results tabulated in Table 9 with the data in Table 2 show that the pH during the Background Phase was significantly greater than during the Pumping and Recovery Phases. These tables also reveal that the Pumping Phase experienced lower alkalinity values than the Background Phase, but the alkalinity of the Recovery Phase did not differ significantly from either the Background or Pumping Phases. Finally, the Pumping Phase experienced significantly greater concentrations of dissolved solids than either the Background Phase or the Recovery Phase. This significant difference in dissolved solids concentrations among study phases is perplexing since other closely allied variables (specific conductance and hardness) do not also experience such significance differences.

Because the turbidity data deviate sufficiently from normality (note the turbidity means and medians of Table 2), non-parametric methods were employed to test for significant changes induced by pumping. After performing a Kruskal-Wallis Test on the turbidity data, which indicated a significant difference (p < 0.001) existed among the three study phases, three pair-wise (Background verses Pumping, Pumping verses Recovery, and Background verses Recovery) Wilcoxon Rank Sums Tests (Table 10) were performed, again using the Bonferroni Correction. These results (Table 10) alongwith Table 2 indicate that the turbidity values measured during the Background Phase were significantly lower than those recorded for either the Pumping Phase or the Recovery Phase.

DISCUSSION AND CONCLUSIONS

The data reported in Table 2 show characteristics of the water within a 3.2 km radius of the Penry Wellfield, under pumping and non-pumping conditions. The groundwater is basically neutral with pH study-phase means that ranged from 6.84 to 6.99 (falling within the National Secondary Drinking Water Regulations of the United States Environmental Protection Agency [1994] of 6.5-8.5) and possessed mean alkalinity values for study phases between 352-371 mg/L (as CaC[O.sub.3]). All three study phases had specific conductance means that exceeded 825 [micro]S/cm and the mean dissolved solids concentrations commonly (seven of the nine sampling rounds) surpassed the National Secondary Drinking Regulations (United States Environmental Protection Agency 1994) of 500 mg/L. The water is very hard, with means of the study phases exceeding 535 mg/L (as CaC[O.sub.3]) and has high iron concentrations with all study-phase means greater than 2.65 mg/L, exceeding the National Secondary Drinking Regulations (United States Environmental Protection Agency 1994) of 0.3 mg/L. The turbidity medians of all three study phases exceeded 0.70 NTU and the median turbidity value for each of the sampling rounds exceeded 0.50 NTU, with six of the sampling rounds exceeding 0.90 NTU. In fact, 52% of all turbidity measurements made during the investigation exceeded 1.0 NTU. The Background Phase had medians that ranged from 0.52 NTU to 0.90 NTU while the medians of the Pumping and Recovery Phases ranged from 1.29 NTU to 1.55 NTU. Finally, the S[O.sub.4.sup.2-] analysis (120 mg/L) from 1998 (Table 11) shows that the S[O.sub.4.sup.2-] concentration of the Penry groundwater probably falls below the National Secondary DrinkingWater Regulations of 250 mg/L. The findings of this study support the conclusions from recent (2000, 2002, and 2006) Ohio EPA 305(b) reports that noted the Ohio carbonate aquifers commonly display groundwater chemistry averages above secondary drinking-water regulations.

The measured values for the water-quality parameters reported in this study closely agree with those reported previously for both local and regional studies of Ohio carbonate aquifers. It is not surprising that the water quality documented in this study is very similar to that reported in 1998 from well TW-6a (Table 11). The results from the present study are also similar to those Norris and Fidler (1973) reported (Table 11) in northwestern Delaware County. Finally, although the parameter means measured during this study are similar to the means reported in other studies of central Ohio (Table 11), S[O.sub.4.sup.2-] concentrations measured at the wellfield in 1998 were noticeably less than reported in these other studies (Table 11). In fact, the 1998 S[O.sub.4.sup.2-] concentrations measured at the Penry Wellfield were lower than 75% of the wells of carbonate aquifers included in the 2002 Ohio EPA 305(b) report. The distribution of a number of hydrogeochemical groundwater facies within the carbonate aquifer of Ohio, recently delineated by Eberts and George (2000) for the Midwestern Basins and Arches Aquifer System, probably accounts for the discrepancy of S[O.sub.4.sup.2-] concentrations measured between the Penry Wellfield and the Deering and others (1983) and both Ohio EPA (2000, 2002) 305(b) reports. Deering and others (1983) and the Ohio EPA 305(b) reports (2000 and 2002) generated parameter means by pooling data from more than one of the hydrogeochemical groundwater facies delineated by Eberts and George (2000). Based on the data collected in this study, the Penry groundwater corresponds to the Ca-Mg-HC[O.sub.3] hydrogeochemical groundwater facies. Although the carbonate-aquifer hydrochemical-facies map of Eberts and George (2000) shows the Penry Wellfield should reside within a Ca-Mg-S[O.sub.4] facies, an area characterized by the Ca-Mg-HC[O.sub.3] facies lies immediately to the north and west of the Penry Wellfield. Given that no control points existed in that study between this Ca-Mg-HC[O.sub.3] area to the north and west and the two control points within a Ca-Mg-S[O.sub.4] area to the south of the Penry Wellfield, the Ca-Mg-HC[O.sub.3] hydrogeochemical facies of the Penry Wellfield can easily be accommodated in the map by simply shifting the Ca-Mg-HC[O.sub.3] boundary southward by about 11 km. Eberts and George (2000) also noted that the "islands" of the Ca-Mg-HC[O.sub.3] facies situatedwithin thelarger Ca-Mg-S[O.sub.4] area in northern Ohio occur near areas of recharge. Close inspection of the potentiometric maps of Norris and Fidler (1973), Raymondi (1997), and Eberts and George (2000) does indicate that the Penry Wellfield is in such a recharge setting.

Within Study Phase Analyses

This investigation failed to document any significant links, in either a predictive or general sense, between water-quality parameters and geography, depth, time (sampling round), or combinations of these variables during the Background Phase of the study. This finding may be caused simply by the relatively small size (24.6 [km.sup.2]) of the study area, or the fact that geography, well depth, or flowpath do not correlate with parameter concentrations. A number of other investigations support the latter possibility. Eberts and George (2000) found a general absence of chemical evolution along flow paths in the Midwestern Basins and Arches Aquifer System and that aquifer lithology, rather than flow-path evolution, controlled hydrogeochemical facies. Recently, Chowdhury and others (2003), evaluating the groundwater in the Chippewa Creek Watershed in northern Ohio, observed that no appreciable trends in groundwater hydrogeochemistry existed across their study area, an area of similar size to the studied area surrounding the Penry Wellfield. With respect to influences of depth on groundwater chemistry, Eberts and George (2000) noted, for most of the Midwestern Basin and Arches Aquifer a similarity between groundwater from the shallow and deeper portions of the carbonate aquifer indicated that no depth trends exist. Similarly, other groundwater-quality data also indicate a lack of correlation between major cations and anions and depth, for example the 2000 EPA 305 (b) report showed that [Mn.sup.2+] concentrations in the carbonate aquifers of Ohio are invariant with respect to depth and that total dissolved solids and C1- concentrations, within the uppermost 122 m of the aquifer, exhibit only increasing variability with depth.

The water-quality parameters measured during the Pumping Phase of the investigation also failed to show any significant correlation with respect to geography, distance from pumping wellfield, depth, time (sampling round), or combinations of these variables. Additionally, changes in water-quality parameters did not correlate to water-level elevations, change in water level, length of open-bore exposure caused by pumping, or distance from the pumpingwellfield. It appears that the month-long Pumping Phase did not produce any significant correlations between water quality and these various independent variables. Similarly, water-quality parameters measured during the Recovery Phase failed to show any correlation to geography (west-to -east), distance from the wellfield, well depth, or combinations of these independent variables.

Among Study Phase Analyses

Parametric methods documented that only pH, alkalinity, and dissolved solids differed significantly among study phases. Although all pH values measured during the investigation were within the National Secondary Drinking Water Regulations of 6.5-8.5, the average pH of the Background Phase (6.99) differed significantly from the pH of the Pumping Phase (6.89) and the Recovery Phase (6.84). Interestingly, eight of the 11 wells that had elevated (more than one standard deviation above the mean) pH values during the Background Phase also experienced elevated (more than one standard deviation above the mean) pH levels during either or both of the other two (Pumping and Recovery) phases. Thus it appears that the wells are responding in a similar manner to the pumping. Alkalinity concentrations measured during the Pumping Phase were significantly lower (with a mean of 352 mg/L) than the Background Phase (371 mg/L). All six wells that experienced low (more than one standard deviation below the mean) alkalinity values during the Pumping Phase also experienced low (more than one standard deviation below the mean) values during the Background Phase. The Pumping Phase experienced significantly larger dissolved solids concentrations (639 mg/L) than either the Background Phase (499 mg/L) or the Recovery Phase (521 mg/L). All eight wells that experienced elevated (more than one standard deviation above the mean) dissolved solids concentrations during the Pumping Phase also had elevated (more than one standard deviation above the mean) concentrations during either or both of the other two (Background and Recovery) phases. These eight wells simply experienced larger concentrations more often during the Pumping Phase than the other two phases. The finding that dissolved solids concentrations showed a statistically significant increase during the Pumping Phase, but other related parameters such as hardness and specific conductance experienced no such differences, remains puzzling. Finally, non-parametric procedures revealed that median turbidity measured during the Background Phase (0.75 NTU) was significantly lower than measured during the Pumping Phase (1.48 NTU) or the Recovery Phase (1.36 NTU).

Since the present paper used a subset of the data presented in the Collector Wells International Inc. (2001) report, a comparison of results is warranted. Fortunately, the structure of these studies allows the comparison of four significant changes (pH, turbidity, alkalinity, and dissolved solids) between the Background and the Pumping Phases identified in the present paper to the results for those same comparisons in the Collector Wells International Inc. (2001) report. Because these studies approached data analysis differently, naturally their specific conclusions are of a slightly different nature, but not, with the possible exception of the pH results, incongruent with each other. As previously stated, the present paper found pH values of the Pumping Phase (6.89) to be significantly lower than the Background Phase (6.99); whereas the Collector Wells International Inc (2001) report simply stated that no appreciable differences existed between Background and Pumping phases and suggested that any variations in pH were probably within the precision range of the pH meter. Addressing the turbidity data of the Background Phase and the Pumping Phase, Collector Wells International Inc. (2001) concluded that the majority of wells did not experience substantial changes due to pumping, while the present paper found a significant difference in the median value of the Pumping Phase (1.48 NTU) compared the Background Phase (0.75 NTU). Collector Wells International Inc. (2001) used the duplicate sample results (Table 1), to erect threshold values for recognizing "real changes" induced by the pumping of the Penry Wellfield. Collector Wells International Inc. (2001) chose that threshold to be the value of the 90th percentile of the difference in the duplicate results. So, in order for wells to be considered as experiencing real differences due to pumping, dissolved solids concentrations needed to have changed by more than 77 mg/L and alkalinity concentrations needed to have changed by 16 mg/L or more. Using these thresholds, Collector Wells International Inc. concluded that all but three wells showed greater concentrations of dissolved solids during the Pumping Phase than Background Phase. This result is compatible with the findings of the present paper that found the Pumping Phase (639 mg/L) had significantly greater dissolved concentrations than the Background Phase (499 mg/L). Addressing alkalinity concentrations, Collector Wells International Inc. (2001) concluded that while 13 wells showed lower (by more than 16 mg/L) concentrations during the Pumping Phase and no wells showed increases greater than the threshold (16 mg/L), alkalinity concentrations did not change substantially in most of the wells during pumping. The present paper found that alkalinity concentrations of the Pumping Phase (352 mg/L) were significantly lower than during the Background Phase (371 rag/ L): a conclusion that is not necessarily statistically or logically in opposition with Collector Wells International Inc. (2001).

In conclusion, while it has been shown that a month-long pumping interval significantly affected pH, alkalinity, and dissolved solids concentrations of water within a 3.2 km radius of the Penry Wellfield, it did not cause those values to fall outside EPA guidelines. Although significant statistical differences indeed were found in pH values between the Background Phase and the other two phases, such small differences (0.10 and 0.15) in pH, from a practical standpoint, in all likelihood, will not cause or reflect substantial changes in water chemistry. While the background water quality already exceeded secondary regulations in many cases, it appears that the pumping of the Penry Wellfield did increase turbidity levels; however, these increased turbidity levels may be temporary as material mobilized due to drawdown may get flushed out of the well with time. Finally, the pumping did not cause [H.sub.2]S concentrations to rise above detectable levels.

ACKNOWLEDGMENTS. Bringing this study to fruition required the cooperation of many people with differing backgrounds, training, and points of view. I would like to express my appreciation to these people that assisted directly and indirectly in the completion of the study and this manuscript. I would like to thank Thomas Courtice (past-President of Ohio Wesleyan University) and Dick Fusch (Academic Dean, Ohio Wesleyan University) for introducing me to a number of city officials, including Tom Galitza (previous Public Utilities Director, City of Delaware). Next I would like to thank Sam Stowe (Vice President of Technical Services, Collector Wells International) and Brad Gamble (Collector Wells International) for overseeing the Penry Wellfield study. Thanks are also in order to Brad Gamble, Collector Wells International, and Rick Shamblen, Malcolm Pirnie Inc, for collecting samples during sampling round 9. I would like to convey my gratitude to Tom Marshall (Public Utilities Director, City of Delaware) for his evenhanded approach to the management of this project. Next, I wish to express my appreciation to Mark Hemans, Larry Sparling, and Amy Tovar (residents and Trustees of Troy Township and also members of the Technical Liaison Committee) for their diligence and equitable approach to the study. Through the efforts and composure of Tom Marshall, Mark Hemans, Larry Sparling, and Amy Tovar a potentially contentious issue was handled with aplomb and equanimity. I must also thank Richard Miller (Water Treatment Division Superintendent, Public Utilities, City of Delaware, Ohio), Paul Spahr (Hydrogeologist, Water Resource Section, Division of Water, ODNR) and the Delaware Department of Health for their assistance at various points during the investigation. I am truly indebted to Dan Hlavin, Andrew Durniat, Magan Panfil, Bryan Sams, Ariel Terranova-Web, and Amy Thwaite. These students proved to be dependable, knowledgeable, and a joy to work with--I thank you for your hard work and commitment. I would like to thank a number of colleagues at Ohio Wesleyan University: Scott Linder (Department of Mathematics and Computer Science) for his guidance and insightful comments regarding the statistical portion of this study and also Heather Grunkemeyer and Kim Lance (both of the Department of Chemistry) for various discussions concerning analytical laboratories. I also need to thank both Dorothy Carina and Barbara Williams (Ohio Wesleyan University) for helping with a variety of logistical tasks that enabled the sampling phase of the study to proceed smoothly. Most importantly, I would like to express my appreciation to the many residents of Troy Township who welcomed us to their homes to collect the data necessary to complete this study. Finally, I thank the reviewers for their insightful comments that improved the quality of the manuscript.

LITERATURE CITED

Allred BJ. 1999. Appendix A. fractured glacial till geotechnical engineering considerations. In: Haefner RJ editors. A Conference on Fractured Glacial Tills. Water Management Association of Ohio Spring Conference on Fractured Tills. 27 p.

Allred BJ. 2000. Survey of fractured glacial till geotechnical characteristics: hydraulic conductivity, consolidation, and shear strength. Ohio Journal of Science: 100(3/4):63-72.

Arihood LD. 1994. Hydrogeology and paths of flow in the carbonate bedrock aquifer, northeastern Indiana. Water Resources Bulletin 30(2):205-218.

Brockman CS, Szabo JP. 2000. Fractures and their distribution in the tills of Ohio. Ohio Journal of Science: 100(3/4):39-55.

Bugliosi EF. 1989. Ohio-Indiana carbonate-bedrock and glacial regional aquifer analysis--plan of study. In: Swain, LA and Johnson, AI. editors. Regional Aquifer Systems of the United States, Aquifers of the Midwestern Area: American Water Resources Association Monograph Series 13:135-148.

Bugliosi EF. 1990. Plan of study for the Ohio-Indiana carbonate-bedrock and glacial aquifer system. U.S.G.S. Open-File Report 90-151.25 p.

Casey GD. 1994. Hydrogeology of the Silurian and Devonian carbonate-rock aquifer system in the Midwestern Basins and Arches Region of Indiana, Ohio, Michigan, and Illinois. U.S. Geological Survey Open-File Report 93-663. 14 p.

Chowdhury SH, Iqbal MZ, Szabo JP. 2003. Comprehensive approach of groundwater resource evaluation: a case study in the Chippewa Creek watershed in Ohio. Ohio Journal of Science: 103(5):134-142.

Collector Wells International Inc. 1998. Hydrological evaluation for the development of an additional ground water source for the City of Delaware, Ohio, prepared for the City of Delaware, Ohio, Prepared by Collector Wells International, Inc., Columbus, Ohio.

Collector Wells International Inc. 2001. Penry Road wellfield evaluation, City of Delaware, Ohio. Prepared by Collector Wells International, Inc., Columbus, Ohio.

Deering MF, Mohr ET, Sypniewski BF, Carlson EH. 1983. Regional hydrogeochemical patterns in ground water of northwestern Ohio and their relation to Mississippi Valley-type Mineral occurrences. Journal of Geochemical Exploration 19:225-241.

Dumouchelle DH, Schiefer MC. 2002. Use of Streamflow records and basin characteristics to estimate ground-water recharge rates in Ohio. Bulletin --State of Ohio, Department of Natural Resources, Division of Water. 46, March 2002. 45 p.

Eberts SM, George LL. 2000. Regional ground-water flow and geochemistry in the Midwestern basins and arches aquifer system in parts of Indiana, Ohio, Michigan, and Illinois. U.S. Geological Survey Professional Paper: 1423-C. 103 p.

Fausey NR, Hall GF, Bigham JM, Allred BJ, Christy AD. 2000. Properties of the fractured till at the Madison County, Ohio, field workshop pit site. Ohio Journal of Science: 100(3/4): 107-112.

Grisak GE, Cherry JA, Von Hof JA, Blumele JP. 1976. Hydrogeologic and hydrochemical properties of fractured till in the interior plains region. In Legget RF editor. Glacial Till- An Interdisciplinary Study. The Royal Society of Canada, Ottawa, Special Publication No. 12:269-291.

Grisak GE, Pickens JF, Cherry JA. 1980. Solute transport through fractured media. 2. Column study of fractured till. Water Resources Research 16(4):731-739.

Ground Water Associates. 1991. Ground Water Resource Evaluation, Prepared for the City of Delaware, Ohio. Ground Water Associates, Inc, Westerville, Ohio.

Ground Water Associates. 1992. Phase II Hydrogeologic evaluation Delaware water treatment plant Delaware, Ohio, prepared for the City of Delaware Ohio. Ground Water Associates, Inc, Westerville, Ohio.

Hanover RH. 1994. Analysis of ground-water flow along a regional flow path of the Midwestern Basins and Arches Aquifer System in Ohio. USGS Water-resources Investigations Report 94-4105.29 p.

Harstine LJ. 1991. Hydrologic atlas for Ohio--average annual precipitation, temperature, streamflow, and water loss for a 50-year period 1931-1980. Ohio Department of Natural Resources, Water Inventory Report No. 28.

Keller CK, Van Der Kamp G, Cherry JA. 1988. Hydrogeology of two Saskatchewan tills, I. Fractures, bulk permeability, and spatial variability of downward flow. Journal of Hydrology 101:97-121.

Larsen GE, Schumacher GA, Shrake DL, Slucher ER, Swinford EM. 1992a. Preliminary structure contour map of the top of the Ordovician undifferentiated for west-central Ohio, Ohio Division of Geological Survey Open File Map 274.

Larsen GE, Schumacher GA, Shrake DL, Slucher ER, Swinford EM. 1992b. Preliminary structure contour map of the top of the Silurian Salina undifferentiated for west-central Ohio. Ohio Division of Geological Survey Open File Map 279.

Larsen GE, Schumacher GA, Shrake DL, Slucher ER, Swinford, EM. 1992c. Preliminary structure contour map of the top of the Devonian Delaware Limestone for west-central Ohio. Ohio Division of Geological Survey Open File Map 281.

McKay LD, Cherry JA, Gillham RW. 1993. Field experiments in a fractured clay till: 1. hydraulic conductivity and fracture aperture. Water Resources Research 29:1149-1162.

Norris SE, Fidler RE. 1973. Availability of water from Limestone and dolomite aquifers in southwest Ohio and the relation of water quality to the regional flow system. USGS Water-resources Investigations 17-73, 42 p.

Ohio EPA. 2000. Ohio's Ground Water Quality, 2000 305(b) Report, July 2000. Columbus, Ohio Environmental Protection Agency: Division of Drinking and Ground Waters.

Ohio EPA. 2002. Ohio's Ground Water Quality, 2002 305(b) Report, May 2003. Columbus, Ohio Environmental Protection Agency: Division of Drinking and Ground Waters.

Ohio EPA. 2006. 2006 305(b) Report Ohio's Ground Water Quality, June 2006. Columbus, Ohio Environmental Protection Agency: Division of Drinking and Ground Waters.

Prudic DE. 1986. Ground-water hydrology and subsurface migration of radionuclides at a commercial radioactive-waste burial site, West Valley, Cattaraugus County, New York. U.S. Geological Survey Professional Paper 1325, 83 p.

Raymondi RR. 1997. Aquifer tests in carbonate rocks overlain by glacial sediments in north-central Ohio. Ohio Journal of Science 97:24-29.

Shaver RH. 1985. Midwestern Basin and Arches Region--Correlation of stratigraphic units of North America (COSUNA) project: American Association of Petroleum Geologists.

Sparling DR. 1983. Conodont biostratigraphy and biofacies of Lower and Middle Devonain Limestones, North-central Ohio. Journal of Paleontology 57:825-864

Sparling DR. 1985. Correlation of the subsurface Lower and Middle Devonian of the Lake Erie region: alternative interpretation and reply, Geological Society of America Bulletin 96:1213-1220.

Strobel ML. 1993. Hydraulic properties of three types of glacial deposits in Ohio: U.S. Geological Survey Water-Resources Investigations Report 92-4135.41 p.

United States Environmental Protection Agency. 1994. National Primary Drinking Water Standards, EPA 810-F-94-001A.

United States Government. 2004. National Secondary Drinking Water Regulations. http://www.epa.gov/safewater/contaminants/index.htm (accessed 26 January 2007).

Weatherington-Rice J, Christy AD, Angle MP, Aller L. 2006. DRASTIC hydrogeologic settings modified for fractured till: part 1 theory. Ohio Journal of Science: 106(2):45-50.

Westgate LG. 1926. Geology of Delaware County. Geological Survey of Ohio, Fourth Series, Bulletin 30. 147 p.

Keith O. Mann (1), Department of Geology and Geography, Ohio Wesleyan University, Delaware, OH.

(1) Corresponding author: Keith O. Mann, Department of Geology and Geography, Ohio Wesleyan University, Delaware, OH 43015. Email: komann@owu.cdu

Table 1
Laboratory precision of MASI Environmental Laboratories:
values represent difference between duplicate analyses

Date Alkalinity Dissolved Hardness Iron
 (mg/L as Solids (mg/L as (mg/L)
 CaC[O.sub.3]) (mg/L) CaC[O.sub.3])

5 July 2000 12 3 186 0.01

6 July 2000 7 22 39 0.02

12 July 2000 2 0 8 0.08

22 Aug. 2000 2 90 12 0.01

23 Aug. 2000 5 48 10 0.04

24 Aug. 2000 2 20 27 0.36

20 Oct. 2000 2 2 4 0.01

21 Oct. 2000 14 74 4 0.00

22 Oct. 2000 10 12 0 0.02

23 Oct. 2000 12 34 0 0.94

9 Dec. 2000 18 24 30 0.03

10 Dec. 2000 8 32 10 0.01

11 Dec. 2000 6 4 2 0.00

20 April 2001 4 16 0 0.03

21 April 2001 2 64 8 0.11

22 April 2001 15 24 28 0.04

23 April 2001 5 72 36 0.16

27 April 2001 4 20 12 0.18

28 April 2001 8 132 0 0.01

29 April 2001 73 12 10 0.06

11 May 2001 8 12 16 0.09

12 May 2001 2 44 36 0.43

13 May 2001 16 16 4 0.10

29 May 2001 0 32 20 0.03

31 May 2001 11 60 4 0.12

31 May 2001 13 72 16 0.03

13 June 2001 4 8 24 0.00

13 June 2001 349 592 616 3.66

14 June 2001 19 738 0 0.53

15 June 2001 6 0 36 0.21

TABLE 2.
Medians (Mdn), means([bar.X]), and standard deviations (s)
of measurements made during individual sampling rounds.

Sampling pH Alkalinity
 (mg/L as
 CaC[O.sub.3])

Round Mdn [bar.X] s Mdn [bar.X] s

1 7.00 7.03 0.114 378 372 47.7

2 6.95 6.96 0.099 371 366 44.5

3 7.01 6.99 0.107 359 358 45.2

4 6.98 6.98 0.136 391 386 39.3

1--4 6.99 6.99 0.116 375 371 44.8

5 7.01 6.98 0.165 343 342 46.0

6 6.85 6.86 0.117 365 360 44.9

7 6.87 6.83 0.133 357 355 48.2

5--7 6.89 6.89 0.153 357 352 45.8

8 6.83 6.82 0.181 351 353 45.9

9 6.87 6.87 0.094 369 367 47.1

8--9 6.87 6.84 0.145 365 360 46.5

Sampling Specific Conductance Dissolved Solids
 ([mu]S/cm) (mg/L)

Round Mdn [bar.X] s Mdn [bar.X] s

1 880 885 149.5 542 528 91.3

2 790 806 140.6 586 593 90.3

3 800 833 145.8 454 464 87.9

4 790 816 176.8 432 414 125.1

1--4 820 835 154.8 503 499 119.5

5 850 862 142.2 640 625 144.5

6 830 845 140.5 638 601 124.4

7 880 880 173.5 660 690 152.0

5--7 850 862 152.4 644 639 144.8

8 770 788 161.1 490 523 143.6

9 950 942 187.8 560 518 146.0

8--9 830 865 189.7 528 521 143.3

Sampling Hardness Turbidity
 (mg/L as (NTU)
 CaC[O.sub.3])

Round Mdn [bar.X] s Mdn [bar.X] s

1 549 569 133.5 0.52 5.19 21.026

2 574 559 35.5 0.75 1.68 2.777

3 502 489 80.0 0.59 3.50 9.371

4 535 537 84.4 0.91 2.82 4.362

1--4 546 539 94.6 0.75 3.30 11.764

5 520 536 88.5 1.55 4.05 9.116

6 580 567 92.2 1.48 3.22 4.724

7 588 571 104.0 1.29 7.33 16.123

5--7 566 558 95.5 1.48 4.87 11.105

8 548 553 106.6 1.40 3.46 6.181

9 576 555 84.4 1.35 4.84 15.208

8--9 556 554 95.2 1.36 4.15 11.510

Sampling Iron
 (mg/L)

Round Mdn [bar.X] s

1 2.270 3.140 3.0301

2 2.455 2.772 1.1159

3 2.660 3.447 3.2050

4 2.700 2.963 1.3781

1--4 2.480 3.080 2.3524

5 2.330 2.725 1.1983

6 2.600 3.268 3.0670

7 2.430 3.519 3.1089

5--7 2.490 3.171 2.6179

8 2.470 3.144 2.3683

9 2.465 2.695 0.9380

8--9 2.468 2.920 1.7971

Table 3
Background Phase: adjusted [R.sup.2] values for linear
and multivariate regression analyses.

Sampling Model pH Alkalinity Specific
Round Conductance

1 Well Depth 0.05 0.09 -0.04

1 Geography -0.02 0.04 0.16

1 Geography & Well 0.02 0.11 0.13
 Depth

1 Geography, Well 0.06 0.10 0.17
 Depth, & (Geography
 x Depth)

1--4 Well Depth 0.08 0.16 <0.00

1--4 Geography -0.02 0.03 0.08

1--4 Time <0.00 <0.00 <0.00

1--4 Geography & Well 0.09 0.18 0.09
 Depth

1--4 Geography, Well 0.08 0.23 0.09
 Depth, & (Geography
 x Depth)

Sampling Model Dissolved Hardness
Round Solids

1 Well Depth 0.04 -0.04

1 Geography 0.22 0.29

1 Geography & Well 0.24 0.27
 Depth

1 Geography, Well 0.22 0.24
 Depth, & (Geography
 x Depth)

1--4 Well Depth <0.00 <0.00

1--4 Geography 0.09 0.16

1--4 Time 0.19 0.03

1--4 Geography & Well 0.07 0.17
 Depth

1--4 Geography, Well 0.18 0.17
 Depth, & (Geography
 x Depth)

Sampling Model Turbidity Iron
Round

1 Well Depth -0.04 -0.03

1 Geography -0.02 -0.04

1 Geography & Well -0.05 -0.08
 Depth

1 Geography, Well -0.09 -0.12
 Depth, & (Geography
 x Depth)

1--4 Well Depth <0.00 <0.00

1--4 Geography -0.02 <0.00

1--4 Time <0.00 -0.01

1--4 Geography & Well <0.00 -0.01
 Depth

1--4 Geography, Well <0.00 -0.01
 Depth, & (Geography
 x Depth)

Table 4
Pumping Phase: adjusted [R.sup.2] values for linear
and multivariate regression analyses.

Sampling Model pH Alkalinity Specific
Round Conductance

7 Well Depth -0.04 0.21 -0.04

7 Distance -0.04 0.01 0.17

7 Distance & Well Depth 0.08 0.24 0.13

7 Distance, Depth, & -0.12 0.22 0.29
 (Dist. x Depth)

5--7 Well Depth 0.01 0.20 -0.01

5--7 Distance <0.00 0.03 0.09

5--7 Time (sampling round) -0.09 <0.00 -0.01

5--7 Distance & Well Depth -0.01 0.25 0.08

5--7 Distance, Depth, & 0.01 0.27 0.24
 (Distance. x Depth)

5--7 Distance, Depth, & 0.14 0.25 0.07
 Time (sampling round)

Sampling Model Dissolved Hardness
Round Solids

7 Well Depth -0.04 -0.02

7 Distance 0.16 0.16

7 Distance & Well Depth 0.12 0.15

7 Distance, Depth, & 0.26 0.29
 (Dist. x Depth)

5--7 Well Depth -0.01 0.03

5--7 Distance 0.08 0.13

5--7 Time (sampling round) 0.03 0.01

5--7 Distance & Well Depth 0.08 0.16

5--7 Distance, Depth, & 0.16 0.30
 (Distance. x Depth)

5--7 Distance, Depth, & 0.11 0.18
 Time (sampling round)

Sampling Model Turbidity Iron
Round

7 Well Depth 0.13 -0.02

7 Distance -0.02 -0.04

7 Distance & Well Depth -0.12 -0.06

7 Distance, Depth, & 0.18 -0.10
 (Dist. x Depth)

5--7 Well Depth 0.08 -0.01

5--7 Distance -0.01 0.01

5--7 Time (sampling round) 0.01 <0.00

5--7 Distance & Well Depth 0.06 <0.00

5--7 Distance, Depth, & 0.07 0.01
 (Distance. x Depth)

5--7 Distance, Depth, & 0.07 0.04
 Time (sampling round)

Table 5
Round 7: adjusted [R.sup.2] values fir linear regression.

Model [DELTA] [DELTA] [DELTA] [DELTA]
 pH Alkalinity Specific Dissolved
 Conductance Solids

Water level 0.04 -0.04 0.14 0.14
elevation

[DELTA] Water 0.02 -0.04 0.18 0.10
level

Open-bore 0.07 0.07 0.16 -0.14
exposure

Distance to Well -0.04 -0.04 0.21 0.15
TW-5

Model [DELTA] [DELTA] [DELTA]
 Hardness Turbidity Iron

Water level 0.12 -0.03 -0.03
elevation

[DELTA] Water 0.12 0.07 -0.04
level

Open-bore -0.06 0.14 0.03
exposure

Distance to Well 0.07 0.01 -0.04
TW-5

Table 6
Recovery Phase: adjusted [R.sup.2] values for linear
and multivariate regression analyses.

Sampling Model pH Alkalinity Specific
Round Conductance

9 Well Depth 0.11 0.30 0.04

9 Distance -0.04 <0.00 0.22

9 Geography -0.03 <0.00 0.29

9 Distance & Well 0.07 0.34 0.20
 Depth

9 Distance, Depth, 0.02 0.35 0.27
 & (Disc. x Depth)

9 Geography & Well 0.07 0.29 0.26
 Depth

9 Geography, Well 0.18 0.33 0.37
 Depth, & (Geog.
 x Depth)

8--9 Well Depth 0.07 0.26 -0.02

8--9 Distance 0.07 0.02 0.13

8--9 Geography 0.12 <0.00 0.13

8--9 Distance & Well 0.15 0.29 0.12
 Depth
8--9 Distance, Depth, & 0.18 0.31 0.20
 (Distance x Depth)

8--9 Geography & Well 0.16 0.25 0.12
 Depth
8--9 Geography, Well 0.18 0.33 0.21
 Depth, & (Geog.
 x Depth)

Sampling Model Dissolved Hardness
Round Solids

9 Well Depth 0.01 0.00

9 Distance <0.00 0.20

9 Geography <0.00 0.28

9 Distance & Well 0.02 0.23
 Depth

9 Distance, Depth, -0.02 0.41
 & (Disc. x Depth)

9 Geography & Well <0.00 0.26
 Depth

9 Geography, Well 0.10 0.28
 Depth, & (Geog.
 x Depth)

8--9 Well Depth -0.01 -0.02

8--9 Distance 0.10 0.22

8--9 Geography 0.08 0.20

8--9 Distance & Well 0.09 0.20
 Depth
8--9 Distance, Depth, & 0.11 0.44
 (Distance x Depth)

8--9 Geography & Well 0.06 0.18
 Depth
8--9 Geography, Well 0.13 0.26
 Depth, & (Geog.
 x Depth)

Sampling Model Turbidity Iron
Round

9 Well Depth 0.03 -0.04

9 Distance 0.01 0.01

9 Geography <0.00 0.05

9 Distance & Well -0.02 -0.03
 Depth

9 Distance, Depth, -0.07 -0.01
 & (Disc. x Depth)

9 Geography & Well -0.03 0.01
 Depth

9 Geography, Well -0.05 <0.00
 Depth, & (Geog.
 x Depth)

8--9 Well Depth -0.01 -0.02

8--9 Distance 0.01 0.04

8--9 Geography 0.01 0.01

8--9 Distance & Well <0.00 0.02
 Depth
8--9 Distance, Depth, & 0.03 0.07
 (Distance x Depth)

8--9 Geography & Well <0.00 -0.01
 Depth
8--9 Geography, Well 0.01 0.02
 Depth, & (Geog.
 x Depth)

Table 7
Significance probabilities for ANOVA analyses addressing
geographic trends.

Variable Background Phase
 (West-East)

 Round 1 Round 2 Round 3 Round 4

pH 0.4988 0.4870 0.0243 0.8199

Alkalinity 0.1732 0.1903 0.4774 0.5510

Specific Conductance 0.0257 0.0588 0.0384 0.8503

Dissolved Solids 0.0112 0.1683 0.0290 0.0999

Hardness 0.0030 0.1525 0.0168 0.0639

Turbidity 0.4350 0.4383 0.4397 0.5531

Fe 0.9256 0.0607 0.9541 0.2954

Variable Pumping Phase
 (Distance from TW-5)

 Round 5 Round 6 Round 7

pH 0.3576 0.2837 0.8188

Alkalinity 0.3270 0.2876 0.2669

Specific Conductance 0.3469 0.1505 0.0235

Dissolved Solids 0.5404 0.0620 0.0280

Hardness 0.1406 0.0539 0.0287

Turbidity 0.3732 0.9533 0.4899

Fe 0.1179 0.1860 0.9233

Variable Recovery Phase
 (West-East)

 Round 8 Round 9

pH 0.0064 0.5667

Alkalinity 0.6371 0.3559

Specific Conductance 0.2022 0.0033

Dissolved Solids 0.0304 0.3453

Hardness 0.0509 0.0039 *

Turbidity 0.3287 0.3372

Fe 0.4561 0.1574

* = Significant difference, using the Bonferoni
Correction (K= 7 and [[alpha].sub.i] = 0.007)
calculated for each sampling round. Underlined
values = trend opposite to what would be expected
given increasing mineralization along flow paths.

TABLE 8
ANOVA table for parameters measured during background,
pumping, and recovery phases.

Parameter Source d.f. Sum of Squares

pH Between Phases 2 0.8370209
 Error 220 4.0499065
 Total 222 4.8869274
Specific Between Phases 2 44,302.4
Conductance
 Error 220 5,783,438.1
 Total 222 5,827,740.5
Turbidity Between Phases 2 104.026
 Error 220 28,919.735
 Total 222 29,023.761
Alkalinity Between Phases 2 14,251.19
 Error 218 455,657.31
 Total 220 469,908.50
Hardness Between Phases 2 17,581.4
 Error 220 1,981,050.5
 Total 222 1,998,631.9
Dissolved Between Phases 2 888,830.33
Solids Error 220 3,925,710.10
 Total 222 4,814,540.40
Iron Between Phases 2 1.8953
 Error 220 1,195.4410
 Total 222 1,197.3362

Parameter Mean Square F Ratio Prob > F

pH 0.418510 22.7344 <0.0001 *
 0.018409

Specific 22,151.2 0.8426 0.4320
Conductance
 26,288.4

Turbidity 52.013 0.3957 0.6737
 131.453

Alkalinity 7,125.59 3.4091 0.0348 *
 2,090.17

Hardness 8,790.70 0.9762 0.3784
 9,004.77

Dissolved 444,415 24.9054 <0.0001 *
Solids 17,844

Iron 0.94763 0.1744 0.8401
 5.43382

df: degrees of freedom

* Significant (a [alpha] = 0.05)

Table 9

Student's t-tests of pH, alkalinity, and dissolved solids among
sampling phases

Parameter Phase Pumping Recovery

pH Background <0.0000 * <0.0000 *
 Pumping -- 0.0971
Alkalinity Background 0.0079 * 0.1579
 Pumping -- 0.3707
Dissolved Background <0.0000 * 0.3365
Solids Pumping -- <0.0000 *

* Significant using Bonferroni Correction (K = 3, [alpha] = 0.016)

Table 10
Wilcoxon Test for turbidity among study phases.

Phase Pumping Recovery

Background 0.001 * 0.006 *
Pumping -- 0.778

* Significant using Bonferroni Correction (K = 3, [alpha] = 0.016)

Table 11
Water-quality parameters of central Ohio.

 This Collector Collector
 Study Wells Wells
 1-4 Int. 1998 Int. 1998
 Mean Well TW-4 Well TW-6a

Depth n/a 215 235

Temperature [degrees] C 13.7 -- 11.4

pH 6.99 6.80 6.99

Alkalinity 371 440 391
(mg/L as CaC[O.sub.3])

Specific conductance 835 -- 695
([micro]S/cm)

Dissolved solids (mg/L) 499 628 585

Hardness 539 530 480
(mg/L as CaC[O.sub.3])

[Ca.sup.2] (mg/L) -- 130 120

[Mg.sup.2+] (mg/L) -- 51 44

[Na.sup.+] (mg/L) -- 8.3 7.00

Fe total (mg/L) 3.100 2.200 1.100

Cl (mg/L) -- 9 7

S[O.sub.4.sup.2-] (mg/L) -- 120 120

 Deering
 Norris and and
 Fidler others. Ohio EPA
 1973 1988 2000
 Well CP-13 Mean Mean

Depth 350 -- --

Temperature [degrees] C 11 -- 13.4

pH 7.7 7.3 7.28

Alkalinity -- -- 302
(mg/L as CaC[O.sub.3])

Specific conductance 1000 1037 781
([micro]S/cm)

Dissolved solids (mg/L) 684 * -- 667

Hardness 550 -- 492
(mg/L as CaC[O.sub.3])

[Ca.sup.2] (mg/L) 140 155 117.1

[Mg.sup.2+] (mg/L) 48 62 48.4

[Na.sup.+] (mg/L) 21 43 32.4

Fe total (mg/L) 5.8 1.08 1.4410

Cl (mg/L) 17 19 23.0

S[O.sub.4.sup.2-] (mg/L) 250 302 216.9

 Ohio EPA Ohio EPA
 2002 2006
 Mean Mean

Depth -- --

Temperature [degrees] C 13.4 13.5

pH 7.25 7.25

Alkalinity 292 294
(mg/L as CaC[O.sub.3])

Specific conductance 850 889
([micro]S/cm)

Dissolved solids (mg/L) 769 771

Hardness 554 540
(mg/L as CaC[O.sub.3])

[Ca.sup.2] (mg/L) 136.4 133.0

[Mg.sup.2+] (mg/L) 50.6 50.8

[Na.sup.+] (mg/L) 35.3 36.8

Fe total (mg/L) 1.0041 1.230

Cl (mg/L) 27.9 28.4

S[O.sub.4.sup.2-] (mg/L) 280.7 284.7

* = calculated in study