Final Report
June 2006

Can Gravel Mining and Water Supply Wells Coexist?

Executive Summary • Introduction • Problem Statement • Objectives • Methods • Results (field observations, lab results, statistical analysis) • Discussion • Acknowledgements • Appendices

Methods
The project started with a compilation of geological maps, gravel pit locations, and well locations from published reports and State of Maine files (Department of Environmental Protection, Maine Geological Survey, and Drinking Water Program). Many of the recent data publications were geo-referenced and available in a GIS format. The inventory first addressed public water supplies because of the larger number of private wells. Also, public water supplies have more consequences per supply well since many people are served from the same source. The evaluation of private wells was greatly assisted by previous work conducted by concerned citizens in Lamoine.

The objective of the well evaluation was to determine the number of wells that could be affected by gravel mining and to establish a network of monitoring points across the mapped sand and gravel aquifer. Since the limits of the aquifer are not always precisely located, all private or public wells within 0.4 km (0.25 miles) of the mapped aquifer boundary were included. Wells were located by overlaying tax maps onto the sand and gravel aquifer maps. Owners of lots within the targeted areas were contacted by telephone, or mail, through the assistance of volunteers in the towns of Ellsworth, Hancock, and Lamoine. Land ownership was verified from local tax records.

Once land ownership and uses were determined, the lots were checked for well locations. Many rural lots were found to be undeveloped, while some lots had both older dug wells and newer drilled bedrock wells. Owners were surveyed by interviews to collect information on: well construction, well age, history of water-related complaints or concerns, and the availability of water quality testing. The actual wellhead locations were measured with a Trimble R3 GPS system, with a resolution <3 meters for differential positioning. Depth to water level was measured to the nearest 0.01 meter (0.02 feet) using a Solinst Model 101 water level meter. Two easily accessible wells were selected as reference wells so that data collected on different days could be compared.

Gravel pits located on the aquifer were also verified in the field, distinctive landmarks or pit centers were measured using the Trimble GPS system, and photographed. Owners were surveyed by interview to collect historical and current uses of gravel pits. At the time of this study, there were twenty-three active gravel pits identified and eleven separate owners in this sand and gravel-aquifer system.

The budget for water quality testing was small. In order to maximize our ability to re-sample under similar condition, water quality testing was prioritized to springs. Springs were selected as natural points of water discharge from the aquifer. Springs were located from local information sources, maps, and field exploration. A total of seven springs were located and sampled. In addition two seepage ponds located within the sand and gravel deposit (Simmons and Blunts Ponds) were included in this study. A seepage pond has now inlet and sometimes no outlet, all of its water comes from groundwater and precipitation. Streams were sampled when no other source (spring or seepage pond) was available. Water samples were collected during the winter when the ground was frozen to make sure that the water collected was groundwater and not recent precipitation. Water samples were analyzed for the following parameters:

• water temperature (measured in the field);
• pH (acidity);
• conductivity (a measure of how much matter is dissolved in the water);
• calcium (major natural ion in water);
• magnesium (major natural ion in water);
• sodium (a natural ion in water or a contaminant from either septic systems or road salt);
• chloride (a natural ion from sea salts or from road salt);
• nitrate (a nutrient from fertilizers or septic systems);
• sulfate (naturally occurring ion, may also be a pollutant); and
• dissolved organic carbon (a measure of organic matter, clean groundwater should have concentrations less than 1 to 2 ppm).

Water temperature was measured in-situ and water was collected into one pre-cleaned 500 mL HDPE bottle and two 40 mL glass VOA vials. The water samples collected were analyzed at the Watershed Research Laboratory at the University of Maine.

The water quality results were tested for an effect due to closeness to gravel pits using the non-parametric Kruskal-Wallace test. The Kruskal-Wallace test was used to determine if any of the chemical parameters were statistically different based upon sample location (defined as greater than or less than one kilometer from a gravel pit). Care must be taken in interpreting statistical test results because the number of samples analyzed was small. Significance for this study has been set at α<0.1 (90% level). A correlation test was used to determine how the chemical parameters varied with respect to each other. The correlation value (Pearson r) is calculated for any two variables tested. For example, if two parameters both change in the same manner they will have a high correlation r-value. In this study, two parameters were considered to correlate if greater than 50% of the variation was similar (where the Pearson correlation value, r was >0.5).