¹ Agency for Toxic Substances and Disease Registry, Atlanta, Georgia

² Georgia Institute of Technology, Atlanta, Georgia

³ Because of numerical approximation and roundoff, contribution of water from all wells and well fields may sum to slightly less or slightly more than 100%; see text for complete details.

⁴ In this report some maps are shown in reduced size as figures in the text. However, all maps are provided as full-size plates under separate cover.

⁵ The number of customers refers to the number of water-utility connections for metering and billing purposes. All of the population was not necessarily serviced by the water utility; some of the population obtained their potable water from privately owned groundwater wells—see “Background” section of report for additional details.

⁶ In a water-distribution system, consumption should equal production if there are no losses through leaks, pipe breaks, or non-metered consumption. The water utility estimated that annual losses in the UWTR system were less than 10% of total production (ATSDR 1999, p. 31). For the purpose of the historical reconstruction analysis, and the intended use of model simulations, these losses were considered negligible.

⁷ The EPANET 2 model uses the “Node-Link” concept to represent pipeline junctions and segments associated with a pipeline network. In EPANET 2 terminology, pipeline junctions or model nodes are used to represent the end points of a section of pipeline and a link is used to represent the length of a pipeline section.

⁸ In some water-distribution system analyses, the terms consumption and demand are used interchangeably. In this report, however, consumption will refer to those data derived from direct metering of either groundwater production or customer usage of water. Demand will refer to the fractional component of consumption that is applied to the EPANET 2 model at pipeline node locations.

⁹ Each node in the pipeline network is not necessarily assigned a demand value that is derived from water-distribution system production. Some nodes do not have an associated demand value because of their location in the pipeline network (zero-demand value assigned in EPANET 2). Other nodes are used to represent groundwater-well production and supply to the water-distribution system (negative-demand value assigned in EPANET 2). For the 35 annual pipeline networks used for the historical reconstruction analysis (Plates 3 through 37), the percentage of positive-demand nodes (those nodes to which a component of monthly consumption was spatially distributed) relative to the total number of nodes in the pipeline network was about 90%.

¹⁰ A spatial join is defined as the merging of records and attributes for unrelated yet overlapping databases (Clarke 1999).

¹¹ This assumption—that system operations over a month-long time period could be represented by a “typical” 24-hour operating schedule—will be tested in the “Sensitivity Analysis” section of the report.

¹² See the EPANET 2 Users Manual for a description of pattern factors and Maslia et al. (2000a) for a description of how the EPANET 2 pattern factors were applied to the present-day (1998) water-distribution system serving the Dover Township area.

¹³ The term treatment plant is used by the water utility to identify all distribution-system facilities associated with a particular point of entry such as wells, storage tanks, water treatment, and high-service or booster pumps.

¹⁴ For purposes of modeling, water treatment, such as the type listed in Table 2, was not included in the distribution system.

¹⁵ This assumes that the net change in a storage tank over a 24-hour period is zero because the starting water level (at hour 0) must equal the ending water level (at hour 24) in accordance with the “Master Operating Criteria” (Table 4).

¹⁶ Hourly-specific information is defined as written or digital information that describes an hourly schedule by which water-utility operators control the on-and-off cycling of wells and high-service and booster pumps.

¹⁷ This approach was also suggested by the external expert panel (ATSDR 2001e)—see “Background” section.

¹⁸ Results are shown for all model nodes (pipeline junctions) with simulated proportionate contribution equal to or greater than 1%. For values of less than 1%, results are not shown.

¹⁹ Data files included with this report on CD-ROM represent the digital (or electronic) results shown on Plates 52 through 153. Contained in the data files are the values of simulated proportionate contribution of water from each operating well or well field to all model nodes. These results were obtained using the manual adjustment process. The files are prepared in text, Excel, and DBF formats.

²⁰ In this section, a summary of the analyses conducted for the historical period of January 1962–December 1996 is presented. More in depth and detailed analyses of results for the seven selected year—1962, 1965, 1971, 1978, 1988, 1995, and 1996—are presented in the next section of the report, “Review of Simulation Results for Selected Years and Months.”

²¹ For well fields that have multiple wells, such as Holly, South Toms River, Parkway, and Berkeley, see Appendix B for information on specific wells in operation during the historical period of analysis.

²² For more exact proportionate contribution results for the seven selected years—1962, 1965, 1971, 1978, 1988, 1995, and 1996—readers should refer to Tables G-1 through G-7, located in Appendix G.

²³ Proportionate contribution results for any month of the historical period (January 1962—December 1996) can be obtained by conducting a source-trace analysis using the appropriate monthly input data file and the EPANET 2 program included with this report on the accompanying CD-ROMs. Readers should refer to Tables F-1 through F-7 (Appendix F) for source-node identifications needed to be used with the input data files, and conduct the simulations according to the description provided in the “Water-Quality Modeling (Source-Trace Analysis)” section of this report.

²⁴ As previously described in the section on “Hydraulic Modeling”, supply nodes were used as a surrogate method (the SNL simulation method, Figure 19B) to represent wells and storage tanks linked to high-service and booster pumps (the WSTP simulation method, Figure 19A). Therefore, in EPANET 2, pattern factors were assigned to wells discharging directly to the distribution system and to supply nodes representing wells and storage tanks linked to high-service and booster pumps to describe a 24-hour operating schedule.

²⁵ The GA optimization approach was previously described in the section on “Hydraulic Modeling.” Appendix E provides details of the development of the methodology and its application to the operation of water-distribution systems.

²⁶ See Maslia et al. (2000a, p. 55) for a discussion of stationary water-quality dynamics (“dynamic equilibrium”) for the water-distribution system serving the Dover Township area.

²⁷ The use of the “stacked” column graph format for presenting simulated proportionate contribution results was described in “Historical Reconstruction Analysis” section of this report.

²⁸ Representation of nodes used to simulate wells linked to storage tanks and high-service and booster pumps as shown in Figure 19.

²⁹ Exact values for the pattern factors can be obtained from the appropriate EPANET 2 input data file provided with this report on the CD-ROMs.

³⁰ Study locations correspond to model node locations and were selected and provided by the New Jersey Department of Health and Senior Services to ATSDR without personal identifiers and status.