The New Jersey Department of Environmental Protection (NJDEP) Well Head Protection Program, developed in response to the 1986 Federal Safe Drinking Water Act Amendments, requires delineation of Well Head Protection Areas (WHPA's), commonly called contributing areas, for all public and non-community water-supply wells in New Jersey. Typically, WHPA's for public community water-supply wells in New Jersey are delineated using a two-dimensional ground-water flow model incorporating the regional hydraulic gradient; however, NJDEP guidelines allow for the use of a three-dimensional flow model to delineate contributing areas to wells in complex hydrogeologic settings.
The Puchack well field in Pennsauken Township, Camden County, N.J., is an area of strong hydraulic connection between the Lower aquifer of the Potomac-Raritan-Magothy aquifer system and the Delaware River. Interactions among and within the public-supply well fields in the area are complex.
To delineate the contributing area to the Puchack well field, the U.S. Geological Survey, in cooperation with the NJDEP, developed an 11-layer ground-water flow model of the Potomac-Raritan-Magothy aquifer system in the Pennsauken Township area to simulate flow in the vicinity of the well field. The model incorporates the interaction between the aquifer system and the Delaware River, and includes boundary flows from an existing regional model of the Camden area. Recharge used in the model ranged from 4.5 to 14 inches per year, and horizontal hydraulic conductivity ranged from 50 to 250 feet per day. Values of vertical hydraulic conductivity ranging from 0.001 to 0.5 feet per day were assigned to zones created on the basis of variations in hydrogeologic conditions observed in geophysical logs from wells.
A steady-state simulation was used to calibrate the model to synoptic water-level data collected in March 1998. Near the Puchack well field, simulated heads generally were within 1 foot of the measured heads in both the Middle and Lower aquifers. Simulated water-level differences across the confining units at most of the nested wells were within ? 0.5 feet of the differences calculated from measured water levels.
The existing flow model was modified to meet NJDEP guidelines for delineating contributing areas in complex hydrogeologic settings. These modifications included rediscretizing the model grid to a finer grid and preparing the water-use data set for use in the rediscretized model. The contributing area to the Puchack well field was delineated by means of particle tracking.
An uncertainty analysis was conducted in which 36 model-input parameters were both increased and decreased until the resulting change in simulated heads exceeded the model-calibration criterion of ? 5 feet at any model cell. Porosity most affected the size and shape of the contributing area. The distribution of withdrawals at the Morris/Delair well field and variations in recharge affected both the size and shape of contributing area to the Puchack well field and the source of water to the Puchack wells.
The results of the uncertainty analysis were combined to determine the 'aggregate' contributing area to the Puchack well field--a composite of areas on the land surface that contributed flow to the Puchack well field in less than 12 years in any uncertainty simulation. The shape of the aggregate contributing area was most similar to that associated with a reduction in porosity, which indirectly affected the size and shape of the contributing areas by changing travel time.