Reports
Simulated hydrologic effects of the development of the Patapsco aquifer system in Glen Burnie, Anne Arundel County, Maryland
1991, Achmad, G.
Report of Investigations 54
Abstract
The lower Patapsco aquifer is an important source of ground water in the Glen Burnie area of Anne Arundel County, Maryland, an area of increasing population growth. The ground-water withdrawal rate from the lower Patapsco aquifer increased from 3.80 million gallons per day (Mgal/d) in 1965 to about 9.70 Mgal/d in 1985. Water-level measurements in the area indicate a decline in areal water levels of 6 to 29 feet in the lower Patapsco aquifer during the twenty years. The amount of drawdown in the upper Patapsco aquifer during 1965-85 is estimated to be about 5 to 10 feet. For the period 1944-52 the average base flow for Sawmill Creek basin, derived from streamflow hydrographs measured at a Glen Burnie gaging station, was 7.4 cubic feet per second. The stream gage was reestablished in 1984-85 and appreciably lower base flows were observed (3.4 to 0.28 cubic feet per second).
The Patapsco aquifer system attains a thickness of about 250 feet in the Glen Burnie area. In the vicinity of Sawmill and Marley Creeks, the upper Patapsco aquifer is under unconfined conditions and is separated from the lower Patapsco aquifer by a clayey confining unit. Water in the upper Patapsco aquifer flows from the northern and western recharge areas toward Sawmill and Marley Creeks and the Curtis Bay tributaries . Vertical leakage occurs from the upper Patapsco aquifer downward to the lower Patapsco aquifer in the northwestern recharge areas, and from the lower Patapsco aquifer upward to the upper Patapsco aquifer near the areas of tidal surface water of the Chesapeake Bay tributaries.
In order to assess the ground-water resources of the Glen Burnie area, a ground-water-flow model of the Patapsco aquifer system was constructed. The model area encompasses about 40 square miles, which include the drainage basins of Sawmill Creek and Marley Creek. The hydrologic properties of the aquifer system were acquired from an earlier study and were reevaluated and combined with other hydrologic and climatological data obtained during this investigation. In the model, transmissivity values of the lower Patapsco aquifer range from 1,500 to 6,000 feet squared per day, with the high values being assigned to the southeastern downdip areas where aquifer tests indicate high transmissivity and the aquifer tends to be relatively thick . The upper Patapsco aquifer was assumed to be homogeneous and its average hydraulic conductivity was estimated to be 33 feet per day. The confining-unit leakance ranged from 10-8 second-1 in the model area to 10-6 second-1 in the areas of tidal surface water where the confining unit consists of more sandy material. A long-term average recharge rate of 20.5 inches per year was used throughout most of the simulation period, with the exception of 1984-85, which was assigned a recharge rate of 14.4 inches per year. The lower recharge rate reflects a precipitation deficiency of about 9.6 inches per year in the Glen Burnie area during June 1984 to May 1985. Average precipitation for 1955-85 was 41.5 inches per year. The steady-state simulation indicates that the effective ground-water evapotranspiration rate for the Sawmill Creek and Marley Creek basins was 4.59 inches per year.
The model was calibrated under steady-state and transient conditions by matching the simulated water levels and stream base flows with those measured in the field . Ninety percent of the water levels match with differences of less than 12 feet ; the maximum difference is 14 feet and the mean difference is 6 feet. Eighty-five percent of the base flows match with differences less than 0.46 cubic foot per second, and the maximum difference is 0.6 cubic foot per second. The calibrated model was subsequently used to simulate hydrologic effects of increased pumpages on ground-water levels and stream base flows.
An evaluation of the ground-water-supply potential of the Patapsco aquifer system was made by using the results of the model simulations. The following three constraints on additional pumpage were considered: (1) The possibility of dewatering Sawmill Creek at Glen Burnie due to the effects of additional pumpage on stream base flow; (2) the possibility of saltwater intrusion due to ground-water levels declining below sea level in the vicinity of areas of tidal surface water; and (3) the possibility of dewatering the aquifer in a well field due to excessive drawdown from pumping wells.
Simulations that increase pumpage by 0.98, 1.95, 2.93, and 3.58 Mgal/d over the 1985 maximum of 11.57 Mgal/d indicate that the base flow of Sawmill Creek at Glen Burnie would be reduced to 0.67, 0.40, 0.08, and 0.001 cubic feet per second, respectively. A simulation of an increase in pumpage of 4.55 Mgal/d over the 1985 maximum (1l.57 Mgal/d) suggests that areal ground-water levels in the Glendale and Harundale areas would be lowered to about sea level, a condition that could lead to saltwater intrusion from the Chesapeake Bay and Curtis Bay tributaries. In all the scenarios, simulated drawdown expressed as a cell average does not cause water levels to decline below the top of the aquifer, which suggests that the water-yielding capacity of the aquifer is not exceeded. At the operational level, however, several scenarios simulating an additional pumpage of 0.32 to 0.97 Mgal/d at the Dorsey Road well field produced drawdowns that cause water levels to decline below the top of the aquifers in at least one pumping well (Bd 95). Likewise, the simulated water level in a hypothetical pumping well at the Thelma Avenue well field declined 5 feet below the top of the aquifer when pumpage is increased from 0.006 to 0.91 Mgal/d. Both of these cases suggest that additional wells, adequately spaced to minimize interference effects, would be needed to increase production at these well fields. In general, allocating additional pumpage to well fields in the southwestern part of the model area would reduce the adverse effects of the conditions cited above.
The simulation of a hypothetical 2-year drought, assuming 60 percent of the long-term average recharge, demonstrates that even with pumpage held to the 1985 maximum (11.57 Mgal/d), water-level declines are sufficient to dewater Sawmill Creek at Glen Burnie. The effects of the drought, however, are temporary and reversible once average recharge conditions are reimposed.
