56
H. Rubin et al.
The coastal plain water-supply aquifer of Israel was first discovered to be
contaminated with kerosene in 1983. Site characterization studies (Kanfi 1986)
found a kerosene layer of up to 70 cm floating on top of the groundwater table,
which at that time was at an elevation of about 2 m below sea level. Since the
discovery of the kerosene contamination, the groundwater table has been elevated
by recharge to a height of about 4 m above sea level (6 m elevation increase) to
avoid saltwater intrusion into the coastal aquifer. Coinciding with the increased
elevation of the groundwater table was the nearly complete disappearance of the
formerly floating kerosene layer. Recent soil core information indicates that a
majority of the kerosene is located below the water table, entrapped within the
permeable blocks of the sandstone.
The integrity of water-supply aquifers is often threatened by contamination
from nonaqueous phase liquids (National Research Council 1994). Fractured
permeable water-supply aquifers are not immune to these hazards, as exemplified
by the kerosene contamination of the coastal plain aquifer in Israel, causing the
abandonment of several water-supply wells (Kanfi 1986). Field evidence from the
coastal plain aquifer is consistent with experimental studies of nonaqueous phase
liquid (NAPL) migration in an artificial fractured permeable formation (Schwille
1981, 1988). The experimental results showed that a substantial proportion of
NAPL can. flow through both fractures and permeable blocks, but that fractures
offer minimal NAPL storage following NAPL drainage and redistribution. Other
experimental work in homogeneous consolidated and unconsolidated materials has
shown that, following redistribution, the NAPL is eventually entrapped as
disconnected ganglia (blobs) encompassing one or more porebodies (Chatiz et al.
1983; Powers et al. 1991). Based on inference from these field and laboratory
studies, a conceptual model of NAPL contamination in a fractured permeable
formation is developed and shown in Fig. 1. The entrapped immobile residual
NAPL is assumed to be present in the permeable blocks as disconnected ganglia
filling one or more porebodies and NAPL storage in the fractures is considered
negligible.
Entrapped NAPL residual is a source of long-term chronic
groundwater contamination because it cannot be effectively mobilized or easily
dissolved, due to capillary forces and its generally low aqueous solubility (Mackay
et al. 1985; Mercer and Cohen 1990).
Theoretical and practical case studies have shown that traditional pump-andtreat methods are inefficient for removal of residual NAPL (Powers et al. 1991;
National Research Council 1994).
Nevertheless, pump-and-treat methods
continue to be used for aquifer restoration purposes (National Research Council
1994). Factors contributing to inefficiencies of pump-and-treat remediation may
include nonequilibrium interphase partitioning at high water velocities, or
diminishing interfacial contact area (Miller et al. 1990; Powers et al. 1992, 1994;
Imhoff et al. 1993), and dilution and bypassing effects resulting from
heterogeneous permeability and NAPL distribution (Anderson et al. 1992a;
Brusseau 1992). Therefore, the focus of this study is to develop and apply basic
quantitative tools for the evaluation of remediation alternatives in fractured
permeable formations, such as the coastal plain aquifer described above.
H. Rubin et al.
The coastal plain water-supply aquifer of Israel was first discovered to be
contaminated with kerosene in 1983. Site characterization studies (Kanfi 1986)
found a kerosene layer of up to 70 cm floating on top of the groundwater table,
which at that time was at an elevation of about 2 m below sea level. Since the
discovery of the kerosene contamination, the groundwater table has been elevated
by recharge to a height of about 4 m above sea level (6 m elevation increase) to
avoid saltwater intrusion into the coastal aquifer. Coinciding with the increased
elevation of the groundwater table was the nearly complete disappearance of the
formerly floating kerosene layer. Recent soil core information indicates that a
majority of the kerosene is located below the water table, entrapped within the
permeable blocks of the sandstone.
The integrity of water-supply aquifers is often threatened by contamination
from nonaqueous phase liquids (National Research Council 1994). Fractured
permeable water-supply aquifers are not immune to these hazards, as exemplified
by the kerosene contamination of the coastal plain aquifer in Israel, causing the
abandonment of several water-supply wells (Kanfi 1986). Field evidence from the
coastal plain aquifer is consistent with experimental studies of nonaqueous phase
liquid (NAPL) migration in an artificial fractured permeable formation (Schwille
1981, 1988). The experimental results showed that a substantial proportion of
NAPL can. flow through both fractures and permeable blocks, but that fractures
offer minimal NAPL storage following NAPL drainage and redistribution. Other
experimental work in homogeneous consolidated and unconsolidated materials has
shown that, following redistribution, the NAPL is eventually entrapped as
disconnected ganglia (blobs) encompassing one or more porebodies (Chatiz et al.
1983; Powers et al. 1991). Based on inference from these field and laboratory
studies, a conceptual model of NAPL contamination in a fractured permeable
formation is developed and shown in Fig. 1. The entrapped immobile residual
NAPL is assumed to be present in the permeable blocks as disconnected ganglia
filling one or more porebodies and NAPL storage in the fractures is considered
negligible.
Entrapped NAPL residual is a source of long-term chronic
groundwater contamination because it cannot be effectively mobilized or easily
dissolved, due to capillary forces and its generally low aqueous solubility (Mackay
et al. 1985; Mercer and Cohen 1990).
Theoretical and practical case studies have shown that traditional pump-andtreat methods are inefficient for removal of residual NAPL (Powers et al. 1991;
National Research Council 1994).
Nevertheless, pump-and-treat methods
continue to be used for aquifer restoration purposes (National Research Council
1994). Factors contributing to inefficiencies of pump-and-treat remediation may
include nonequilibrium interphase partitioning at high water velocities, or
diminishing interfacial contact area (Miller et al. 1990; Powers et al. 1992, 1994;
Imhoff et al. 1993), and dilution and bypassing effects resulting from
heterogeneous permeability and NAPL distribution (Anderson et al. 1992a;
Brusseau 1992). Therefore, the focus of this study is to develop and apply basic
quantitative tools for the evaluation of remediation alternatives in fractured
permeable formations, such as the coastal plain aquifer described above.