to   264.98(d) and 264.99(f), or must be an alternate sampling procedure specified under

 264.97(g)(2). This minimum monitoring frequency is inadequate for karst terranes

dominated by conduit flow, because the storage time of the water is low and the

concentrations of contaminants in conduit systems can vary over a short period of time.

Therefore, hourly sampling of aquifers dominated by conduit flow is recommended before,

during, and after storm or other runoff (e.g., snowmelt) events, although this is a site specific

determination. As described by Quinlan (1990), in aquifers dominated by conduit flow,

"sampling should start at base flow, before the beginning of a storm or meltwater event, and

continue until 4 to 30 times the time to the hydrograph peak has elapsed, depending upon the

extent to which an aquifer is characterized by conduit flow as opposed to diffuse flow.

Sampling may have to be done as often as at 1 to 6 hour intervals in the early part of a

precipitation event and at 4 to 24 hour intervals in the waning part of its hydrograph."

Data from the samples collected during the peak runoff should be compared with

samples collected during base flow under fair weather conditions at other times of the year.

This should enable a reliable assessment of the ground water quality to be made. Quinlan

and Alexander (1987) discuss the rationale for sampling frequencies of ground water in karst

aquifers dominated by conduit flow in much greater detail. The work performed by Quinlan

and Alexander (1987) is site specific, and while it illustrates the considerable extent to which

spring discharge should be evaluated, the results obtained by Quinlan and Alexander should

not be taken as representative of all sites located above karst aquifers dominated by conduit

flow.

The previous discussion highlights the various difficulties associated with ground 

water monitoring in karst aquifers. Site characterization of these hydrogeologic settings is

complex, time consuming, and potentially costly. The Agency again emphasizes its belief

that most facilities sited in karst settings will not be able to meet the ground water monitoring

requirements of Subpart F; therefore, alternative locations for land disposal of hazardous

wastes are preferred.

5.2.6

Fracture Trace Analysis

The detection of ground water contamination in fracture controlled aquifers can be

problematic due to the difficulties in locating the fracture systems that often dominate the

ground water flow system. Fracture traces have been mapped for the purpose of locating

zones of increased weathering, porosity, and permeability that act as preferential pathways of

contaminant migration. Strong correlations between well yields and fracture traces in

carbonates, igneous rock, metamorphic rock, fractured siltstones, and fractured sandstones

have been documented by numerous authors (Jansen and Taylor, 1988). Fracture trace

analysis should be performed at sites where hydrogeologic data indicate that contaminant

migration may occur along fractures.

November 1992

5 20