Passive DSs have been used to examine ground-water VOCs primarily in two types of applications. The first type of application involves obtaining VOC concentrations in ground-water immediately prior to its discharge to surface-water. This application is of interest because dilution of VOCs in surface-water and lateral transport by currents make contaminant-discharge zones difficult or impossible to locate by examining surface-water samples. The second type of application involves deploying DSs in wells for well monitoring. In wells having saturated screen intervals longer than 1.5 m, multiple DSs typically are deployed to investigate the entire screen length. This application is of interest in appropriate wells because it has the potential to greatly reduce sampling costs by eliminating purge requirements and the need for expensive well pumps, and by decreasing the manpower and time normally required for sampling. A series of investigations in 2002-2003 [19] examined the potential for application of PDB samplers for use in monitoring wells at 14 Air Force bases. Data from these studies showed an average cost savings of 75% when switching from conventional sampling to PDB samplers. An additional advantage of DSs is the potential to delineate VOC stratification in the screened or open interval of the well.

13.2.1 VOCs in ground-water at the ground-water/surface-water interface

VOCs in ground-water at the ground-water/surface-water interface have been used for several years as a tool for examining contaminant-discharge zones. Because VOCs are volatile, there is a tendency for VOCs moving through pore water to partition into any vapor phase, such as a gas bubble, that they contact. VOCs in ground-water discharging to surface-water through organic-rich bed sediments can come into contact with methane bubbles forming as the bed sediments decay by methanogenesis. Thus, some VOCs partition into the methane bubbles and are released to the overlying water and atmosphere during ebullition (bubble emission). One study found that the VOC content of methane gas bubbles in bottom sediment could be used to identify discharge zones of VOC-contaminated ground-water [28]. In a like manner, PVD samplers provide a vapor phase for VOC partitioning at the ground-water/surface-water interface. This type of sampler has been used widely to identify VOC-contamination zones in ground-water beneath surface-water (e.g. [3,16-18,29,30]). In one study, data from the samplers were used to identify specific fracture zones contributing contamination to a stream [16]. PVD samplers provide a vapor concentration in equilibrium with the environmental aqueous concentration of VOCs. In some cases, water-filled PDB samplers have been deployed in bed sediments to obtain direct aqueous concentrations of VOCs in pore-water [29,31].

The PVD sampler, an empty, uncapped vial (Fig. 13.2A) enclosed in two LDPE bags (Fig. 13.2B), is buried approximately 18-45 cm in the bottom sediment of gaining reaches in surface-water bodies and allowed to equilibrate. The sampler is attached to a wire surveyor flag for sampler identification and recovery. Sampler deployment often can be done by wading and using a shovel. In some circumstances, however, deployment and recovery may involve divers or other means of deep-water emplacement. The samplers are allowed to remain in place for at least 2 weeks for the samplers to equilibrate and for the pore-water to recover from the environmental disturbances caused by sampler deployment.

During sampler recovery, the outer LDPE is removed from the vial opening to prevent entrained sediment from interfering with the seal. The inner LDPE bag is left intact, and the vial is sealed by capping over the inner bag. A septated cap (Fig. 13.2C) is used to allow the trapped vapor to be sampled by a syringe.

An example of a study using PVD samplers in deep water is at Johns Pond, Western Cape Cod, MA, USA [29]. PVD samplers were deployed in Johns Pond to confirm that VOCs in a ground-water plume emanating from the Massachusetts Military Reservation were discharging into the pond. An array of 134 PVD samplers was buried by divers about 15 cm below the pond bottom in the presumed discharge area and allowed to equilibrate for about 2 weeks. At selected sites, water-filled PDB samplers also were buried.

Two areas of high VOC concentrations were identified by the PVD samplers (Fig. 13.3). One area was a broad discharge zone (about 335 m wide) approximately 30-106.5 m offshore, having trichloroethene (TCE) and tetrachloroethene (PCE) with vapor concentrations as high as 890 and 667 parts per billion (ppb) by volume, respectively [32]. This zone represented the discharge area of a plume moving toward the pond from the northwest, identified as the Storm Drain-5 plume. Samples from the second area were located closer to shore than the larger

Fig. 13.3. Discharge areas of TCE plumes, as identified by passive vapor diffusion (PVD) samplers, August and December 1998, Johns Pond, Massachusetts (modified from Ref. [29]).

contamination zone and contained unexpectedly high concentrations of TCE (>40,000 ppb by volume in a PVD sampler and 1200 mgL-1 in an adjacent water-filled PDB sampler). Confirmatory ground-water samples collected with a drive-point sampler near the second area had aqueous TCE concentrations as high as 1100 mgL-1.

Following the initial investigation, a more closely spaced array of110 PVD samplers was installed to map the second area of elevated TCE concentrations (Fig. 13.3). The discharge area detected with the samplers was about 22.9 m wide and extended from about 7.6 to 60 m offshore. Subsequent drilling into the pond bottom and onshore confirmed that the second area was the discharge zone of a distinct TCE plume not associated with Storm Drain-5 plume discharging in the first area. The second plume discharged closer to shore because it was shallower than the Storm Drain-5 plume. Further investigation showed that the second plume was related to a ground-water contamination plume west of Ashumet Pond that had flowed under Ashumet Pond and was discharging into Johns Pond [33] (Fig. 13.3). Thus, in this investigation, PVD samplers confirmed the suspicion that the Storm Drain-5

plume was discharging to Johns Pond, and were used to map the discharge zone and to detect and map a previously unsuspected discharge zone of a ground-water plume from the opposite side of an adjacent pond.

13.2.2 VOCs in ground-water in monitoring wells

Investigations have found that in permeable formations, under natural gradient conditions, there can be sufficient flow through the screened interval of a monitoring well to maintain water isolated from the overlying stagnant water stored in the unscreened part of the well (e.g. [34]). In permeable formations, the general consensus is that flow through the screen is expected to occur, with most of the flow coming from the most permeable strata in the screened interval. Where such through-flow is occurring and there is no mixing with the overlying stagnant water in short-screened wells (3.05 m or less), DSs placed in the screened interval have the potential to provide VOC concentrations comparable with those obtained by traditional pumping methods. Several field investigations have shown good agreement between pumped-sample and DS VOC concentrations in wells (see Ref. [22] for examples).

A number of factors should be considered when using DS in wells. These factors include the potential for contaminant stratification and the potential for in-well mixing.

Contaminant stratification is commonly seen in aquifers when multilevel short-screened wells are utilized [35]. In the absence of in-well mixing, this stratification can manifest itself within the screened interval of the well, as shown by investigations of standing water in monitoring wells using dialysis cells isolated between baffles to limit in-well mixing and vertical flow [24,36].

In the absence of the flow-limiting baffles in the well, in-well mixing can obscure or eliminate the development of chemical stratification within the screened or open intervals of wells. A variety of factors can facilitate the mixing in boreholes. These factors include thermally driven convection cells [37], diffusive mixing [38], and vertical in-well flow [39]. A field test conducted by Church and Granato [40] found that well screens can act as conduits for vertical flow because they can connect zones of differing head and transmissivity, even in relatively homogeneous aquifers. They found that vertical flow can mask the presence of discrete contaminated horizons in the screened interval and can contaminate zones of the aquifer that would not otherwise become contaminated. In these situations, little vertical variation in VOC

concentrations is observed across the length of the screened interval or the zone undergoing vertical flow.

Even in wells without flow-limiting baffles, however, solute stratification can sometimes be seen by using multiple DSs. Harter and Talozi [27], using regenerated cellulose membranes, found large contrasts in salinity and nitrate concentrations over approximately 3 m intervals in 40 wells. They found non-uniform nitrate profiles in 80% of the wells they examined. Stratification of inorganic solutes in the screened intervals of wells has been observed in a variety of open-screened intervals using multiple regenerated cellulose passive samplers [23,41].

Stratification of VOCs has also been observed in the screened intervals of monitoring wells by using multiple PDB samplers at a variety of sites [41-44] (Fig. 13.4). The source of this stratification may include

Explanation -Q- Passive diffusion bag (PDB) sampler Pumped sample

Fig. 13.4. Comparison of diffusion and pumped samples in ground-water showing vertical stratification of TCE in the screened interval (modified from Refs. [43,44]).

such factors as vertical differences in contaminant concentrations outside the well screen, vertical flow through a portion of the screen, density contrasts, or, in wells screened at the water table, volatilization loss at the air/water interface.

In situations where there is vertical stratification of contaminants in the well bore, the use of a borehole flow-meter can sometimes aid in understanding the distribution of contamination and the relation between the pumped-sample concentrations and the DS concentrations, particularly when the pumped-sample and DS VOC concentrations disagree. For example, borehole flow-meter testing in well IRP-31 at Andersen Air Force Base, Guam, showed little or no vertical movement of water within the limits of the flow-meter under ambient conditions (Fig. 13.5A) [41]. Under pumped conditions, most of the water entered the well near the top of the screen (Fig. 13.5B). PDB samplers in the open screen (without flow-limiting baffles) showed substantial stratification of TCE (Fig. 13.5C). The TCE concentration was about 211-218 mgL-1 at a depth of 136.2 m and only 20 mgL-1 at a depth of 141.2 m. The upward increasing concentration in well IRP-31 implies that there may be higher concentrations at shallower depths than the uppermost zone sampled by the PDB samplers (Fig. 13.5C). Because during pumping, most of the water in this well is derived from a horizon shallower than the PDB samplers (Fig. 13.5B), it is probable that the PDB samples represent local concentrations, and the pumped sample primarily represents water derived from a more contaminated zone at a depth of about 135-136 m, slightly shallower than the PDB samplers (Fig. 13.5C). Therefore, it is not surprising that the uppermost PDB sample TCE concentration is slightly lower than the adjacent pumped sample.

The TCE concentration in well IRP-31 at the depth from which pumped samples typically were collected (about 139.5 m) was higher in the pumped samples (150-153 mgL-1) than in the PDB samples (57-63 mgL-1). Examination of the flow-meter data and the vertical distribution of contamination, however, show that the difference does not mean that the PDB concentrations are inaccurate. The vertical concentration and flow-meter profiles in the well show that most of the water during pumping is derived from a shallower zone having higher TCE concentrations than the typical pumped-sample collection depth (Fig. 13.5C). Therefore, the most probable explanation for the difference is that the pumped sample represents a higher concentration in the water from a shallower horizon transported downward in response to pumping.

0 0

Post a comment