Report Cover   Table of Contents   Sec. 1   Sec. 2   Sec. 3   Sec. 4   Sec. 5   Sec. 6   Sec. 7   Sec. 8

Download the Report PDF (874 KB) ("Right-click", then "Save as..." to your desktop)
View Report Appendices

SECTION 2

DESCRIPTION OF TECHNOLOGIES

No-purge samplers rely on the natural flow of groundwater through a well screen, and therefore the results obtained using these devices will not always be comparable to results obtained using conventional sampling methods which induce groundwater flow into a well by creating a hydraulic gradient through well purging. In the absence of vertical flow, the no-purge devices will primarily monitor groundwater migrating through the well screen at the discrete depth intervals at which the samplers are placed. If vertical flow exists in the well, no-purge sampler results likely will be representative of the aquifer zone with the highest hydraulic head. Groundwater flows from high- to low-head zones, and the zone with the highest hydraulic head will be the source for groundwater flowing vertically through the well, and will therefore be the zone monitored by the no-purge sampler.

As described in Section 1, a total of four diffusion (PDBS, RPPS, PsMS, and RCS) and two grab (HydraSleeve® and Snap Sampler™) sampling devices were selected for this demonstration. Additionally, these methods were compared to two conventional sampling methods (low-flow purge/sample and three-volume purge/sample). Specific design and method details for each of these sampling techniques are presented in Table 2.1 and the following subsections. Note that the sampler dimensions and volumes listed in Table 2.1 correspond to the versions used in this McClellan AFB field demonstration; other versions of these samplers may be available.

2.1 DIFFUSION SAMPLERS

For diffusion samplers, chemical constituents in the groundwater diffuse across the membrane over time, and the chemical content of the water inside the sampler reaches equilibrium with the chemical content of groundwater in that interval of the well. The sampler is subsequently removed from the well, and the water in the diffusion sampler is transferred to a sample container and submitted for laboratory analysis. Once a diffusion sampler is placed in a well, it remains in place until chemical equilibrium is achieved between the water in the well casing and the water in the diffusion sampler. There is a time-lag between the time groundwater enters a well and the diffusion of the chemicals in the groundwater into a diffusion sampler. This time-lag is variable depending on several factors such as the groundwater temperature, the physicochemical properties of the compound of interest, and the diffusive membrane used in the sampler. Because of this quality, diffusion samplers are representative of a time-weighted average of chemical concentrations in groundwater.

TABLE 2.1
SUMMARY OF NO-PURGE SAMPLING DEVICES TESTED NO-PURGE SAMPLER DEMONSTRATION MCCLELLAN AFB, CALIFORNIA

2.1.1 Passive Diffusion Bag Sampler (PDBS)

The PDBS used in this demonstration is constructed of a 45-centimeter (cm)-long section of 5.08-cm-diameter, 4-mil-thick, low-density polyethylene (LDPE) tubing that is permanently sealed on one end and sealed on the other end with a high-density polyethylene (HDPE) cap (Figure 2.1). The pore size of the LDPE is approximately 0.001 micron, which does not permit the flux of water molecules (i.e., it does not leak). The sampler, which holds approximately 350 milliliters (mL) of purified water, is placed in “flex-guard” polyethylene mesh tubing for abrasion protection, attached to a weighted rope, and lowered to a predetermined depth within the screened interval of a well. The rope is weighted to ensure that the sampling devices are positioned at the correct depth and that they do not float upward through the water column.

FIGURE 2.1
STANDARD PDBS
NO-PURGE SAMPLER DEMONSTRATION
MCCLELLAN AFB, CALIFORNIA

Depending on the hydrogeologic characteristics of the aquifer, the diffusion samplers can reach equilibrium within 3 to 4 days (Vroblesky, 2001). Groundwater samples collected using the diffusion samplers are thought to be representative of water present within the well during the previous 24 to 72 hours. However, the recommended minimum equilibration time for water temperatures above 10 degrees Celsius (°C) is two weeks (ITRC, 2004).

PDB samples are not susceptible to matrix interferences caused by turbidity because the membrane used in the device is not permeable to colloids or other particles larger in diameter than approximately 0.001 micron. PDB samples also are not subject to volatilization loss by degassing during effervescence when the samples are acidified for preservation in highly alkaline waters because the alkalinity from the aquifer does not penetrate the membrane.

2.1.2 RIGID POROUS POLYETHYLENE SAMPLER (RPPS)

RPPSs have recently been tested in a laboratory setting by the US Geological Survey (USGS). The tested samplers consisted of a 1.5-inch outside diameter (OD), 6.2-inchlong, rigid polyethylene tube having a pore size of 6 to 15 microns (Figure 2.2). Given the relatively large pore size, the RPPS could potentially be used to sample for a relatively wide variety of volatile and non-volatile analytes. The bench-scale test results indicated that this type of sampler can yield accurate results for VOCs (including MTBE), chromium, and chloride (Vroblesky, 2004). Potential disadvantages of this sampler include the following:

• The porous polyethylene sampler pores tend to retain air even when submerged. Because the entrapped air reduces sampler permeability, the air should be removed prior to use by flushing the samplers with water.
  
  

FIGURE 2.2
STANDARD RPPS
NO-PURGE SAMPLER DEMONSTRATION
MCCLELLAN AFB, CALIFORNIA

• Tests performed to date indicate that the maximum feasible sampler dimensions are approximately 1.5 inches OD by 7.5 inches long (volume equal to approximately 175 mL). Use of a longer sampler would result in leakage of water out of the sampler walls due to the higher head pressure present in the sampler (Vroblesky, 2004).
  
  

2.1.3 POLYSULFONE MEMBRANE SAMPLER (PsMS)

Testing of ‘Peeper’ samplers performed by (among others) Dr. Andrew Jackson of Texas Tech University has indicated that dissolved concentrations of non-volatile groundwater constituents can pass through a polysulfone (e.g., HT® Tuffryn) membrane having a sufficient pore size (Jackson, 2003). Peeper samplers are rigid structures that can hold volumes of water separated from the environment by porous membranes to monitor dissolved constituents in saturated environments. The same polysulfone material used in some Peeper samplers also can be used to construct PSDs. The samplers constructed for use in the McClellan study were comprised of a rigid 2-inch-long section of 2-inch-OD PVC pipe that was covered on both ends with the flexible polysulfone membrane. The polysulfone membrane was held in place by sliding a PVC coupling over the end of the pipe (Figure 2.3). The coupling was held in place by friction. The samplers were filled with purified water prior to deployment. The pore size of the polysulfone material that was used is 0.2 micron. The volume of each sampler canister was approximately 108 mL, and two of these canisters were deployed at each sample depth. One conclusion from a previous diffusion sampler demonstration at Grissom Air Reserve Base (Parsons, 2004b) was that the orientation of the porous membrane relative to the assumed direction of groundwater flow was potentially an important consideration. Because of this, samplers were deployed in an orientation such that the plane of the membrane was positioned orthogonally to horizontal groundwater flow. Due to the lack of field- or bench-scale testing of PsMSs, potential advantages or disadvantages of this sampler have not been quantified.

FIGURE 2.3
STANDARD PSMS
NO-PURGE SAMPLER DEMONSTRATION
MCCLELLAN AFB, CALIFORNIA

2.1.4 REGENERATED CELLULOSE SAMPLER (RCS)

Regenerated cellulose samplers have been successfully tested in wells for inorganic and volatile organic constituents in groundwater (Vroblesky et al., 2002; Ehlke et al., 2004). The sampler used in this investigation consisted of a perforated PVC pipe inside a sleeve of high-grade regenerated cellulose tubular dialysis membrane (Membrane Filtration Products, Inc., Seguin, Texas) with an outer protective LDPE mesh (Figure 2.4). The membranes have a nominal molecular-weight cutoff of 8,000 daltons, or about 0.0018 micron pore size, and a flat width of about 3 inches. The diameter of the filled sampler is about 1 inch and the length is about 13 inches, with a capacity of approximately 400 mL. A potential disadvantage of this sampler is that it may begin to biodegrade in some groundwater systems (Vroblesky and Pravecek, 2002); however, the ability of the samplers to produce chemical concentrations comparable to other methods in previous investigations indicates that, during short-term deployment, the susceptibility of the cellulose membrane to biodegradation does not significantly affect the sampler’s usefulness in at least some groundwater environments.

Ehlke et al. (2004) found that VOC concentrations in RCSs equilibrated within 3 days and iron and bromide concentrations equilibrated within 3 to 7 days. In an unpublished study, Vroblesky (personal communication) found that VOC and chloride concentrations had reached equilibrium by the first sampling event at 8 days. Vroblesky et al. (2002) state that concentrations of inorganic constituents in RCSs equilibrated within 20.5 to 92 hours.

FIGURE 2.4
STANDARD RCS
NO-PURGE SAMPLER DEMONSTRATION
MCCLELLAN AFB, CALIFORNIA

2.2 GRAB SAMPLERS

In contrast to the diffusion samplers, grab sampling devices represent more of an equilibrated instantaneous “snap-shot” in time of groundwater conditions. For these devices, the sampler is deployed in a well and is left there until groundwater conditions have re-equilibrated. At that time the groundwater is captured by the device, and the resulting sample is submitted to the laboratory for analysis.

2.2.1 SNAP SAMPLER™

The Snap Sampler™ (patent pending) was developed by ProHydro, Inc. and was initially designed to collect a representative VOC sample in situ without the need for purging. Samples collected with the Snap Sampler™ can be analyzed for more than VOCs. Utilizing minimum sample volume requirements, this sampler can also be used for analyzing a larger number of physical and/or chemical water quality parameters. The Snap Sampler™ employs standard-sized 40 mL glass volatile organics analysis (VOA) vials with double end-openings (Figure 2.5). Specialty Teflon® end closure caps seal water within the Snap Sampler™ vial with an internal closure spring. The closure spring is made of perfluoroalkoxy (PFA Teflon®)-coated stainless steel. To deploy the sampling device, the VOA vial is placed inside the Snap Sampler™, and the end closure caps are attached to the sampler’s trigger mechanism in an open position. Both ends of the VOA vial are open to the well environment during the deployment period.

FIGURE 2.5

STANDARD SNAP SAMPLER™
NO-PURGE SAMPLER DEMONSTRATION
MCCLELLAN AFB, CALIFORNIA

Up to three Snap Samplers™ can be connected in series with a single suspension/trigger cable. The suspension/trigger cable consists of a 1/32-inch-diameter stainless steel wire rope within ¼-inch HDPE tubing. The HDPE tubing attaches to the samplers and the wire rope attaches to the release mechanism of the sampler. The samplers are lowered into the well to a predetermined depth using the suspension/trigger cable. The suspension/trigger cable is secured at the surface at a well-head docking station that does not interfere with well-head locks or water level measuring devices. The Snap Sampler™ is left for an appropriate length of time to allow the well to return to equilibrium with the surrounding groundwater. When ready to collect samples, the internal trigger cable is manually pulled at the wellhead to activate the sampler release mechanism. The trigger releases the vial caps, which close onto the VOA vial by action of the internal closure spring. The vial caps and spring seal the groundwater within the sampling container.

The samplers are then retrieved from the well, VOA bottles are removed from the Snap Sampler™, preservative is added (if necessary) using a method that does not require the sample bottle to be uncapped (Parsons, 2004a [SOP can be accessed via vendor website at www.snapsampler.com]), and end caps are secured with standard VOA vial screw caps. The VOA vials can be used with standard laboratory autosampling equipment designed for 40 mL vials. From the well to the autosampler, water samples are never exposed to ambient air. A 125-ml sample bottle is currently in development to accommodate larger volume needs. Other sampler and bottle material compositions are available or are being developed to accommodate different sampling needs. For example, a fully non-metallic sampler is now available for metals sampling.

The diameter of the sampler apparatus used at McClellan was 1.6 inches. The length of the device was approximately 10 inches with a single sampler and vial, 17 inches with two samplers and two vials, and 23 inches with three samplers and three vials. The longest distance between the end openings of the three-vial configuration was 17 inches. The current configuration uses a new connector that changes these dimensions slightly as follow: diameter = 1.66 inches, length = 8 inches with a single sampler and vial, 16 inches with two samplers and two vials, and 24 inches with three samplers and three vials. The longest distance between the end openings of the current three-vial configuration is 19 inches.

2.2.2 HYDRASLEEVE® SAMPLER

The HydraSleeve® sampler (US patents #6,481,300 and #6,837,120), manufactured by GeoInsight (www.hydrasleeve.com), is designed to collect a representative sample for most physical and chemical parameters without purging the well. It collects a water sample from a defined interval within the well screen without mixing fluid from other intervals. Physically, it is a section of lay-flat polyethylene tubing, sealed at the bottom end, and built with a polyethylene reed-valve at the top end (Figure 2.6).

FIGURE 2.6
STANDARD HYDRASLEEVE®
NO-PURGE SAMPLER DEMONSTRATION
MCCLELLAN AFB, CALIFORNIA

The empty sampler is weighted at the bottom, attached to a line, and then lowered to a predetermined depth within the well screen. It is typically left in the well for a period of time to allow the well to re-equilibrate following sampler deployment. Once the well has re-equilibrated, the sampler can be activated for sample collection. Prior to activation, the sampler remains in a collapsed (i.e., empty) state and therefore takes up minimal space within the well. To activate, the sampler is pulled up a distance equal to 1 to 2 times the sampler length (2.5 to 5 feet for a 30-inch-long sampler). As the sampler rises through the water column, the reed valve opens, allowing the sampler to “core” the water column through which it is being raised. Once full, the reed valve closes, which prohibits any more water from entering the sampler. An alternate approach to activating the sampler is to raise and lower it multiple times over a distance equal to the sampler length. However, this approach is less attractive because the raising and lowering of the sampler can result in increased agitation of the water in the well and higher turbidity levels in the sample.

The 24- to 30-inch-long sampler can be purchased in either 1.5- or 2.5-inch diameter models; the 30-inch sampler has volumes of 1,000 mL and 2,500 mL for these diameters, respectively.

2.3 CONVENTIONAL SAMPLING METHODS

One of the scoping guidelines described in Section 1.5 was to have results from at least one other traditional sampling method that could serve as a “baseline” for comparison purposes to the diffusion and grab sampling technologies. In order to address this scoping guideline, conventional sampling methods used as “baseline” measurements were:

1. Sampling following low-flow/minimal drawdown purging , and
2. Sampling following conventional purging of at least three well-casing volumes of water and stabilization of water quality parameters.

The objective of low-flow sampling is to remove a small volume of water at a low flow rate from a small portion of the screened interval of a well without mixing water among vertical zones. Ideally, by placing the inflow port of a pump at a prescribed depth within the screened interval of a well, and by withdrawing water at a slow rate, groundwater will be drawn from the aquifer into the well only in the immediate vicinity of the pump. This theoretically depth-discrete sampling allows for vertical definition of contamination in the aquifer. In practice, however, when a low-flow sample is collected, determining the portion of the screened interval of the aquifer that contributed water to the sample can be problematic.

Groundwater sampling using the three-volume purge method involves removing a large volume of water (three to five well-casing volumes) from the well over a short time. The objective of this method is to remove all stagnant water present within the well casing, as well as groundwater present in the surrounding well filter pack. Theoretically, by removing this water quickly, the “stagnant” water that resided in the well and filter pack will be replaced with “fresh” groundwater from the surrounding formation with minimal mixing. The “fresh” groundwater that is then sampled is considered to be representative of the local groundwater. Rapid drawdown of the water level in a well is not uncommon, and wells are often purged dry using this method.

Conventional sampling at McClellan that is part of regularly scheduled LTM is performed using both low-flow and three-volume purge techniques. Low-flow sampling is only performed at wells in which dedicated bladder pumps have been installed, while three-volume sampling is performed using submersible pumps that are moved from well to well. McClellan is in the process of installing dedicated bladder pumps in all of their regularly sampled wells so that all future conventional sampling will be performed using the low-flow technique.

In order to maximize consistency and comparability between the historical conventional sampling record for McClellan and the conventional sampling performed as part of this demonstration, similar procedures were followed to the extent possible. However, as described in the Work Plan (Parsons, 2004a) the presence of dedicated pumps in a well automatically excluded that well for use in this demonstration. Therefore, the low-flow sampling that was performed during this demonstration did not strictly adhere to the SOP for low-flow sampling provided in the McClellan QAPP (URS, 2003).

A submersible pump (i.e., Grundfos RediFlo2®) and new, clean dedicated LDPE tubing were used to perform all purging and sampling of the wells. The pump intake was positioned at the midpoint of the saturated portion of the well screen, and the flow rate was controlled to minimize drawdown in the well (during low-flow purging only). Average pump rates varied from approximately 0.09 to 0.19 gallon per minute (gpm) for the low-flow purge and from approximately 0.71 to 4.0 gpm for the three-volume purge. Drawdown was monitored throughout the low-flow purge using a water-level probe. Field parameters including temperature, pH, conductivity, dissolved oxygen (DO), oxidation-reduction potential (ORP), and turbidity also were monitored in a flow-through cell during both low-flow and three-volume purging. Once well stabilization was achieved, as demonstrated by stabilized field parameters (described in the Work Plan [Parsons, 2004a]), samples were collected. For the low-flow technique, sample bottles were filled directly from the pump discharge. For the three-volume purge, samples were collected using a bailer following completion of the purge, as specified in the McClellan QAPP (URS, 2003).

For all wells, the low-flow sample was collected first, after which time the pump rate was increased and the three-volume purge sample was collected following evacuation of the required purge volume and field parameter stabilization.

Report Cover   Table of Contents   Sec. 1   Sec. 2   Sec. 3   Sec. 4   Sec. 5   Sec. 6   Sec. 7   Sec. 8