Modeling Ground Water Quality Sampling Decisions

Questions such as what, where, when, and how often to sample play a central role in the development of monitoring strategies. Limited resources will not permit sampling for many contaminants at the same frequency at all well sites. Therefore, a resource allocation strategy is necessary to arrive at answers for the preceding types of questions. Such a strategy for a ground water quality monitoring program is formulated as an integer programming model (an optimization model). The model will be of use in the process of deciding what constituents to sample and where to sample them so as to maximize a given objective, subject to a set of budget, sampling, and regulatory constraints. The maximization objective in the model is defined as a weighted function of population exposure to a scaled measure of observed chemical concentrations. The sampling constraints are based on the observed variability of contaminants in the aquifer, needed precision in estimates, a chosen level of significance, the available budget for implementing the program, and selected regulatory constraints. The model is tested with field data obtained for 10 selected constituents from more than 650 wells in the Cambrian-Ordovician aquifer in Iowa. Results from two alternative formulations of the model are compared, analyzed, and discussed. Further avenues for research are briefly outlined.     

Innovative Sampling Techniques for Ground Water Monitoring at Hazardous Waste Sites

The authors have recently used several innovative sampling techniques for ground water monitoring at hazardous waste sites. Two of these techniques were used for the first time on the Biscayne Aquifer Super-fund Project in Miami, Florida. This is the largest sampling program conducted so far under the U.S. Environmental Protection Agency (EPA) Superfund Program.
One sampling technique involved the use of the new ISCO Model 2600 submersible portable well sampling pump. A compressed air source forces water from the well into the pump casing and then delivers it to the surface (through a pulsating action). This pump was used in wells that could not be sampled with surface lift devices.
Another sampling technique involved the use of a Teflon manifold sampling device. The manifold is inserted into the top of the sampling bottle and a peristaltic pump creates a vacuum to draw the water sample from the well into the bottle. The major advantage of using this sampling technique for ground water monitoring at hazardous waste sites is the direct delivery of the water sample into the collection container. In this manner, the potential for contamination is reduced because, prior to delivery to the sample container, the sample contacts only the Teflon, which is well-known for its inert properties.
Quality assurance results from the Superfund project indicate that these sampling techniques are successful in reducing cross-contamination between monitoring wells. Analysis of field blanks using organic-free water in contact with these sampling devices did not show any concentration at or above the method detection limit for each priority pollutant.     

Sampling Strategies for Estimation of Parameters Related to Ground Water Quality

Multiple theoretical sampling designs are studied to determine whether sampling designs can be identified that will provide for characterization of ground water quality in rural regions of developing nations. Sampling design in this work includes assessing sampling frequency, analytical methods, length of sampling period, and requirements of sampling personnel. The results answer a set of questions regarding whether using innovative sampling designs can allow hydrogeologists to take advantage of a range of characterization technologies, sampling strategies, and available personnel to develop high-value, water-quality data sets. Monte Carlo studies are used to assess different sampling strategies in the estimation of three parameters related to a hypothetical chemical observed in a ground water well: mean concentration (MeanC), maximum concentration (MaxC), and total mass load (TML). Five different scenarios are simulated. These scenarios are then subsampled using multiple simulated sampling instruments, time periods (ranging from 1 to 10 years), and sampling frequencies (ranging from weekly to semiannually to parameter dependent). Results are analyzed via the statistics of the resulting estimates, including mean square error, bias, bias squared, and precision. Results suggest that developing a sampling strategy based on what may be considered lower quality instruments can represent a powerful field research approach for estimating select parameters when applied at high frequency. This result suggests the potential utility of using a combination of lower quality instrument and local populations to obtain high frequency data sets in regions where regular monitoring by technicians is not practical.     

An Environmentally Friendly Decontamination Protocol for Ground Water Sampling Devices

Several detergent-washing/air-drying decontamination protocols were tested to determine their ability to remove residual contamination from two types of ground water sampling devices. We tested a relatively simply constructed device, a bailer, and a much more complex, and theoretically more difficult to decontaminate, bladder pump. The devices were decontaminated after sampling ground water that was contaminated with organics that varied in their hydrophobic nature and propensity to be sorbed by the materials in the devices. These studies showed that a hot-detergent wash, hot-water rinse, and hot-air drying protocol was effective.     

The Challenge of Ground Water Quality Monitoring

"Valuable information pertaining to contaminant sources, contaminants, and ground water quality was derived using the state-supplied data."     

An Analysis of Low-Flow Ground Water Sampling Methodology

Low-flow ground water sampling methodology can minimize well disturbance and aggravated colloid transport into samples obtained from monitoring wells. However, in low hydraulic conductivity formations, low-flow sampling methodology can cause excessive drawdown that can result in screen desaturation and high ground water velocities in the vicinity of the well, causing unwanted colloid and soil transport into ground water samples taken from the well. Ground water velocities may increase several fold above that of the natural setting. To examine the drawdown behavior of a monitoring well, mathematical relationships can be developed that allow prediction of the steady-state drawdown for constant low-flow pumping rates based on well geometry and aquifer properties. The equations also estimate the time necessary to reach drawdown equilibrium. These same equations can be used to estimate the relative contribution of water entering a sampling device from either the well standpipe or the aquifer. Such equations can be useful in planning a low-flow sampling program and may suggest when to collect a water sample. In low hydraulic conductivity formations, the equations suggest that drawdown may not stabilize for well depths, violating the minimal drawdown requirement of the low-flow technique. In such cases, it may be more appropriate to collect a slug or passive sample from the well screen, under the assumption that the water in the well screen is in equilibrium with the surrounding aquifer.     

The Effects of Ground Water Sampling Devices on Water Quality: A Literature Review

This paper reviews both field and laboratory studies that tested or compared the ability of various types of sampling devices to deliver representative ground water samples. Several types of grab samplers, positive displacement devices, and suction-lift devices were evaluated, Gas-lift and inertial-lift pumps were also evaluated. This study found that most of these devices can. under certain circumstances, alter the chemistry of ground water samples, das-lift pumps, older types of submersible centrifugal pumps, and suction-lift devices are not recommended when sampling for sensitive constituents such as volatile organics and inorganics, or inorganics that are subject to oxidation/precipitation reactions. In general, of the devices reviewed in this paper, bladder pumps gave the best recovery of sensitive constituents. However, better performance could be achieved for several devices if improved operational guidelines were developed by additional testing, especially at lower flow rates. Clearly, further research is warranted. Future studies should focus on pumping rate, flow control mechanisms, and dedication or decontamination of sampling devices.     

An Evaluation of Some Systems for Sampling Gas-Charged Ground Water for Volatile Organic Analysis

Loss of volatile organics during sampling is a well-recognized source of bias in ground water monitoring; sampling protocols attempt to minimize such loss. Such bias could be enhanced for ground water highly charged with dissolved gases such as methane. Such ground water was the object of this study. A positive-displacement bladder pump, a momentum-lift pump and a suction-lift, peristaltic pump were employed in sampling both methane-charged ground water for volatile aromatic hydrocarbons and a CO₂-charged reservoir water for volatile chlorinated hydrocarbons. In both cases, the suction-lift pump produced samples with a significant negative bias (9 to 33 percent) relative to the other methods. Little difference between samples produced by the other pump Systems was noted at the field site, but in sampling the reservoir, the bladder pump produced samples that were 13 to 19 percent lower in halocarbon concentration than were samples from the momentum-lift pump.
These negative biases are tentatively interpreted as losses due to volatilization during sampling. Slightly greater negative biases occur for compounds of higher volatility as estimated from their Henry's law constants. Additional studies appear to be warranted in order to adequately establish the scientific basis for recommending protocols for sampling ground water in which degassing could enhance the loss of volatile organics during sampling.     

A Systematic Approach for Evaluating the Quality of Ground Water Monitoring Data

The recognition and assurance of the quality of ground water monitoring data are crucial to the correct assessment of the magnitude and extent of a ground water contamination problem. This article addresses an approach being developed to systematically evaluate the quality of a given set of ground water monitoring data collected during site investigation/ remedial action efforts. The system consists of a checklist of criteria, grouped into four major categories, which can be applied to laboratory or field measurements.
The first category, basis of measurement, considers whether the appropriate sampling, boring and/or analytical methods were chosen to obtain the measurement and the limitations of each method. Secondly, application of the method is assessed. This includes examination of the extent to which procedures were correctly performed, the use of quality control measures and calibration, and possible sources of error in the measurements. Third, evaluation of applied statistical methods is made, with consideration given to which statistics are meaningful in a given context and whether measurements are reproducible. The final category, corroborative information, considers whether independent data or other information are available that add credibility to the values measured.
In this approach, a "high quality" data value is defined as one in which accuracy is supported by meeting the preceding criteria. When accompanied by precision information, high quality data allow for defensible assessments and actions. This evaluation system is useful in developing monitoring programs and in guiding documentation of field and laboratory methods during data collection. It relies heavily on experienced judgment and can be catalyst for the beneficial exchange of knowledge and ideas among ground water professionals.     

A New Multilevel Ground Water Monitoring System Using Multichannel Tubing

A new multilevel ground water monitoring system has been developed that uses custom-extruded flexible 1.6-inch (4.1 cm) outside-diameter (O.D.) multichannel HOPE tubing (referred to as Continuous Multichannel Tubing or CMT) to monitor as many as seven discrete zones within a single borehole in either unconsolidated sediments or bedrock. Prior to inserting the tubing in the borehole, ports are created that allow ground water to enter six outer pie-shaped channels (nominal diameter = 0.5 inch [1.3 cm]) and a central hexagonal center channel (nominal diameter = 0.4 inch [1 cm]) at different depths, facilitating the measurement of depth-discrete piezometric heads and the collection of depth-discrete ground water samples. Sand packs and annular seals between the various monitored zones can be installed using conventional tremie methods. Alternatively, bentonite packers and prepacked sand packs have been developed that are attached to the tubing at the ground surface, facilitating precise positioning of annular seals and sand packs. Inflatable rubber packers for permanent or temporary installations in bedrock aquifers are currently undergoing site trials. Hydraulic heads are measured with conventional water-level meters or electronic pressure transducers to generate vertical profiles of hydraulic head. Ground water samples are collected using peristaltic pumps, small-diameter bailers, inertial lift pumps, or small-diameter canister samplers. For monitoring hydrophobic organic compounds, the CMT tubing is susceptible to both positive and negative biases caused by sorption, desorption, and diffusion. These biases can be minimized by: (1) purging the channels prior to sampling, (2) collecting samples from separate 0.25-inch (0.64 cm) O.D. Teflon sampling tubing inserted to the bottom of each sampling channel, or (3) collecting the samples downhole using sampling devices positioned next to the intake ports. More than 1000 CMT multilevel wells have been installed in North America and Europe to depths up to 260 feet (79 m) below ground surface. These wells have been installed in boreholes created in unconsolidated sediments and bedrock using a wide range of drilling equipment, including sonic, air rotary, diamond-bit coring, hollow-stem auger, and direct push. This paper presents a discussion of three field trials of the system, demonstrating its versatility and illustrating the type of depth-discrete data that can be collected with the system.     

An Inexpensive, Multi-Use, Dedicated Pump for Ground Water Monitoring Wells

A simple, inexpensive sampling pump has lately come into use in ground water monitoring. The pump is referred to as an inertial pump; its only downhole components are a foot valve connected to a length of tubing or pipe. The operating principle of the pump is based on the inertia of a column of water within the riser tubing. Ground water is drawn through the foot valve and up the riser tubing by rapid up and down movements of the tubing. This pumping method is not new, but has only recently been applied to monitoring wells. Foot valves are available in a variety of materials and sizes and can be used in monitoring wells as small as 19mm (3/4 inch) I.D. Flexible polyethylene or Teflon® tubing, and in some cases stainless steel tubing or rigid PVC pipe, is used as the riser. The inertial pump satisfies most of the criteria normally cited for an "ideal" sampling device. The pump is easy to operate, reliable, durable, portable, and virtually maintenance-free. It can be operated manually from as deep as 40m or from as deep as 60m using a motor drive. The pump is inexpensive, and therefore suitable for use as a dedicated sampling pump. Recent tests have shown the pump to be suitable for sampling volatile organics. The inertial pump has a high flow capacity and performs well in silty/sandy environments, which makes it useful for developing and purging monitoring wells. It may also be used to perform field hydraulic conductivity tests.     

A New System for Ground Water Monitoring

This paper describes a new system for ground water monitoring, "the BAT System," which includes the following functions: (a) sampling of ground water in most types of soils, (b) measurement of pore water pressure, and (c) in situ measurement of hydraulic conductivity. The system can also be used for tracer tests. The system utilizes a permanently installed filter tip attached to a steel or PVC pipe. Installation is normally performed by pushing the tip down to the desired depth. The filter tip can also be buried beneath a landfill. The primary feature of the new system is that the filter tip contains a self-sealing quick coupling unit, which makes it possible to temporarily connect the filter tip to adapters for various functions, e.g. water sampling and for measurement of pore pressure and hydraulic conductivity. The new technique makes sampling of both pressurized water and gas possible. Samples are obtained directly in hermetically sealed, pre-sterilized sample cylinders. Sampling of ground water and measurement of pore pressure can be repeated over a long period of time with undiminished accuracy. This technique is also well-adapted for taking water samples from different strata in a soil profile, in both the saturated and unsaturated zones. Actual installations range from 0.5 to 60m depth.     

Sampling of VOCs with the BAT Ground Water Sampling System

In the BAT ground water sampling system, a stainless steel probe with a porous filter element is pushed vertically to the desired sampling depth. An evacuated glass sampling tube is then lowered down the penetration rods where it makes contact with the filter via a hypodermic needle and draws a pore fluid sample.
An investigation of the system was carried out at a number of sites contaminated by leaking underground gasoline storage tanks. Ground water samples obtained using the BAT system and adjacent monitoring wells were analyzed for volatile organic compounds (VOCs).
Because the BAT system is an in situ penetration device with a small filter length, it is possible to determine variations in contaminant concentration with depth. BAT samples in general exhibited higher recovery of VOCs than did bailer samples from adjacent monitoring wells screened over large intervals.
Much higher levels of VOCs were recovered when the probe was used with its 316 stainless steel filter than when using the high-density polyethylene (HDPE) filter. Significant sorption apparently occurred on the latter filter.
Because the BAT sample tubes are sealed and remain a closed system, the in situ water pressure is maintained. No significant loss of VOCs was found in sampling tubes containing headspace. Samples from the upper tube in the cascaded setup with headspace recovered levels of VOCs as high, or in a few cases higher, than the lower, no-headspace tubes.     

Global Optimal Design of Ground Water Monitoring Network Using Embedded Kriging

We present a methodology for global optimal design of ground water quality monitoring networks using a linear mixed-integer formulation. The proposed methodology incorporates ordinary kriging (OK) within the decision model formulation for spatial estimation of contaminant concentration values. Different monitoring network design models incorporating concentration estimation error, variance estimation error, mass estimation error, error in locating plume centroid, and spatial coverage of the designed network are developed. A big-M technique is used for reformulating the monitoring network design model to a linear decision model while incorporating different objectives and OK equations. Global optimality of the solutions obtained for the monitoring network design can be ensured due to the linear mixed-integer programming formulations proposed. Performances of the proposed models are evaluated for both field and hypothetical illustrative systems. Evaluation results indicate that the proposed methodology performs satisfactorily. These performance evaluation results demonstrate the potential applicability of the proposed methodology for optimal ground water contaminant monitoring network design.     

Inorganic Water Quality Monitoring Using Specific Conductance in Mexico

Five inorganic geochemical data sets with ion balances less than or equal to 5% (for different hydrogeologic basins in Mexico) were used to construct specific conductance (SC) vs. total dissolved solids (TDS) curves for each aquifer. These curves allow one to proxy the specific conductance for inorganic water quality. The data from these five curves were integrated to form a "global" curve for Mexico. The equation that defines the line is SC = 1.40 TDS + 18. We propose that, in addition to allowing ground water monitoring, establishing a SC vs. TDS curve may be implemented to monitor drinking water supplies such as wells and reservoirs. If anomalously high values are observed, this would warrant detailed geochemical sampling.     

Multiport Sock Samplers: A Low-Cost Technology for Effective Multilevel Ground Water Sampling

The importance of obtaining depth-specific ground water samples is now well recognized among practitioners and scientists alike. Many methods and technologies are available for level discrete or depth-specific ground water sampling in consolidated aquifers. All methods have their associated advantages and drawbacks, however. One common disadvantage is that they are expensive. A large number of point discrete ground water samples were required for a UK research project aimed at quantifying natural attenuation processes in ground water contaminated by a former coal carbonization plant. Based on experience from a previous project to develop novel level accurate sampling methodologies for use in existing boreholes, the Ground Water Protection and Restoration Research Unit (GWPRRU) produced and tested a low-cost design multiport sock sampler for ground water monitoring. The sock sampler design allowed the recovery of multiple depth-specific ground water samples from depths of 150 feel (45 m) from individual boreholes in the sandstone aquifer at the field site. Because of their use of inexpensive materials, simple design, installation and use that does not require gravel packs, packers, or grouting, sock samplers were found to be the most cost effective, convenient, and reliable method of obtaining multiple depth-specific ground water samples at the project field site.