- Program Areas
- Installation Energy and Water
- Environmental Restoration
- Munitions Response
- Resource Conservation and Resiliency
- Weapons Systems and Platforms
Improved Understanding of Sources of Variability in Groundwater Sampling for Long-Term Monitoring Programs
The objective of this project was to identify key sources of variability that influence volatile organic chemical (VOC) concentration measurements in water samples collected from groundwater monitoring wells using current sampling and analysis techniques. By understanding the sources of variability in groundwater monitoring results, improved sampling and analysis methods can be developed to reduce and control for sampling variability.
The project was implemented through the completion of three project tasks:
Evaluation of Existing Datasets (Task 1): The initial project task utilized large existing groundwater monitoring datasets to identify key sources of variability in groundwater monitoring results. The data mining study involved identification of large databases of groundwater monitoring results, selection of datasets from these databases that were suitable for statistical analysis, and exploratory analysis and statistical analysis of the datasets to identify factors associated with monitoring variability. For this task, three large groundwater monitoring databases were identified: (i) the Hill Air Force Base (AFB) database, (ii) the Marine Corps Logistics Base (MCLB) Albany database, and (iii) a database of monitoring results from 48 underground storage tank sites compiled by GSI. Because the Hill AFB database included the most comprehensive documentation, many of the analyses could only be conducted using the Hill AFB database.
Evaluation of Short-Term Variability (Task 2): For this task, sources of short-term variability in groundwater monitoring results were characterized through a field program that involved collecting and analyzing a large number of groundwater samples from a set of monitoring wells over a short period of time. Eighteen wells at Hill AFB were identified for sampling, including six low variability wells, six medium variability wells, and six high variability wells as determined by the analysis of the Hill AFB groundwater database conducted in Task 1. An intensive sampling program was conducted at each of the wells consisting of the following elements: (i) collection of a series of samples from the middle of the well screen by low-flow sampling over a defined range of purge volumes, (ii) measurement of ambient vertical flow within the screened interval of the well, (iii) collection of samples from the top, middle, and bottom of the well screen by passive sampling, (iv) collection of samples from the top, middle, and bottom of the well screen by low-flow sampling, and (v) collection of a second series of samples from the middle of the well screen by low-flow sampling over a defined range of purge volumes.
Methods to Reduce Short-Term Variability (Task 3): For this task, the effect of sample collection method on monitoring variability was evaluated, as well as whether improved sampling methods and procedures could reduce the short-term variability in groundwater monitoring results. The field program involved collecting samples from a set of eight monitoring wells using five different sample collection methods. Each method was used for three sampling events, resulting in a total of 15 sampling events. The five sampling methods evaluated were Low-Flow Sampling with Purge to Parameter Stability (reference method), Low-Flow Sampling with Constant 24L Volume Purge, No Purge Low-Flow Sampling without In-Well Mixing, SNAP (No Purge Passive Sampling), and No Purge Low-Flow Sampling with In-Well Mixing.
The results support the following observations:
- For chlorinated solvents, the long-term change in concentration is typically slow: Attenuation half-lives are usually greater than five years and are often greater than 10 years.
- Most concentration change in monitoring records is due to short-term variability: Most of the variability in conventional groundwater monitoring records based on quarterly or semi-annual monitoring of conventional monitoring wells is attributable to short-term variability rather than long-term changes in constituent concentrations. Only 30% to 40% of the variability is due to the long-term trend. Monitoring less frequently (i.e., an average frequency of once per year or less) will improve the efficiency of long-term monitoring.
- There are large differences in the amount of short-term variability between different monitoring wells: Within a single plume or site, there can be large differences in short-term variability between monitoring wells. Identifiable characteristics (e.g., aquifer permeability, depth of well screen below water table, depth to groundwater, and well location within the plume) explain only a small amount of the observed differences in variability.
- Sample collection methods and procedures can affect short-term monitoring variability: The specific sample collection methods and procedures used to collect groundwater samples from monitoring wells can affect short-term variability. Certain types of no-purge sampling may increase monitoring variability. When using low-flow purge and sample collection methods, the consistent use of certain sampling procedures such as ensuring a constant sample collection depth within the screened interval, using the bottom fill procedure for volatile organic analysis (VOA) vials, and not removing small bubbles from VOA vials can reduce variability.
Based on the results of this project, the following changes are recommended to make long-term monitoring programs more efficient and cost effective:
- Monitor annually or less frequently
- Consider monitoring schemes that vary the time period between sample collection
- Utilize improved (i.e., more reproducible) groundwater sampling procedures
- Watch out for stratified monitoring wells
- Reconsider and re-evaluate the utility and purpose of field duplicates
- Don’t overreact to short-term concentration changes
- Consider replacing highly variable monitoring wells.