Manuscript (Go to Abstract)
Contaminated Groundwater Analytical, QAQC, Detection Limits
At many locations, groundwater has been contaminated by spills, leaks, and disposal of chemicals and metals. Unfortunately, metals, chlorinated solvents, gasoline components, pesticides, fertilizers, etc. are frequently detected in groundwater at concentrations which create a threat to human health and/or the environment. It is necessary to define the nature and extent of the contaminated groundwater and to design and implement remedial response action to reduce the threat to acceptable levels.
The site assessment, the remedial investigation, the remedial design, the remedial action, and the site closure are driven by sampling and analysis of the affected media (typically: air, soil, sediments, and water). From the initial "discovery" to the final site "closure" analytical results control the entire process. Table 1 shows the interaction of the project phases with the sampling, analytical, and QAQC issues.
Sampling, sample handling, analytical QAQC, and data management can be expensive and time-consuming. In some cases, there is a tendency to limit the media tested, the number of samples, the analytical slate, the detection limits, and the QAQC detail in order to save time and cost; it is critical to make objective decisions regarding the analytical issues and to think ahead to the overall project.
Each step of the overall project depends on the data generated during the previous steps; errors, lack of data, and incorrect assumptions (based on limited data) can have a negative impact on the overall project.
Known or observed spill or leak, taste or odor in air or drinking water, historical site activity, visible chemical impacts (stained ground) are all indicators of possible contaminated media. The objective during this phase is to review and evaluate all the available data and decide if there is a human health or environmental risk; the quality of the available data will be evaluated (QAQC, detection limits, consistency, etc.) Can the chemicals of concern (COCs) be determined?
There are state and federal protocols for performing environmental site assessments; the protocols include sampling, analysis, and QAQC guidelines. However, the objective is to generate the data needed to determine the risk posed by the site and to decide if remedial response will likely be required. It is critical to generate sufficient correct data. Site assessments are typically done in phases with subsequent phases based on the data generated by the previous phases.
Sufficient analytical data is required to determine if the site represents a significant health or environmental risk. Some sites require a complete analytical slate (VOC, SVOC, metals, pesticides, PCBs, etc.) to determine the COCs; many sites only require VOC and metal analyses. Figure 1 shows a large-scale Superfund site that contained metals, PCBs, and over 200 other chemicals; the site analytical database reduced the COCs to benzene, benzo-a-pyrene, 1,2-DCA, vinyl chloride, PCB242, and arsenic.
Remedial Investigation, Risk Assessment, Feasibility Study
The objectives of this phase are to:
Figure 2 shows a schematic of a complex site which contained chlorinated solvents, aromatics, and alcohols in the shallow groundwater; remedial action was required in eight separate areas of the site; the comprehensive analytical database determined the remediation sequence for each separate area.
During the design phase, the physical details of the remedial action are determined and the several steps and facilities are connected. Analytical data provides the basis for the design. Some "new" analytical data may be generated during this phase to "fix" the location and size of specific components such as extraction wells, sample stations, nutrient distribution lines, etc.; this analytical data will typically focus on the COCs at higher concentrations.
Some groundwater analytical data will be generated during the construction phase to confirm COC trends in the major active remediation areas and to insure continuous protection of the at-risk receptors; this data will not have a significant impact on the construction phase.
Operation, Maintenance, Refinements
This phase is usually the most costly and largest duration of a project and is driven primarily by the historical and current analytical trends. It is critical to generate sufficient groundwater analytical data to allow efficient conduct of the remedial action. As the remedial action proceeds, "things" will change, and refinements or adjustments are necessary to insure effective remediation; the analytical data defines the nature of the needed refinements; the sampling frequency, the QAQC, and the detection limits can be adjusted based on concentrations and overall progress.
Figure 3 shows a typical groundwater progress curve; the TBA detection limit will be 50 ppb until benzene approaches the compliance criteria; the TBA detection limit can be lowered as necessary when (and if) TBA becomes the "critical" contaminant.
Most groundwater remediation projects are converted from active remediation to natural attenuation when the COC concentrations have decreased to the level where natural processes will protect human health and the environment. During natural attenuation, the COC concentrations are decreased by bioremediation, diffusion, dilution, dispersion, etc. A comprehensive, active, groundwater analytical database is critical to converting to natural attenuation and to monitoring the status of natural attenuation. Figure 4 illustrates the data used for the natural attenuation decision. Figure 5 shows eight years of natural attenuation monitoring data.
Typically, the site closure requirements are precisely defined by regulatory programs and/or by agreements between the PRP and the authorities; in most cases the closure requirements are COC concentration levels in the exposure pathway (water, air, soil); groundwater is a common exposure pathway. The groundwater remediation criteria is typically the federal MCLs for the COCs at the point of exposure; the QAQC and detection limit requirements need to produce credible, precise, and reproducible groundwater data to justify closure.
The duration of long-term monitoring can be 20-30 years after closing; the purpose is to track natural attenuation progress and to detect any trends that may require response action. Figure 6 shows a typical long-term data trend. Figure 7 shows an area where significant "rebound" has occurred.
Large Scale, Complex Superfund Site
Benzene/TBA Progress Curves
The entire groundwater remediation process is controlled by groundwater analytical data. Analytical data can be expensive so there is a tendency to reduce the frequency and precision of groundwater analysis to save time and money. It is important to generate the analytical data necessary to effectively control the project; effective project control will save more money than reducing the analytical program. It is groundwater analytical data trends that control the remediation process; it is preferable to have many data points over time even with less intense QAQC validation; precise, validated data on a limited number of occasions can miss actual critical concentration trends.
Chemicals of Concern (COCs)
Many instances of contaminated groundwater involve many chemicals. For example, spilled or leaked gasoline contains over 200 chemicals all of which can impact groundwater to a degree; however, not all of these chemicals are critical in terms of human health or environmental risk or in terms of the remedial response action.
Review of the site history may suggest some chemicals of concern. Initial groundwater analyses at low detection limits for the full list of VOCs, SVOCs, pesticides, PCBs, and RCRA metals is necessary to establish the COCs; the samples need to represent the vertical and areal extent of the contaminated groundwater.
Common degradation products of the spilled/leaked chemicals may become COCs over time; this issue needs to be factored into the review of the groundwater analytical data for the purpose of selecting COCs; the QAQC, detection limits, and surrogate recoveries should anticipate likely degradation products. Figure 8 shows vinyl chloride concentrations over time for a dry cleaning site; initially PCE, the DCE and DCA were the COCs but the DCE and DCA degraded to vinyl chloride which became a COC as the remediation progressed. Since vinyl chloride typically has a remediation criteria of 2 ug/L (MCL) at the exposure point, full QAQC data validation is required.
Risk-Based Critical Chemicals
At most sites it is the risk created by one or more of the COCs which drive the remediation process. Once the COCs are determined, it is critical to review the MSDSs for the COCs to determine which COCs create the greatest risks; likely COC degradation products also need to be considered. For example, aerobic in-situ bioremediation of groundwater contaminated by leaked gasoline can yield TBA and acetone as degradation products; the TBA and acetone concentrations will likely increase during the active remediation. However, the low toxicity of TBA and acetone (high MCLs) would leave benzene (low MCL) as the primary COC. Figure 9 shows the TBA, MTBE, acetone, and benzene at a progress monitoring well for an eastern PA gas station.
Sampling and Analysis Plan (SAP)
Typically a groundwater sampling and analysis plan is developed for each site. For complex sites it may be desirable to develop a sampling and analysis plan for each major project phase. The plan will detail sampling location and frequency, well purging, sample preservation and handling, analytical slate, analytical methods and detection limits, QAQC requirements, and analytical data reporting requirements; the details will vary with the analytical data requirements for each project phase.
Detection limits are detailed in the SAP. It is critical to recognize just what detection limits mean in regards to specific contaminants. Tables 2, 4, and Figure 12 are typical groundwater summary reports for volatile organic chemicals and metals. Usually the detection limit is the lowest level at which the COCs can be quantified for a particular sample. "ND" means non-detect, and there is a concentration associated with all "NDs"; "ND" (for example, at 50 ppb for benzene in a particular sample) means that benzene may be present in the sample but the concentration is lower than 50 ppb and can not be quantified. Detection limits vary with samples, media, and chemicals. Some analytical reports list >50, meaning the detection limit for that particular analyte in that particular sample is 50 ppb.
To determine if a particular "ND" chemical is present in a sample, it is necessary to inspect the GCMS graph. If the chemical is present, but the concentration is below the detection limit, there will be a peak on the GCMS graph at the time specific for that chemical; Figure 10 illustrates this issue. If the detection limit for a COC is above the risk-based remediation criteria for that COC, then it will be necessary to rerun the original sample at lower dilution rates until the detection limit is less than the remediation criteria.
Table 5 illustrates the variability of detection limits in response to chemical concentrations in analytical samples; the elevated levels of butyl alcohol and naphthalene drive up the detection limits for the other chemicals. If it were necessary to quantify vinyl chloride at <10 ppb on these samples, it would be necessary to analyze the sample with little or no dilution; this would force the laboratory to run several "cleaning blanks" to remove the residual TBA and naphthalene from the instrument.
EPA and various state agency protocols define the QAQC requirements for analytical samples, the objective is to insure that the analytical results are sufficiently precise and accurate for the intended use of the results.
Compliance samples, receptor protection samples, and site closure samples typically will require full QAQC validation in terms of surrogate recoveries, calibration, trip blanks, instrument blanks, field duplicates, etc. However, many samples, used to track progress and to evaluate trends, do not require full QAQC validation. The level of QAQC required should be specified on the sample chain-of-custody form. Thus, it is critical to develop a close working relationship with the analytical laboratory.
Concentrations Reported as BDL
Program Controlled Analytical Capability
Analytical Data Summary
Analytical Data Summary
It is important to recognize what properly validated groundwater analytical results actually represent. Over the last several years, remediation criteria for many chemicals have decreased; for example, the benzene criteria at many sites is 5 ppb, and the vinyl chloride criteria is 2 ppb. As the progress trends are evaluated for vinyl chloride and benzene on a particular site, the analytical variability may exceed the remediation criteria. Figure 13 shows the analytical variability in Laboratory Performance Evaluation Samples; for example, the actual benzene content was 43 ppb with an analytical range of 23 ppb to 66 ppb; the 1,2-DCA content was 61 ppb with a range of 30 ppb to 93 ppb.
It is essential to generate sufficient groundwater analytical data over time so that the real COC trends can be evaluated without distraction by analytical variability.
There are few remediation project issues which can not be resolved by "more" data. However, an effective database management system is needed so that "new" data can be readily incorporated into the database. The same database should be used for all groundwater remediation data; sampling data, field data, chain of custody, analytical results, and QAQC validation are all included in the database; the database is controlled by a central person with appropriate support.
Scatter versus Trends
As discussed, there is a lot of variability in groundwater analytical results especially at lower detection limits. Figure 14 shows the benzene, vinyl chloride, and 1,2-DCA concentrations at MW INT-134 over an 8-year period (1995-2003); there is considerable scatter in the analytical data, but the trends over time are consistent with the remediation plan. Usually time and supplemental data will distinguish scatter form trends.
Contaminated groundwater needs to be evaluated and remediated in a timely, cost-effective manner to protect human health and the environment. Analytical results control and drive this entire remedial response process, and the project team must actively direct the analytical program.
Send mail to webmaster with questions
or comments about