Incremental Sampling Methodology
Incremental Sampling Methodology (ISM) is a technique designed to statistically reduce or limit variability associated with discrete sampling. It provides unbiased, representative and reproducible estimates of the mean concentration of analytes in a specific area of interest, called a “decision unit.”
Interest in (ISM) has grown in recent years largely because the approach, when applied correctly, can significantly reduce sampling uncertainty. This in turn can increase the probability that sample data is more representative of average site conditions at hazardous waste sites, thereby strengthening decision making at the site.
DoD Directive 4715.11 requires an assessment of munitions on operational firing ranges and the Military Munitions Response Program (MMRP) requires similar assessments of closed ranges. These requirements have historically been met by dividing the sites into exposure units and collecting discrete or small-scale composite samples for explosives analysis by EPA Method 8330A. Results were assumed to be normally distributed and representative of the entire exposure unit. The U.S. Army Corp of Engineers Cold Regions Research and Engineering Laboratory (CRREL) demonstrated that these assumptions were false. Traditional field sampling and laboratory sub-sampling methodologies are not able to accurately represent conditions at the site. As a result, improved sampling methodologies became a priority.
Aspects of ISM have been utilized for several years. The work on surface soil sampling at DoD firing ranges and relevant methodologies were published in 2007 by CRREL in a document titled “Protocols for Collection of Surface Soil Samples at Military Training and Testing Ranges for the Characterization of Energetic Munitions Constituents.” State regulatory agencies have started to adopt and issue guidance for using MIS protocols at hazardous waste sites, and these guidelines are encompassing broader constituents than explosive residues (Alaska DEC, Hawaii DoH HEER, Ohio EPA VAP). The Interstate Technology Regulatory Council (ITRC) began working on ISM in 2009 and is scheduled to issue guidance in early 2011. Several states are represented in this workgroup including Florida, California, Massachusetts, Texas, Oklahoma, Arizona, and New Mexico. These activities are expected to increase interest and use of ISM at hazardous waste sites around the country.
On the surface, ISM might be confused with more traditional compositing techniques; but this assessment misses the point of incremental sampling. It is designed to statistically reduce or limit the variability associated with discrete sampling. This variability is attributed to compositional and distributional heterogeneity. Composite sampling with limited increments does not adequately address these issues. It tends to add uncertainty unless the entire sample is already homogeneous (an unlikely occurrence in the field). ISM differs from typical composite sampling in two ways: (1) the number of grabs is much greater, and (2) the combined grabs represent the entire area of interest (defined as a decision unit). This process is controlled through a detailed sampling plan.
Heterogeneity introduces error to the sampling process; ISM addresses the two main sources of sampling error attributed to heterogeneity (refer to figure 1 for a comparison of discrete and incremental sampling).
- Compositional heterogeneity describes the distribution of contaminant concentration across the range of particle sizes making up the population. It introduces “fundamental error” (FE) when insufficient mass is collected and analyzed. ISM addresses this source of error by controlling the sample size that is collected and analyzed.
- Distributional heterogeneity is a function of spatial variability and occurs when particles are not randomly distributed across the population. It introduces “grouping and segregation error” (GSE) when the sample consists of too few increments to capture or represent the spatial variability. ISM addresses this source of error through the collection of multiple, randomly located sample increments.
There are two components to ISM: field sampling and laboratory analysis. In the field, an approximately 1 – 5 kg sample is collected and sent to the laboratory for processing in its entirety. Thirty to 100 increments (or more depending on the expected distributional heterogeneity for the decision unit) of uniform size are collected across a grid formation that represents the entire decision unit. Determining the appropriate size of the decision unit is a critical aspect of ISM and one that is usually detailed in a sampling plan that is reviewed and approved prior to field mobilization. Some sampling plans may require that the sample be field dried and size reduced; but more often, the entire 1 – 5 kg sample is delivered to the laboratory for additional processing.
In the lab, the sample is processed in accordance with program requirements, site specific sampling plans, or by the laboratory’s SOP. In general, these procedures include air drying, sieving, particle reduction (grinding or milling), and a multi-increment sub-sampling performed in accordance with established guidelines. The entire sample is spread into a grid formation and the sub-sample is generated using similar techniques employed in the field, only on a much smaller scale. The sub-sample is typically in the 10-30g size. Alternatively, some sampling plans will allow the use of a sample splitter (e.g., rotary riffle splitter or equivalent). Essentially the same objective is achieved regardless of the technique used.
This entire sub-sample is used for analysis. Batch QC includes replicate analysis to verify homogeneity; multi-incremental sample replicates are usually normally distributed with very few outliers. Thus, the goal of limiting discrete sample variability is achieved.
ISM may not be appropriate to apply universally at a site. It was designed as a surface sample collection technique, with the goal of reducing sources of sampling error for the entire decision unit. More than one sampling approach may be applicable for complete site characterization. For example, ISM may be useful in characterizing surface soils, excavation floors and walls, or contaminated stockpiles, but not for meeting other site characterization goals. Sampling guidelines issued by state and federal agencies will typically include details about the use of ISM in site characterization.
Method and/or Analyte Class Considerations
EPA Method 8330B has incorporated ISM. This was a significant change when compared to the previous method, 8330A. Method 8330B requires air drying and sieving of samples to less than 2 mm. The protocol then requires milling when samples are collected at firing range sites, or mortar and pestle grinding for ammunition plans and depots. The Department of Defense has issued specific guidelines for implementing EPA SW-846 8330B to address sampling and analytical methodologies, concepts, and QC requirements of the method.
There are a few key differences between different versions of EPA 8330x:
- 8330 and 8330A produce detection limits ~10x higher than 8330B due to differences in the extraction procedures;
- All versions of the method require air-drying, sieving, and grinding, but only 8330B requires ISM protocols;
- Grinding procedures differ by sample source:
- mortar and pestle for ammunition plants and depots,
- ring puck mill or equivalent mechanical grinder for firing ranges;
- 8330B always requires ISM, regardless of the sample source.
EPA Method 8330B is currently the only method with specific guidance on the use of ISM to reduce sampling error. The concepts may apply to other groups of analyses. Sampling and analysis plans should include procedural guidelines to ensure that field and laboratory applications meet data quality objectives at the site. The use of milling or grinding will not be appropriate for all classes of target analyses. For example, milling may be appropriate for explosives and metals, but inappropriate for volatile organics. State specific guidelines (in Alaska, Hawaii, and Ohio) are useful in assessing the applicability of ISM and considerations that must be taken into account during the development of the sampling plan. These guidelines include information on analyte-specific considerations, minimum sub-sample size for analysis, and techniques for obtaining a representative sub-sample.
Alaska and Hawaii include volatile organics in state specific ISM guidance documents. The procedure for collecting the ISM sample for volatiles includes preservation in methanol. This precludes low level sample analysis, thereby resulting in detection limits that may be significantly higher than data quality objectives for other states. Therefore, ISM may not be appropriate for collection of volatile organics outside of a narrow application.
The US Army Corp of Engineers is actively conducting research and demonstration projects to evaluate the use of ISM on a number of contaminated sites. Contaminants of concern include metals, polycyclic aromatic hydrocarbons, perchlorate, and white phosphorous. Each of these classes of compounds presents unique sampling challenges. Sampling plans should be written to include guidance to assess sampling precision in the field and in the laboratory.
Benefits and Limitations
According to the National Defense Center for Energy and Environment (NDCEE), benefits of ISM include:
- The approach adds flexibility to the planning process as regulators and stakeholders evaluate the configuration and the number of decision units needed to characterize the site;
- Depending on the size of a decision unit, ISM sampling can reduce the number of samples being analyzed at the site thereby reducing analytical costs;
- The process offers improved precision (e.g., lower relative standard deviation between field replicates and between laboratory replicates) indicating that ISM sampling provides a more accurate representation of average site conditions;
- More representative data supports effective decision making;
- Indirect cost benefits are derived from the higher level of confidence in the data (i.e., consistency of data across sampling events, fewer anomalies, and more informed decision making).
There are a few limitations that need to be considered:
- Limited acceptance and regulatory guidance still exists, although this is changing as more agencies integrate ISM into sampling guidance documents;
- Sampling plans have to define the decision unit to be composited into a multi-incremental sample;
- Fewer labs have the capability to process ISM samples;
- Cost per analysis is higher to account for additional processing steps, although this cost is usually off-set by fewer samples per investigation;
- The technique is not appropriate for all analyte classes or sample types (e.g. shallow surface soil sampling has been evaluated, but cost benefits should be re-examined for deeper sampling).
Regulatory Acceptance and Guidance Documents
- State of Alaska Department of Environmental Conservation, Division of Spill Prevention and Response, Contaminated Sites Program, Draft Guidance on Multi-Incremental Soil Sampling, March 2009
- State of Hawaii Department of Health, Office of Hazard Evaluation and Emergency Response, Technical Guidance Manual for the Implementation of the Hawaii State Contingency Plan, Section 4: Soil Sample Collection Approaches, November 2008
- US Army Corps of Engineers Interim Guidance 09-02, 20 July 2009, Implementation of Incremental Sampling (IS) of Soil for the Military Munitions Response Program
- Ohio EPA Division of Emergency and Remedial Response, Standard Operating Procedure 2.6.1: Multi-Incremental Sampling for Soils and Sediments, January 2007
- Interstate Technology Regulatory Council, Project Introduction: Incremental Sampling Methodology, July 2009
- National Defense Center for Energy and Environment, Office of the Assistant Secretary of the Army for Installations, Energy and Environment, Concurrent Technologies Corporation, Multi-Incremental Fact Sheet
For more information visit Incremental Sampling.