3.1.4.2 Sampling approach

Decisions about the general sampling approach for a project are crucial in ensuring the data will meet the project objectives. Project planners may elect to employ ISM, traditional discrete sampling, or a combination of both. The optimum approach depends on the CSM, the nature and extent of contamination, project objectives, and regulatory requirements. For example, if applicable regulations or policies call for comparison of an action level to the maximum detected concentration, discrete sampling is typically necessary because ISM provides only an estimate of the mean and cannot be used to estimate the maximum. An alternative is to place multiple small SUs over the DU and demonstrate compliance using the maximum value obtained from the SUs. If estimation of the maximum is identified as a sampling objective and a discrete sampling approach is employed, project planners should consider using a high-density discrete sampling plan. A traditional low-density discrete or low-density SU sampling plan is unlikely to produce an accurate estimate of the maximum concentration in the area of concern.

The fact that discrete sample data can be combined and recombined in a variety of data sets after sample analyses is both a strength and a weakness; it is possible to use such data in biased or uninformed combinations. ISM data, on the other hand, are more specifically tied to the individual volumes of soil that were sampled. This too is both a strength and a weakness. Project-specific objectives and constraints as well as the decision mechanism to be used must be considered during the selection of DUs so that multiple issues are addressed.

Project planning decisions are seldom based on technical considerations alone. In addition to the uncertainties associated with different sampling approaches, project planners must consider various regulatory requirements, as well as resource and budget limitations. Nevertheless, to appropriately control decision errors, a clear and complete understanding of the technical strengths and weaknesses of sampling approaches must be considered before an approach is selected. When weighing these strengths and weaknesses, the available information on soil heterogeneity and the project goals for decision errors should be fully considered. As is always the case in environmental management, sampling plans must balance the necessity of controlling costs with the need for a reasonable degree of certainty. The decision to use ISM, discrete sampling, or some combination to investigate areas that have elevated concentrations should be made with a clear understanding of the limitations of both techniques and clear definitions of what constitutes a potential risk on a scale that considers both concentration and volume relative to specific project objectives.

Table 3-1 lists many items to be considered during the planning, implementation and use of ISM data. Included in this table are example project objectives. The project team must identify the data necessary to address all study questions and meet the project objectives.

Table 3-1. Considerations to address during systematic planning for ISM sampling

Factors Issues
Conceptual site model Source(s)
Contaminants
Action levels
Distribution
Migration mechanisms (fate and transport including preferential pathways)
Parameter to estimate

Parameter(s) the project needs to estimate:

  • Mean
  • UCL
  • Other
Project objectives and decisions

Overall project goal:

  • Human health risks of exposure areas
  • Risk to ecological receptors
  • Effectiveness of cleanup/removal
  • Extent
  • Leaching potential
  • Other

 

Source, nature, and numerical value of the action level
Requirements for precision, total error, and decision quality
End use of the data
Decision units
(see Section 3.3)
Number of DUs

Rationale for DU selection:

  • Human health exposure area
  • Ecological exposure area
  • Source area
  • Excavation sidewall or floor
  • Location of DU
Scale of the decision (spatial and/or temporal scale that was originally intended in the development of the action level)
Variability within DU (anticipated and/or measured)
Location of DU

Size of DU (three-dimensional [3-D]):

  • Surface dimensions
  • Subsurface dimensions
  • Depth interval
  • Geologic strata or soil horizon
Sampling units
(see Section 3.3)

Subdivision into SUs:

  • Number if SUs composing each DU (from one to many)
  • Rationale for SUs
  • Location of SUs
  • Size (surface dimensions, depth, interval)

Portion of DU represented by SU:

  • Targeted areas of suspected high contamination
  • Random placement
  • Extrapolation allowed
Background SU
Increments Number of increments in each ISM sample
Increment spacing
Targeted bulk sample volume/mass
Approximate increment volume/mass

Sampling pattern:

  • Simple random
  • Systematic random
  • Stratified random
  • Other
Replicates
(see Section 7.3)
Number of replicates

Type and purpose of replicates:

  • DU replicate
  • SU replicate
  • Field replicate
  • Laboratory replicate
  • Instrument replicate
Targeted soil fraction

Basis for targeted fraction of soil:
Targeted fraction of soil (e.g., grain size, geologic unit, etc.)

  • All fractions
  • Multiple fractions
  • <2 mm fraction or a fraction >2mm
Field procedures
(see Section 5)

Field steps to control sampling errors:

  • Correct sample device
  • All potential increments in the SU equally available to sample device
  • Consistent increment size
  • Coverage of SU
  • Cross section of SU collected
  • Decontamination between ISM samples (not between ISM increments)
  • Field mixing
  • Field subsampling

 

 

Sampling procedures including aspects of “correct sampling” (see Section 2.5)
Lab procedures
(see Section 6)
Mixing in the laboratory

Particle size reduction or selection (where appropriate):

  • Grinding or milling
  • Sieving
  • Sieve size
Subsampling
Statistic calculated
(see Sections 4 and 7)
Arithmetic mean (of replicates)
Variance (of replicates)

95% UCL:

  • Chebyshev
  • Student’s-t
Metric used to evaluate SUs
Decision mechanism
(see Sections 4 and 7)
Source, nature, and numerical value of the action level

How will the decision be made? Which decision mechanism (DM) will be used (i.e., how will you decide whether further action is necessary or not)?

  • DM 1: Comparison of one ISM sample from the DU to the action level
  • DM 2: Comparison of the mean of replicate data from the DU to the action level
  • DM 3: Comparison of the 95% UCL on the mean of replicate data from the DU to the action level
  • DM 4: Comparison to background
  • DM 5: Combining DUs
  • DM 6: Extrapolating from sampled to unsampled areas
  • DM 7: Evaluating oversized DUs
  • Other