Department of Health Seal

TGM for the Implementation of the Hawai'i State Contingency Plan
Subsection 3.4


A decision unit (DU) is an area where a decision is to be made regarding the extent and magnitude of contaminants with respect to potential environmental hazards and associated risks to human health and the environment, as described in Section 13. A DU is necessarily a volume as well as area of soil. Decision units are designated in terms of size and location in a manner that ensures the objectives of the site investigation will be accomplished.

An important goal of a site investigation is to estimate the mean concentration of a contaminant for a designated DU volume of soil. A DU can be thought of as "The entire volume of soil that you would send to a laboratory for testing if this was indeed possible." Tested as a single sample, the concentration of the contaminant reported by the laboratory would represent the true mean for the DU volume of soil as a whole. This is usually not feasible given the large volume of soil assigned to a DU. A representative sample or samples of the soil must instead be collected and tested. As discussed in Section 5 while this can in theory be accomplished using discrete samples, Multi Increment samples are far more reliable and ultimately more cost efficient.

The designation of DUs is, importantly, independent of the sampling method used to characterize the area targeted. Practitioners new to sampling theory and Multi Increment sampling methods (Subsection 4.2) are often concerned about potential “dilution” of contaminated areas “hot spots” in comparison to traditional, discrete soil sampling methods (refer to Subsection 4.3). This is addressed through the designation of well-thought-out DUs that establish the desired resolution of the investigation. Although not directly referred to, the concept of “DUs” was an inherent part of past, discrete soil sample investigations. Discrete soil sample collection points were typically designated based on a desire to characterize contamination in one area versus another. As discussed in the subsections below, the area intended to be represented by a single, discrete sample point (or cluster of sample points) is designated as a separate DU for characterization. A large-mass, Multi Increment sample is then collected from multiple (e.g., 30-75+) locations within this area rather than reliance on a small, discrete soil sample collected from a single location. The number of DUs designated for a particular investigation not coincidentally often corresponds with the number of discrete soil samples or clusters of samples that might have been collected under past approaches.

Decision units can be designated for characterization of surface soils and/or subsurface soils. Designation of DUs for surface soil is described in Subsections 3.4.2 and 3.4.3. Designation of DUs for subsurface soils is discussed in Subsection 3.4.4.

Figure 3-10. Example Spill Area and Exposure Area Decision Units

Establishing clear site investigation objectives and designating DUs to achieve these objectives early in the site investigation design helps develop an effective sampling approach to characterize a site. This ensures that adequate data are available to prepare an Environmental Hazard Evaluation (EHE, see Section 13), develop removal or remedial options, or carry out further investigation. A discussion of DU investigation strategies in the context of EHEs is also included in the HDOH guidance document Evaluation of Environmental Hazards at Sites with Contaminated Soil and Groundwater (HDOH, 2016).


The designation of DUs for characterization is unique to each site and depends in part on the specific type of environmental hazard under investigation (see Step 8 in Subsection 3.2 and Section 13). Decision units generally fall into two categories (Figure 3-10): 1) Spill Areas, and 2) Exposure Areas. As the name implies, a Spill Area DU represents a specific area where releases of hazardous substances are known or suspected.

Example Spill Area DUs might include former waste storage or disposal areas, the area (and volume) of contaminated soil under and around a leaking underground tank or an area of a site where contaminated fill material is suspected to have been placed. Identification and characterization of such areas is a common objective of an environmental investigation.

An Exposure Area DU is an area where receptors, human or ecological, routinely access and could be exposed to hazardous substances. Example Exposure Area DUs might include the yard of a residence or an unpaved area of exposed soil at a commercial or industrial facility. Exposure Area DUs might include isolated (unknown) spill areas, but as described below determination of the exact location of such areas is not required for assessment of potential direct-exposure risk, provided that the sample (or samples) collected from the DU is representative of mean contaminant concentration for the DU as a whole. Large Exposure Area DUs initially anticipated to be clean might be divided up into small DUs for re-characterization if they are found to be contaminated, in order to help optimize removal or remedial actions. The need for smaller decision units for future response actions should in general be considered up front, however, in order to minimize the time and cost required to resample an area.

The appropriate type, size, shape and number of DUs for a given project is necessarily site-specific and must take into consideration the historical, current and future use of the site. A strong understanding of the historical use of a site is critical to the designation of DUs (see Subsection 3.1.1). Historical site plans, insurance maps, historical aerial photos and discussions with past or current workers are very useful for determining initial DU areas. Obvious or suspected spill areas should normally be investigated as separate DUs. This includes structural remnants of potential hazardous substance storage or disposal areas, suspect pits and trenches, stained areas and low points where runoff could have collected.

Investigation objectives and approaches can vary over time as the project proceeds and alternate DUs may need to be established to assist in response actions or long-term management of sites. For example, DUs established for site characterization purposes may need to be refined for the removal or remedial phase of the project to better isolate high-priority areas and optimize resources available for cleanup.

When contamination poses multiple, potential environmental hazards then the smallest DU area and depth (i.e., highest resolution) should be selected to characterize the area. DUs for different media (e.g. soil vs. groundwater vs. soil gas vs. indoor air) should in most cases be treated separately, even if they are investigated concurrently. DUs for different media could overlap but may have different decisions associated with them.


Figure 3-11. Example Exposure Area Decision Units  Figure shows areas where residents or workers may be exposed to contaminants in soil on a regular basis. For ecological receptors, an example exposure area is the home range. The size & shape of exposure area decision units depends on the targeted receptor and the desired scale of the evaluation.

In the absence of a known or suspect spill area, the most appropriate Decision Unit for relatively immobile contaminants that primarily pose direct-exposure, toxicity-based hazards (e.g., lead, arsenic, dioxins, polychlorinated biphenyls [PCBs], polynuclear aromatic hydrocarbons [PAHs], etc.) is the assumed exposure area for the site receptor(s). "Exposure areas" are areas frequented by residents or workers or wildlife that may come in contact with contaminants in soil on a regular basis (see Section 13). Examples include residential yards, schoolyards, playgrounds, gardens, open areas on commercial/industrial properties, home range, etc. (refer to Figure 3-11). For exposure area DUs, the primary question is "What is the mean concentration of the target contaminant across the exposure area as a whole?" Exposure area DUs can be based on current land use (e.g., an open area of a commercial or industrial site) and/or future use of the area (e.g., proposed residential lots).

The top two to six inches of soil is generally considered for surface soil DUs, depending on the site-specific DQOs (USEPA 2011d; CalEPA 2013). The top 0-6 or 0-4 inches of soil are commonly selected for surface soil Dus in Hawaiian Islands investigations. Exposure area DUs for residential properties typically encompass the entire yard, but could also focus on play areas or other areas of the yard that are frequented most. DUs for apartment complexes should focus on open common areas. For future redevelopment projects that involve single-family homes, the size of a hypothetical residential lot is generally assumed to be 5,000 ft2 (see Subsection 3.5).

The location and size of exposure area DUs for commercial or industrial sites is necessarily site specific. DUs should be based on the location of exposed areas of soil and use of the site by workers. As a default and especially for undeveloped properties, exposure areas should be initially limited to half an acre or approximately 20,000 ft2. When possible, designate DUs and investigate the site in a manner that allows future, unrestricted land use (i.e., residential land use, see Section 13). This will minimize the need for restrictions on future site use or delays in redevelopment.

Recall that soil EALs are not intended for direct comparison to individual, discrete sample data, although this may be useful for general screening purposes (refer to Step 8 in Subsection 3.2.1). The EALs are intended for comparison to the mean concentration of a contaminant in soil within the designated DU. Soil action levels for direct exposure risks, for example, are intended to address long-term, "chronic" exposure to very low concentrations of contaminants in soil. Exposure is assessed in terms of random exposure to contaminants in soil throughout the DU area over a period of many years (see Section 13).

It is not necessary and indeed not practical to attempt to identify the "maximum" and "minimum" concentration of a contaminant within the DU. If a contaminant is present, then at some, minute mass of soil the concentration of the contaminant will necessarily be 100%, or 1,000,000 mg/kg. It is also highly likely that the contaminant is entirely absent in scattered, small masses of soil within the same area. Estimation of the representative, mean concentration of a contaminant for a DU requires that a representative proportion of both higher and lower contaminant concentration areas be included in the sample data. As discussed in more detail in Section 4, this is best accomplished by the collection of a Multi Increment (versus discrete) sample from the DU. Replicate field samples are routinely collected from DUs to evaluate the precision of the original sample data.

Examples of DUs based on exposure areas are included in Subsection 3.5. Decision units based on exposure areas can also be established for ecological risk assessments. Additional guidance on decision units for ecological risk assessments will be included in the TGM in the future.


For use in this guidance, a "spill area" is defined as a discernable area of elevated contamination in soil that can be mapped with respect to surrounding areas. Examples include areas with obviously contaminated and stained soils, unpaved areas used to store or mix hazardous chemicals, known waste disposal areas, areas immediately adjacent to transformer pads or other types of equipment that may have contained hazardous chemicals, releases from pipelines, etc.

The isolation and evaluation of individual spill areas is generally necessary to evaluate environmental hazards associated with soil leaching, vapor intrusion and gross contamination hazards (see Step 8 in Subsection 3.2 and Section 13). This applies to most releases of petroleum, solvents and highly mobile pesticides like atrazine and ametryn. In these cases, the appropriate question is "What is the mean concentration of the contaminant(s) within the volume of soil impacted by the spill"? Spill area DUs are sometimes recognizable based on surface staining but this is not always the case (e.g., PCP-related dioxins and arsenic). Typical spill area DUs are a few hundred to a few thousand feet in area and can extend to varying depths, depending on the nature of the contaminant released (see Subsection 3.5). For stockpiles, a default DU volume of twenty cubic yards is recommended for testing of soil stockpiles impacted by contaminants that could pose potential leaching concerns (HDOH 2011b).

If the target contaminant at the site poses leaching, vapor intrusion and gross contamination hazards, then the spill area should be designated as a separate DU for characterization. For example, a spill area associated with a petroleum release around an aboveground storage tank should be identified as a separate DU and appropriately investigated. This is because petroleum contamination can pose multiple environmental hazards, including leaching of contamination to subsurface groundwater resources, intrusion of vapors into overlying buildings and nuisance or even explosion hazards associated with grossly contaminated soil. It is inappropriate to incorporate data outside of the spill area in the evaluation of these types of hazards.

It may also become useful to identify and isolate small spill areas within a larger exposure area DU if the mean concentration of the target contaminant exceeds action levels and remediation is required. For example, isolating and remediating spill areas at a site heavily contaminated with arsenic can help optimize remedial actions to reduce average exposure concentrations (see Subsection 4.2.9 and 4.3.3). The cost-benefit of subdividing an initially larger DU into smaller DUs for additional, more detailed characterization is necessarily site-specific. The need for smaller DUs and a better resolution of contaminant distribution should ideally be taken into account as part of the initial designation of DUs at the site. Careful planning ahead of time will allow decisions to be made based on the original set of data collected and avoid the added time and expense of additional investigation.

Isolation and remediation of spill areas within an exposure area DU may also be necessary to prevent localized but heavily contaminated soil from being spread out across a larger area during future construction activities. For example, PCB-contaminated soil in the immediate vicinity of a transformer pad may not in itself pose direct exposure hazards to workers or even future residents given the assumed exposure area. However, under a future redevelopment scenario, the soil could be excavated and spread out over a much larger area. This could result in a dramatic increase in the average concentration of a contaminant across the DU(s). Decision units for these types of spill areas as well as other examples are described in Subsection 3.5.

Figure 3-12. Designation of DU Layers for Characterization as Part of a Subsurface Investigation


A similar approach should be followed for designation of subsurface DUs (e.g., review of site history and initial field observations). An investigation of the vertical extent of contamination is typically required for releases of liquids or other chemicals that could migrate downwards and contaminate deeper soil or groundwater. A subsurface investigation may also be required to further delineate contamination documented at the surface, determine the depth of buried contamination, or the extent of contaminated fill material.

A small number (e.g., <30) of Exploratory Borings are usually advantageous during the initial stages of an investigation, similar to initial field inspections of surface soils to identify potential spill areas. The resulting information can be used to identify obviously contaminated intervals (e.g., visual observation of petroleum contamination, ash layers, elevated PID readings, etc.) and assist in designation of subsurface, DU Layers for more detailed characterization.

Designated depth intervals should be targeted for characterization, (Figure 3-12). A subsurface investigation of suspected deeper contamination might, for example, include DU layers designated from 0.5'-1.0', 1.0'-2.0'; 2.0-5.0' and 5.0-10' below ground surface. This is necessarily site specific and dependent on the contaminants and objectives of the investigation. Subsurface DU Layers might, for example, be designated in a manner that allows optimal resolution of contaminant distribution and mass for in situ remedial actions. Subsurface DU layers might also be designated in a manner that allows for optimization of ex situ removal or remedial actions, including segregation of contaminated soil that may require expensive treatment or disposal from otherwise clean areas of soil. Refer to the examples provided later in this Section.

Subsurface DU layers are ideally characterized to the same level of detail as carried out for exposed, surface soil DU, with 30-50+ increments collected and combined to form a bulk, MI sample. This will require the installation of borings through the overlying soil to access the targeted layer below and/or the collection of samples from pits or trenches (see Section 5).

As discussed in Section 4, characterization of subsurface DU Layers using Multi Increment sampling approaches is recommended. The collection of discrete samples from targeted depth points in a core (e.g., every five feet) is unreliable for characterization of subsurface soils. This is due to potential, small-scale, random variability of contaminant concentrations in soil at the scale of a discrete sample and/or the subsample mass tested by the laboratory (see also Schumacher and Minnich, 2000, Feenstra, 2003, HDOH, 2014). In Figure 3-12, the section of the core extracted from a DU layer represents an "increment" under a Multi Increment sampling approach (refer to Section 4). Core increments extracted from targeted DU layers are normally subsampled and combined in the field to prepare bulk MI samples in order to reduce the total mass of soil submitted to the laboratory for processing and analysis.

The thickness of subsurface DUs could vary between boring locations. For example, it may be desirable to determine the mean concentration of lead in an irregular, subsurface layer of fill material or debris across a site. In one area the layer might be a few inches thick, while in other areas the layer might be several feet thick. The mass of soil collected from increments extracted from the layer should be weighted in order to collect a representative sample. This can be done by collecting a fixed mass of soil from a fixed spacing of soil plugs extracted from the core (e.g., five-gram subsamples collected every two inches).

For samples to be analyzed for VOCs, collect regularly spaced (for example every two to six inches) 5 gram plugs from the targeted core interval or DU layer and place them in a sampling container with methanol while in the field. If this option is not feasible (e.g., due to methanol shipping constraints), individual subsample plugs can be collected in small gas-tight coring devices (e.g. Core-N'-One® or Encore®-type containers) and immediately frozen for shipment to the laboratory (see Section 11). The laboratory should be instructed to combine the increments in methanol prior to extraction and analysis.

In some cases it may be desirable or otherwise necessary to use an individual borehole as a DU for decision making. The borehole itself can be a "DU" if this is the scale at which a decision will be made. For example, single boreholes might be used to determine the depth to the bottom of contaminated fill at specific locations within a site in order to assist in remedial plans (Figure 3-13). In other cases, obstructions and other site factors may limit the number of borings that can be installed at a site to estimate the lateral extent of contamination. Investigations of active gas stations with suspected, leaking underground storage tanks (USTs) or small USTs associated with boilers or generators are a common example (Figure 3-13). The installation of a number of borings adequate to collect proper MI samples at such sites (e.g., 30+) is often impractical. The objective of the investigation might instead be to simply determine the presence or absence of contamination at depth at a specific location within the facility. Limited sampling objectives such as these could be carried out with a smaller number of borings, especially for contaminants such as petroleum that are easily identifiable in the field.

Figure 3-13. Example Gas Station and Former Boiler UST Sites Where Single Borehole DUs Might be Used.

These example sites show where single borehole DUs might be used for investigating the lateral and/or vertical extent of contamination. The hypothetical borehole locations are noted by red circles. This is adequate to determine presence or absence of the contaminant(s) only.

Data from a single or otherwise small number of borings will not be adequate to estimate the mean concentration of the contaminant in the area of the boring, but limited objectives for an investigation may be achieved. Narrow fingering of contamination into otherwise clean areas of soil could, however, cause the full extent of contamination to be underestimated. This approach is not reliable for contaminants and/or targeted contaminant layers that are not easily recognizable in the field. The use of Laser Induced Fluorescence or Membrane Interface Probe methods for near-continuous readings of subsurface conditions can assist in providing higher quality data (refer to Section 8 ).

The collection of traditional discrete samples from targeted depths within borings is strongly discouraged, again due to the potential for significant, small-scale variability and resulting "false negatives" and unreliability of data interpolation between sample points. Designation and testing of targeted, DU Layers are especially important if decisions are to be made on data from a single boring. This might include the interval observed or suspected to have the highest probability of being contaminated (e.g., staining, odors, field XRF, debris, etc.). Intervals below and above this zone would be designated as separate, "perimeter" DU Layers to verify anticipated clean soil (see following Section).

For example, Figure 3-14 depicts boring locations and DU Layers for a hypothetical gas station with both lead-contaminated fill material and a leaking underground storage tank. The fill material is marked by pieces of wire, melted glass, burned wood and a distinct lead signature using a field XRF. Soil contaminated by the leaking tank is marked by a distinct, gasoline odor, staining and elevated PID readings.

Figure 3-14. Use of a Single or Small Number of Boring DUs to Estimate the Vertical or Lateral Extent of Contamination at a Specific Location within a Site.

This method can be used with the entire targeted core interval or a representative subsample of the interval submitted to lab for processing and testing. Submitting the entire DU layer to the laboratory for processing and analysis is preferable, since this eliminates potential field error (see also Subsection 3.5.6). This is referred to as "direct inference" (AAFCO, 2015),

Borings are installed to estimate the lateral and vertical extent of contaminated soil associated with the two scenarios. Where encountered in a boring, intervals of fill material and/or gasoline-contaminated soil are designated as separate DU layers for characterization. The core in between these intervals is designated as a "perimeter" (or "confirmation") DU layer and likewise targeted for characterization. Similar DU layers are designated outside of the apparent margins of contamination in order to confirm that the same interval of soil is now clean. For example, this might include testing of the upper two feet in borings to verify the absence of contaminated fill material, the interval of vadose-zone soils previously identified to be contaminated by the tank release and soil at capillary fringe zone at the water table. Including a core interval across the capillary fringe is especially useful, since this has the highest chance of catching the presence of contamination that has reached the water table (see Figure 3-14).

Such core intervals might be considered "discrete" samples in that they are to be independently processed and tested. Reference to the samples as "core intervals" is preferable, however. This will help to avoid confusion with traditional and less reliable "discrete" samples collected from a single depth point within a core.

The installation of a large enough number of borings to estimate the mean concentration of termiticides in soil under an existing building slab likewise may not be practical (e.g., Technical Chlordane or aldrin-dieldrin). In such cases the HEER Office recommends that soil be collected and combined from a minimum of three borings through the slab. The resulting bulk sample should be processed in the same manner as done for a traditional, MI sample and the presence or absence of termiticides noted. The reported concentration of termiticides in the soil should not be relied upon for final decision making purposes.

As discussed in Sections 4 and 5, if the volume of soil in a boring for a targeted interval is small enough (e.g.1-3kgs), then the entire interval should again be submitted to the laboratory for processing and testing in the same manner as carried out for MI samples. If soil is to be tested for volatile organic chemicals (e.g., TPHg and BTEX) then subsampling of the targeted interval in the field and placement in methanol or gas-tight individual containers will be required (see Section 5). Note that plugs of soil removed from the core are not "increments" in the sense of MI samples. They are instead subsamples of core increments. The spacing and number of plugs necessary to collect a representative subsample of the core increment will vary based on anticipated, small-scale heterogeneity of the target contaminant in the core as well as the total, desired mass of soil to be collected from the targeted DU layer as a whole. This is discussed in more detail in Section 5.

The use of limited borings to estimate the lateral and vertical extent of subsurface contamination can be very useful and cost-effective for design of initial remedial actions. A more detailed characterization will normally be required for final decision making purposes (unless preliminary data will already be used to assume a potential environmental hazard exists) or to confirm the effectiveness of cleanup actions.


"Perimeter DUs" are established immediately outside an area of suspected heavy contamination in order to confirm the lateral extent of contamination. Perimeter DUs can also be used to delineate the extent of contamination adjacent to any exposure area DU that has been found to exceed the applicable HDOH EALs. Soil contamination is typically delineated out to levels that fall below HDOH EALs, even if this involves crossing property borders (assuming property access is granted by the neighboring property owner). The number and design of perimeter DUs is necessarily site-specific and based in part on the confidence that the DUs will be placed in areas that are unlikely to be contaminated. For example, avoid letting a small area of contamination cause a much larger perimeter DU to fail action levels and require additional investigation.

Figure 3-15. Primary (DU-1 and DU-2) and Perimeter DUs (DU-3 and DU-4) Designated Around a Building with Suspected Lead and Termiticide Contamination Perimeter DUs should also be designated and used to collect confirmation samples adjacent to and below areas of excavated soil (Figure 3-16; after ITRC, 2012).

Figure 3-16. Perimeter DUs Designated Around a Planned Soil Excavation Area Samples to confirm clean boundaries ideally collected prior to soil removal to avoid need for remobilization (after ITRC, 2012)

In Figure 3-15, DU-1 and DU-2 represent hypothetical spill area DUs designated around the perimeter of a building to test for the presence of lead and termiticides in soil. These DUs are bordered by the perimeter DUs 3 and 4, which are anticipated to be outside of impacts above EALs.

Note that samples can be collected either before or after actual excavation (Figure 3-16). The confirmation of clean perimeter DUs prior to excavation is optimal, since this will negate the need for sample collection afterward and minimize concerns about the potential need for repeated over excavation. No further sampling is required if data for perimeter DUs indicate impacts below applicable action levels and soil within the DUs is completely removed.


Decision unit designation and sampling strategies specific to stockpiles is provided in the HEER Office document Guidance for the Evaluation of Imported and Exported Fill Material (HDOH, 2017d; included as Appendix 3-A; see also Subsection and Subsection 5.5). Background information on a stockpile should be compiled to the extent possible and used to support the sampling approach proposed in the SAP. Fill material that is imported to or exported from sites where significant environmental contamination has been identified or where cleanup projects are underway could pose multiple environmental hazards if not appropriately characterized and managed.

Special considerations for selecting DUs for sampling soil stockpiles include (see HDOH, 2017d):

Table 3-1. Default DU Volumes for Stockpiles
Receiving Site Land Use Default DU Volume Comments
Unrestricted Use (includes single-family homes) 100 yd3 Assumes 5,000 ft2 reuse exposure area and six-inch placement thickness.
Schools and High‐Density Residential Developments 400 yd3 Assumes 0.5-acre exposure area and six-inch placement thickness.
Commercial or Industrial use only (formerly developed fill source) 400 yd3 Assumes 0.5-acre exposure area and six-inch placement thickness.
Commercial or Industrial use only (agricultural field fill source) 400 yd3 or 18 DUs Stockpile divided into minimum 18 DUs for characterization if >7,200 yd3.

Notes (Table 3-1):

  1. See guidance in Appendix 3-A for additional options and recommendations (HDOH, 2011e).
  • The source of the soil in the stockpile;
  • How the stockpile was created (over time, if applicable);
  • How best to access the pile for sampling, especially if it is large and or unstable;
  • Target contaminants.

Large stockpiles could be broken or segregated into separate DUs for characterization. Default DU volumes for testing of stockpiles (or sources of fill material) are summarized in Table 3-1. Division of the stockpile should be based on soil type, source, potential for contamination, potential environmental concerns based on targeted contaminants (e.g., direct exposure vs leaching) and volume limits based on proposed reuse.

Larger DU volumes are appropriate for testing of stockpiles for reuse at high-density residential developments or schools. For example, a 400cy DU represents the approximate volume of soil necessary to cover the default, commercial/industrial exposure area of 20,000 ft2 to a depth of six inches. A similar DU volume serves as a good starting point for testing of stockpiled soil to be used for fill material at a commercial or industrial site.

Assumed exposure areas of one-acre or more and thicknesses of one foot or more might also be appropriate for beach replenishment using dredged material not otherwise suspected to contain significantly elevated levels of contaminants. This would yield DU volumes of several thousand cubic yards. Testing of larger volumes can also be appropriate for screening of large stockpiles that are not anticipated to include contaminated soil but where some level of verification data is desired (e.g., soil from previously tested agricultural fields).

Larger DU volumes might also be acceptable for general screening purposes if other lines of evidence support a low risk of contamination based on the known source of the soil. Smaller DU volumes (e.g., 20 yd³; see HDOH, 2017d) are recommended for soil that might contain pockets of highly leaching contaminants (e.g., triazine herbicides) or contaminants that might pose potential gross contamination concerns (e.g., petroleum). Such high-resolution testing of stockpiled soil for potential reuse is likely to be cost-prohibitive, however.

Disposal of the soil with suspected pockets of highly leachable contaminants or gross contamination at a municipal landfill or long-term management at the source under an EHMP might be the most prudent option. Fill material characterization, sources of fill that should be considered suspect for contamination and other considerations are described in further detail in Guidance for the Evaluation of Imported and Exported Fill Material, Including Contaminant Characterization of Stockpiles (Appendix 3-A). Note that a hazardous waste determination is required for disposal of soil at a landfill (see Subsection 5.10). This could require additional sampling and analysis. Contact the landfill for specific information. Consider the inclusion of any additional testing methods and requirements in the original SAP in order address disposal needs prior to the initiation of site activities.


Detailed guidance on the designation of decision units for sediment investigations is forthcoming. Sediment is defined as unconsolidated material that is currently under water (e.g., harbor bottom sediment) or otherwise associated with deposition in an aquatic environment (e.g., tidal flats, ephemeral stream beds, flood plains, etc.). Factors that affect the lateral and vertical distribution of contaminants in sediments include:

  • The nature of the source area (e.g., periodic vs continuous release),
  • Proximity to contaminant source,
  • Sediment geochemistry,
  • Size distribution of sediment particles,
  • Water flow rate and volume,
  • Location of depositional areas,
  • Local features (natural or artificial),
  • Resuspension and deposition during subsequent flood events or other disturbances,
  • Vertical stratification over time,
  • Seasonal fluctuations, and
  • Bioturbation effects

Stratification can affect the nature of contamination at the water-sediment interface, necessitating seasonal sampling and leading to significant and abrupt vertical variability in contaminant concentrations over short distances.

These factors necessitate careful planning and thought in designation of DUs for characterization. Initial DU designation should focus on migration pathways from suspected source areas to depositional areas. Decision unit sizes and boundaries should be adjusted to address ecological impact concerns and optimization of potential remedial actions as appropriate for the investigation. Vertical designation of DUs layers might focus on the uppermost, active benthic zone, past layers of suspected high contamination, or depth intervals targeted for future dredging.

Samples will typically need to be collected from the biotic zone (e.g., the 0 to 4-inch interval) at the sediment-water interface. Deeper samples might be necessary to delineate the vertical extent of contamination to address other site-specific receptors, or characterize layers of contaminated sediment deposited in the past. If a contamination release is not recent, it is possible that later events could have resulted in deposition of a new, possibly much less contaminated sediment layer above the sediment of concern. Surface samples could result in an incomplete characterization of the DU area. Contaminated sediments currently isolated at depth could pose a threat due to storms or other erosional events or disturbance by future dredging. Conversely, clean sediment of sufficient thickness in an overall constant depositional environment may act as a cap over underlying contamination and indicate that the exposure pathway is incomplete.

Although detailed data are less available, small-scale, random variability of contaminant distribution in sediment in a lateral direction may be less dramatic than for soil in some depositional areas due to mixing during sediment transport and deposition (refer to discussion in Section 4). This could include depositional areas from runoff of agricultural, industrial or urban areas, or aquatic areas impacted by long-term discharges of industrial waste water. If this is the case, replicate field samples could document adequate sample precision with less individual increments per DU than typically utilized for soil sampling sites.

Potential distinct boundaries between clean and contaminated sediment may be anticipated with depth in areas near past contaminant source locations. This requires close consideration of the vertical resolution of a sediment investigation in terms of targeted, DU layers.


The guidance in this Subsection applies to characterization of soil in large open areas, where based on a thorough Phase I Environmental Site Assessment, localized spill areas are not anticipated, or have been identified and will be separately characterized. Examples include former agricultural fields, golf courses, and munitions ranges. The guidance does not directly apply to previously developed commercial, industrial, or residential areas. Characterization of these areas, as well as known or suspect spill areas within the types large areas noted above, should follow DU designation guidance presented in Subsection 3.4.

Optimal methods of employing a DU-MIS strategy for very large areas are a subject of on-going research by various organizations, and refinements of the recommendations provided below will be made in the future, as may be appropriate. CHARACTERIZATION FOR BASELINE INVESTIGATIONS

Baseline investigations of large areas, including agricultural fields, golf courses, and former munitions ranges might be necessary as part of a property transaction, potential redevelopment project or an evaluation of area-wide contamination. These areas can be up to hundreds or thousands of acres with dramatic changes in terrain and historical use. While still challenging logistically in the field, adhering to the same Systematic Planning process described earlier (see Subsection 3.2) will help ensure that investigation objectives are clearly defined and that meaningful data are collected.

The objective of a baseline investigation is to establish environmental impacts under current site conditions. This might include estimation of the mean concentration of residual pesticides in soil at a former golf course or in a large field that was formerly used to produce sugarcane. Other examples include assessment of soil impacts from dispersal of munitions-related contaminants across a former military target range.

The first step in this process should include completion of a thorough Phase I Environmental Site Assessment (Phase 1 ESA) covering the site history and current operations (refer to Subsection 3.1.1). The compilation and review of existing information for a former agricultural site might, for example, include the following:

  • Crop history;
  • Current and past pesticide use;
  • Historic aerial photographs;
  • Historic Sanborn Fire Insurance maps, topographical maps, or other maps;
  • Interviews with former employees;
  • Existing soil investigation reports (including investigations of adjacent or nearby fields using lot-size DUs);
  • Review of other published, historic information (journals, etc.); and
  • Field inspection (current operations, former buildings, suspect dump areas, etc.).

Decision units should be designated to characterize the site. Areas of suspected or known heavy contamination should be identified separately from anticipated less impacted areas (see Subsection 3.4.3). For example, these might include former pesticide mixing areas, storage areas, rail lines, and plantation camps for agricultural sites, and burn pits or areas immediately surrounding low-order detonations at a firing range impact zone. Designate DUs for non-suspect areas based on more localized factors including crop history, soil type, drainage patterns, etc., appropriate for the site investigation objectives and based on the information gained in the Phase I ESA. In the case of golf courses, older sections of the course where different types of pesticides were used might be characterized separately from more recent additions, and fairways might be separated from putting greens.

The total number and size of DUs will necessarily be site-specific. Use of even a single DU might be adequate to accomplish investigation objectives, although this is less likely as the size of the site increases. Past experience has suggested that 10-15 DUs are typically adequate for baseline characterization of very large areas, based on past use history, soil types, drainage patterns, etc. A better resolution of mean contaminant concentration within a single area might also be desirable, for example to verify historic information gathered on past use of the site. A Multi Increment sample should be collected to characterize each DU, with triplicate samples collected from a minimum 10% of DUs (see Section 4). See Section 9 for COPCs for pesticides. Refer to the HDOH Environmental Action Level guidance for a list of potential munitions-related COPCs (HDOH 2016).

Note that the baseline investigation approach also requires a thorough walkthrough of the entire site. This walkthrough will assist in identifying areas suspected of elevated contamination, including previously unknown dumping sites, waste pits, former plantation camp areas, pesticide mixing or storage areas, etc. that might otherwise be missed.

The results of the baseline investigation can be used to estimate the general extent and magnitude of soil contamination within the targeted area at the scale of the DUs designated. This information, in conjunction with a thorough, Phase I ESA review of the site might be adequate for the entity overseeing the investigation and their consultant to draw initial conclusions regarding redevelopment of the site for residential or other purposes. As discussed below, however, a more detailed investigation might be required for a formal site closure determination by HDOH, depending in part on the anticipated land use. For example, a baseline assessment including a strong Phase 1 report might be adequate for concurrence with future commercial/industrial use of the area, while more detailed higher-resolution data will normally be required for residential development. CHARACTERIZATION of very large areas FOR REDEVELOPMENT

Table 3-2. Recommendations for Investigation of Large Areas
Project Classification Area Recommendations
Category 1 <59 Acres
  • Phase 1 ESA
  • One DU per acre
Category 2 >59 to <118 Acres
  • Phase 1 ESA
  • 59 randomly located, one-acre DUs
Category 3 >118 to <590 Acres
  • Phase 1ESA
  • Baseline Investigation
  • 59 randomly located, one-acre DUs
Category 4 >590 Acres
  • Phase 1 ESA
  • Baseline Investigation
  • 90 randomly located, one-acre DUs

Recommendations provided below apply to both residential and commercial/industrial redevelopment. Characterization of very large areas for redevelopment can be challenging. Such projects can cover hundreds or thousands of acres and include hundreds or thousands of individual residential lots. The primary environmental hazard is direct exposure of future residents and workers to residual pesticides, or other contaminants such as metals in the soil. Localized contamination of highly mobile chemicals (e.g. explosives residues) can also pose potential leaching threats to groundwater that might be used to serve the redevelopment in the future.

A default Exposure Area DU size of one-acre is considered acceptable for characterization of large areas where no localized areas of potentially heavy contamination are identified as part of a thorough Phase I ESA (i.e., suspect Spill Areas; refer to Subsection 3.4.3). Variability of mean contaminant concentrations within this default DU size (i.e. at the scale of potentially smaller exposure areas) is assumed to be relatively low based on investigations of former golf courses and agricultural field areas where detailed data has been collected (e.g., refer to studies referenced in Section 9 appendices). Restriction of the default exposure area size to one-acre also helps to ensure that unanticipated, small but heavily contaminated spill areas are captured by DU data (e.g., a former pesticide mixing area). Note that if a thorough Phase 1 ESA has not been completed, the default one-acre Exposure Area DU size may be judged inadequate for evaluation (this applies to all categories of large area sites discussed below).

Table 3-2 summarizes the recommended strategy for characterization of large parcels of land where localized spill areas are not known or anticipated. Division of the site into adjacent, one-acre DUs is recommended for areas 59-acres or less in size (Category 1, <59 acres). Designation of DUs should reflect information garnered during the Phase I ESA to the extent practical (e.g., land-use history, terrain, soil type, etc.).

Random placement of 59, one-acre DUs is recommended for moderately large sites where the DUs will cover at least 50% of the total area (Category 2, <118 Acres). Testing of 59 of the total number of potential, one-acre DUs within the project area allows 95% confidence that the mean contaminant concentration in 95% of one-acre DUs at the entire site will be lower than the highest concentration reported in the one-acre DUs that were tested (USEPA, 1989b). DUs should be placed in a systematic random distribution, and with consideration to adequately represent variability associated with land-use history, pesticide use, soil type, topography and other key factors gained from the Phase I ESA investigation. Note that the 95% confidence criteria will not be met if the highest mean concentration of just one of the 59 decision units exceeds the applicable target action level. Additional sampling would typically be required to adequately identify and address areas of the site with elevated contamination. Consultation with the HEER Office to discuss potential options is recommended.

Inclusion of baseline investigation data as described above is recommended for sites where 59, one-acre DUs will cover less than 50% but at least 10% of the total project area (Category 3, >118 to <590 Acres). The baseline investigation should be conducted first, and will help to identify large-scale variance within the subject site and assist in subsequent DU placement and decision making. DUs should again be placed in a systematic random distribution, and with consideration to adequately represent variability associated with land-use history, pesticide use, soil type, topography and other key factors gained from the Phase I ESA investigation and the baseline investigation. For example, the baseline study might identify somewhat higher but still potentially acceptable levels of arsenic contamination in a portion of a field that was already under sugarcane production in the 1920s and 1930s (see pesticide discussion Section 9). Placement of one-acre DUs within this area or even separate characterization of this area would be warranted. Including a baseline investigation also provides some level of data for the entire project area (in addition to the 59 one-acre DUs), and helps address concerns of prospective residents who understandably might ask about soil testing data for their area.

Confidence in the representativeness of data decreases as the total area encompassed by the one-acre DUs decreases. An increase in the number of one-acre DUs to 90 in addition to a baseline assessment is recommended for projects where less than 10% of the land will be covered by the DUs (Category 4; >590 acres). This provides a 99% confidence that the mean contaminant concentration in 95% of one-acre DUs at the entire site will be lower than the highest concentration reported in the one-acre DUs that were tested (USEPA, 1989b).

The configuration of DUs across very large areas with respect to the planned redevelopment might also be desirable, although this could complicate usage of the data should redevelopment plans change in the future (see example in Subsection 3.5.6). HDOH feels that these recommendations are manageable in terms of the overall cost of large-scale, redevelopment projects. Alternative approaches should be discussed with HDOH on a case-by-case basis.

Decision Units should be placed in a systematic random fashion, and with consideration to adequately represent variations in site characteristics (e.g., land-use history, terrain, soil type, etc.).


As discussed in Section 5, Multi Increment samples are considered to be more reliable for characterization of the mean concentration of a contaminant within a targeted DU. The evaluation of data collected from DUs is discussed in Section 13 of this guidance as well as the accompanying guidance Evaluation of Environmental Hazards at Sites with Contaminated Soil and Groundwater (HDOH, 2016). When using a Decision Unit strategy, the entire area of a Decision Unit is acted upon as a single entity based on the average contaminant data collected from that Decision Unit. If the decision outcome is "contaminated," then the entire area of the DU is treated as being contaminated. If the data indicate that remediation is required, this applies to the entire Decision Unit. If the outcome is "not contaminated," then the entire area of the DU is treated as being not contaminated. As discussed above, this makes the designation of Decision Units very important to ensure that appropriate exposure areas and/or spill areas are identified, and areas of obvious heavy contamination are segregated into separate DUs to reduce the volume of soil that is identified as "contaminated" and requires treatment.