Attachment_J-8_-_4_of_4_NUREG-1757_Vol_2.pdf

APPENDIX I

I-27 NUREG-1757, Vol. 2, Rev. 1 the residual radioactivity is not in the surficial soil, the original area of residual radioactivity may not be as important in scenario development, because some of the primary transport mechanisms result in redistribution of the radionuclides over larger areas (i.e., ground water used as irrigation).

One common mistake in licensee submittals is that area factors are typically not provided for residual radioactivity on building surfaces. The primary reason for this is that such factors could not be calculated by using the DandD, Version 1. Therefore, when the screening DCGLW values were published in the Federal Register (see Appendix H), which were derived from an improved DandD, Version 1, the associated area factors were not published. An alternative approach should be used to calculate area factors for residual radioactivity on building surfaces.

One approach is to use DandD, Version 2.1, to calculate the area factors, although it models area factors conservatively. Another approach that has been successfully used is to develop the area factors by using the RESRAD–BUILD computer code and adjusting these derived area factors to account for the fact that RESRAD–BUILD typically gives less conservative dose estimates.

With this approach, the screening DCGL values are converted into the appropriate concentration unit for RESRAD–BUILD [i.e., from (dpm per 100 cm2) to (pCi/m2)]. Area factors calculated by RESRAD–BUILD can then be adjusted by the ratio of the dose from RESRAD–BUILD to

0.25 mSv/y (25 mrem/y) (i.e., the equivalent dose from DandD).

I.3.3.3.6 Offsite Scenarios

In rare situations, the scenario resulting in the highest exposures from the residual radioactivity will be offsite use scenarios. For these evaluations, the dose limit remains that of 10 CFR 20, Subpart E, even though the situation may seem similar to the clearance of materials prior to license termination (see also Section G.1.1 of this volume). In these scenarios, the exposure to the radioactive material will occur because it has been removed from the current location, and this results in either new or enhanced exposure pathways. For example, a site has poor ground water characteristics (thereby, allowing the licensee to remove the ground water pathway from any applicable scenarios) and the reasonably foreseeable land use is either commercial or industrial. The primary contaminant is Tc-99, which primarily results in dose through either the ground water or vegetable pathways, both of which are not applicable to the physical characteristics of the site or land use assumptions. The residual radioactivity is present in the site’s top soil. A possible offsite scenario is that during construction of any commercial interest on the site after license termination, the removed topsoil is sold for use in a residential setting.

In this case, it is likely that the topsoil with residual radioactivity will be unintentionally mixed with other topsoil at the offsite location. Licensees can use generic analyses to screen the importance of offsite uses with such sources as NUREG-1640, “Radiological Assessments for Clearance of Materials from Nuclear Facilities.” (NRC 2003)

Even if offsite use is not considered reasonably foreseeable, offsite scenarios may be less likely but plausible scenarios and should be analyzed as scenarios, to understand the robustness of the analysis.

NUREG-1757, Vol. 2, Rev. 1 I-28

I.3.3.3.7 Determining the Compliance Scenario

In many situations a licensee will be faced with selecting a compliance scenario from potentially a large suite of scenarios and exposure groups. The licensee is expected to base their demonstration of compliance on the exposure to the highest group, consistent with the definition of the critical group. Licensees may find it advantageous to use an iterative approach to screen all the potential scenarios. This will allow the licensees to focus their more detailed analyses on the important scenarios. Licensees may be able to use information from NUREG/CR-5512, NUREG-1640, and NUREG-1717, as well as other licensees’ analyses to screen their potential scenarios with quantitative methods. Licensees also may be able to provide qualitative arguments to demonstrate that the dose from certain scenarios are bounded by the dose of higher level scenarios (e.g., a residential gardening scenario will bound the dose for the residential non-gardening scenario). The licensees should provide justifications on the basis, method, and results of their scenario screening in their DP.

Even after screening the scenarios, a licensee will likely be left with a few scenarios that may require detailed analyses to determine which will result in the critical group. For licensees with multiple radionuclides, commonly, determining the compliance scenario depends on the final mixture of radionuclides. This can provide a dilemma for licensees creating DCGLs. The licensee must show that the final concentrations at the site meet the dose criteria of 10 CFR Part 20, Subpart E. Three possible approaches that the licensee may use to show compliance are, but are not limited to the following:

1. Use the most limiting DCGL for each radionuclide, regardless of the scenario, and use the sum of fractions, ignoring the scenario basis for each DCGL. This approach requires limited justification. It will always either estimate the same dose as the individual scenarios or overestimate the dose. Generally, it will greatly overestimate the dose for the individual scenarios.

2. Use a surrogate approach to limit the number of radionuclides of importance. A surrogate approach relies on different radionuclides having relatively fixed ratios. For example, assume that at a site with cesium-137 and strontium-90 can show that for every 37 Bq/kg (1 pCi/g) of Cs-137, there will be a 18 Bq/kg (0.5 pCi/g) of Sr-90. By using this relationship, an effective DCGL for combined Cs-137 and Sr-90 can be created. The licensee may be able to reduce the number of critical scenarios, specifically those driven by exposure to Sr-90.

This approach requires that the licensee have the necessary information on relative ratios of the radionuclides.

3. Commit to demonstrating the final dose for each of the important scenarios in the final status survey reports. This approach will require the licensee to establish operational DCGLs to fully utilize MARSSIM (see Section 2.5).

I-29 NUREG-1757, Vol. 2, Rev. 1

The licensee needs to provide either quantitative analysis of or a qualitative argument discounting the need to analyze all the scenarios generated from the less likely but plausible land uses. The results of these analyses will be used by the staff to evaluate the degree of sensitivity of dose to overall scenario assumptions (and the associated parameter assumptions). Analyses of less likely but plausible scenarios are not meant to be ‘worst-case’ analyses and should not utilize a set of ‘worst-case’ parameters. Selection of parameters for less likely but plausible scenarios should be consistent with the guidance in this Appendix. The reviewer will consider both the magnitude and time of the peak dose from these scenarios. If peak dose from the less likely but plausible land use scenarios is significant, the licensee would need to provide greater assurance that the scenario is unlikely to occur, especially during the period of peak dose. The licensee may be able to show that the compliance scenario bounds the results of all or many of the scenarios associated with the less likely but plausible land uses.

I.3.4 Generic Examples

The following examples are provided as situations where the default pathways may be removed or modified. Note, the examples assume that an adequate level of justification has been provided by the licensee.

I.3.4.1 Removal of Ground Water Pathways

A licensee has extensive contamination of the upper soil horizon and the upper aquifer, which is unconsolidated and the licensee wishes to remove the ground water pathway because the upper aquifer would not be used as a water source. The aquifer shows relatively high levels of microbial activity, turbidity, and nitrates. In addition, adjacent to the site is a small patch of wetlands that shows a great deal of communication with the upper aquifer. The potential yield rate of the upper aquifer is sufficient for domestic use, but there is a better-quality, confined aquifer, whose horizon is at a depth of approximately 30 meters (100 feet). The licensee has also demonstrated that the deeper aquifer will not become contaminated from the upper aquifer.

Considering all of these reasons in combination, it is questionable whether the upper aquifer would actually be used. Although it may be possible for someone to treat the contaminants and use the aquifer, there are better sources of water easily available. After consultation with the EPA and the State, it is agreed that it would be unreasonable to assume someone would use the upper aquifer as a water source. Therefore, the licensee is allowed to remove the ground water pathway from the scenario.

I.3.4.2 Scenario Development for Buried Residual Radioactivity

I.3.4.2.1 Example 1: Subsurface Soil

A site has residual radioactivity buried at a few feet below the surface and the licensee is requesting unrestricted release. The residual radioactivity does not have enough highly energetic gamma-emitters to result in an external dose in the current configuration. Two exposure

NUREG-1757, Vol. 2, Rev. 1 I-30 scenarios can be developed (without any other site-specific information): (1) leaching of the radionuclides to the ground water, which is then used by a residential farmer; and (2) inadvertent intrusion into the buried residual radioactivity by house construction for a resident farmer with the displaced soil, which includes part of the residual radioactivity, spread across the surface.

Exposure scenario 2 encompasses all the exposure pathways and, although not all of the source term is in the original position, leaching may occur both from the remaining buried residual radioactivity and the surface soil. Except for cases where an additional 0.6 m (2 ft) of unsaturated zone may make a tremendous difference in travel time to the aquifer, the ground water concentrations should be similar and, therefore, analysis of the second exposure scenario appears to be the appropriate scenario for the critical group exposure. This example is described in greater detail and integrated with the other guidance in Appendix J of this volume.

I.3.4.2.2 Example 2: Embedded Piping

At another site, the licensee is requesting unrestricted release of its site. It is removing the buildings, but is evaluating the need to remove the concrete pads, which have embedded piping that contains the residual radioactivity. Two scenarios can be reasonably envisioned. The first scenario involves a resident farmer onsite. The farmer builds a house on the concrete pad, without disturbing the embedded piping. Possible exposure pathways would be external dose from the piping and exposure to leached materials from the piping through ground water use (e.g., drinking, irrigation, etc.). The second scenario is similar to the building renovation scenario, where the concrete pad and piping are removed from the site. The licensee should investigate both to find the limiting scenario.

I.3.4.3 Scenario Development for Restricted Release

The site restrictions planned for an alternate site include a restriction, for this example, on the deed, on the use of the property for only parkland, and an engineered cover is placed over the residual radioactivity. The engineered cover is contoured for use as parkland and has a vegetative cover (i.e., not a mound covered in rip-rap). Three scenarios are easily envisioned for the restricted release analysis. The first is recreational use of the property as a city park or golf course, which would limit exposure scenarios to possible external exposure. The second would involve offsite use of ground water that contains radionuclides leached from the buried residual radioactivity. The default offsite user would be a resident farmer using the ground water for all water needs. The third scenario would be a worker maintaining the park.

The doses assuming the loss of the institutional control (i.e., the deed restriction) and degradation of the engineered cover also must be evaluated. Again, two main scenarios can be envisioned.

The first scenario is similar to the default exposure and would involve a residential farmer that uses ground water from the aquifer under the site. The engineered cover may have been compromised by the placement of the buildings, but the cover may still work in some degraded

I-31 NUREG-1757, Vol. 2, Rev. 1 function (e.g., the water infiltration rate would increase from the design rate to some higher rate, but probably not as high as the infiltration rate would have been if the cover had never been constructed). Whether buried residual radioactivity had been transported to the surface by the construction of a basement under the resident farmer’s house would depend on the thickness of the engineered cover. If typical basement depth were deeper than the engineered cover’s thickness, some portion of residual radioactivity would be transported to the surface, mixed with the “clean” cover material, and spread over the site.

The second scenario would involve possible erosion of the cover and subsequent exposure of an onsite resident to the buried radionuclides or radionuclides redistributed by surface water. The exposure scenario would still be a resident farmer. The reasonableness of this scenario would depend on the thickness and erosion-resistence of the engineered cover.

I.4 Criteria to Establish Conceptual Models

I.4.1 Introduction

Analyzing the release and migration of radionuclides through the natural environment and/or engineered systems, at a specific site, requires the licensee to interpret the nature and features of the site so that the site can be represented by mathematical equations (i.e., mathematical models).

This simplified representation of the site, including the associated mathematical models, is commonly referred to as the conceptual model of the site.

Figure I.1 depicts the process of conceptual model development. In dose assessments, developing a conceptual model involves making an abstraction of site data into a form that is capable of being modeled. This development should generally involve making simplifying assumptions, including simplification of the appropriate governing equations, to reflect the physical setting. These simplifying assumptions are usually made in describing the geometry of the system, spatial and temporal variability of parameters, isotropy of the system, and the influence of the surrounding. The conceptual model should provide an illustration, or description, of site conditions, which shows, or explains, contaminant distributions, release mechanisms, exposure pathways and migration routes, and potential receptors. In other words, the conceptual model should explain or illustrate how radionuclides enter, move through, and/or are retained in, and leave, the environment.

NUREG-1757, Vol. 2, Rev. 1 I-32 decom-023.ppt 031102

Conceptual Model

Mental Construct

Site Data Scenario

Mathematical Models

Engineering Designs

Computer code

Figure I.1 Conceptual Model Development.

As shown in Figure I.2, developing a conceptual model at a site is Step 3 of the Decommissioning Decision Framework (see Section 2.6 of this volume). Conceptual model development follows after assimilation of site data (Step 1) and definition of scenarios (Step 2), because information from these two steps feeds into its development. In other words, the conceptual model should be based on what is known about the site from data and information gathered as part of Step 1, and how the site evolves during the period covered by the analysis based on the assumed land-use defined under Step 2.

Mathematical models are a quantitative representation of the conceptual model. Because the conceptual model provides the linkage between site conditions and features (Steps 1 and 2) and the computer code(s) (with its associated mathematical models) used in the dose analysis (Step 4 of the Decommissioning Framework), it is a key step in a dose assessment and should not be taken lightly.

I-33 NUREG-1757, Vol. 2, Rev. 1

Dose Assessment4

Scenario Definition/

Pathway Identification

Assimilation of Existing Data and Information

Assimilation of Existing Data and Information

System

Conceptualization

Dose Assessment4 Dose AssessmentDose Assessment44

Scenario Definition/

Pathway Identification22 Scenario Definition/

Pathway Identification Scenario Definition/

Pathway Identification

Assimilation of Existing Data and Information

Assimilation of Existing Data and Information

Assimilation of Existing Data and Information

Assimilation of Existing Data and Information

Assimilation of Existing Data and Information

System

Conceptualization33 System

Conceptualization decom-024.ppt 031102

Figure I.2 Decommissioning Decision Framework.

I.4.2 Issues

Uncertainties in conceptual models can be large, and possibly even larger than uncertainties in parameters used in the analysis (James and Oldenburg 1997). Thus, conceptual model uncertainties can be a significant source of uncertainty in the overall dose assessment.

Uncertainties in the conceptual model(s) are generally caused by incomplete knowledge about the natural system being analyzed and differing views about how to interpret data representing the system.

Development of conceptual models is a subjective process based on interpretation of limited (or in most cases, sparse) site data. From these limited data, the licensee should determine the key processes and features at the site and how they are likely to affect the movement of radionuclides through the environment. Because the conceptual of the site is based on incomplete information, it is possible that multiple interpretations of the same data can be derived. A licensee should also determine the appropriate level of simplification acceptable for representing the site. An overly simplified conceptual model may leave out key site features or conditions that are important in estimating where radionuclides are likely to be transported (thus, where people might be

NUREG-1757, Vol. 2, Rev. 1 I-34 exposed) and when they might get there (thus, the radionuclide concentration when it arrives).

On the other hand, an overly complex conceptual model may introduce unnecessary uncertainty and costs into the analyses. As a broad example, simple models contained in screening codes may oversimplify features and processes at a specific site. The licensee also should ensure that the appropriate level of detail is provided in the conceptual model. It is important that the conceptual model have sufficient detail and scope for a license reviewer to be able to assess the appropriateness of the computer codes used in the analysis and the defensibility of the assumptions made. In summary, key issues in developing and presenting the conceptual model are: (a) identifying the important site features and processes that need to be included in the conceptual model; (b) deciding among possible competing interpretations of the site data; and

(c) determining the level of detail needed to describe those features and processes.

I.4.3 Recommended Approach

I.4.3.1 Screening

An acceptable dose assessment analysis need not incorporate all the physical, chemical, and biological processes at the site. The scope of the analysis, and accordingly the level of sophistication of the conceptual model, should be based on the overall objective of the analysis.

A performance assessment conceptual model can be simple if it still provides satisfactory confidence in site performance. For an initial screening analysis, little may be known about the site from which to develop a conceptual model. Computer codes used for screening analyses are generally intended to provide a generic and conservative representation of processes and conditions expected for a wide array of sites. Accordingly, the generic conceptual model in such codes may not provide a close representation of conditions and processes at a specific site. Such a generic representation is still acceptable as long as it provides a conservative assessment of the performance of the site.

The DandD code has two default land-use scenarios; a building occupancy and a resident farmer scenario. The building occupancy scenario is intended to account for exposure to both fixed and removable residual radioactivity within a building. Exposure pathways included in the building occupancy scenario include external exposure to penetrating radiation, inhalation of resuspended surface residual radioactivity, and inadvertent ingestion of surface residual radioactivity. The resident farmer scenario is intended to account for exposure to residual radioactivity in soil.

Exposure pathways included in the resident farmer scenario include: external exposure to penetrating radiation; inhalation exposure to resuspended soil; ingestion of soil; and ingestion of contaminated drinking water, plant products, animal products, and fish. The predefined conceptual models within DandD are geared at assessing releases of radioactivity, transport to, and exposure along, these pathways.

For the building occupancy scenario, DandD models external exposure to penetrating radiation as an infinite area source, using surface source dose rate factors from Federal Guidance Report No. 12 (EPA, 1993). Exposure to inhalation of resuspended surface residual radioactivity is

I-35 NUREG-1757, Vol. 2, Rev. 1 modeled as a linear static relationship between surface residual radioactivity and airborne concentrations. The model accounts for ingrowth and decay. Exposure to incidental ingestion of surface residual radioactivity is modeled with a constant transfer rate.

The generic conceptual models for the resident farmer scenario are more complicated because of the large number of exposure pathways and considerations of release of radioactivity from the source area and transport of radionuclides in the environment. DandD models external exposure from volume soil sources when the person is outside as an infinite slab of residual radioactivity 15 cm (6 in) thick, using dose rate factors from Federal Guidance Report No. 12, for volume residual radioactivity. When the person is indoors, exposure from external radiation is modeled in a similar manner, except the exposure is assumed to be attenuated through the use of a shielding factor (note: the higher the shielding factor, the lower the assumed attenuation).

Exposure through ingestion of contaminated animal and plant products is modeled simply through the use of transfer factors. Instantaneous equilibrium is assumed to occur between radionuclide concentration in the soil and the concentration in plants, and between animal feed and animal products.

The generic source-term conceptual model in DandD assumes a constant release rate of radionuclides into the water and air pathways. Release of radionuclides by water is assumed to be downward and a function of a constant infiltration rate, constant contaminant zone thickness, constant moisture content, and equilibrium adsorption. DandD assumes that there are no radioactive gas or vapor releases. Release of radioactive particulates is assumed to be upward, instantaneous, uniform, and a function of a constant particulate concentration in the air and the radioactivity within the soil. Radionuclides in the contaminant zone are assumed to be uniformly distributed in a single soil layer, 15 cm (6 in) thick. No transport is assumed to occur within the source zone, but radioactive decay is taken into account. In terms of containment, DandD assumes that there are no containers (or that they have failed), and that there is no cover over the contaminated zone.

The DandD generic conceptual model for the ground water pathway assumes a single hydrostratigraphic layer for each of the unsaturated and saturated zones. The unsaturated zone (vadose zone) can be broken into multiple layers within DandD; however, each layer is assumed to have the same properties. For radionuclides entering the vadose zone, DandD accounts for adsorption-limited leaching by considering the vadose zone to behave as a well-mixed chemical reactor with a constant water inlet and outlet rate set at the infiltration rate. Accordingly, it is assumed that the vertical saturated hydraulic conductivity of the unsaturated zone is greater than or equal to the infiltration rate (i.e., there is neither ponding nor runoff on the surface). The outlet concentration from one unsaturated zone layer to another is assumed to be a function of the constant infiltration rate, equilibrium partitioning, the thickness of the layer, a constant moisture content, and radioactive decay. Radionuclides entering the saturated zone are assumed to be instantaneously and uniformly distributed over a constant volume of water equivalent to the larger of either the volume of infiltrating water (i.e., the infiltration rate times the contaminated area) or the sum of the water assumed to be removed for domestic use and irrigation. Based on the default parameters in DandD, dilution in the ground water pathway is based on the water use.

NUREG-1757, Vol. 2, Rev. 1 I-36

Box 1

Box 2

Unsaturated-Soil Layer

Surface-Soil Layer

Thickness = H2

Thickness = H1

Ground-Water Aquifer Surface-Water Pond

Box 3

Ground-Water Well

I = Infiltration

Precipitation/Evaporation

Irrigation Domestic Use decom-025.ppt 041403

Box 1

Box 2

Unsaturated-Soil Layer

Surface-Soil Layer

Thickness = H2

Thickness = H1

Ground-Water Aquifer Surface-Water Pond

Box 3

Ground-Water Well

I = Infiltration

Precipitation/Evaporation

Irrigation Domestic Use

Box 1

Box 2

Unsaturated-Soil Layer

Surface-Soil Layer

Thickness = H2

Thickness = H1

Ground-Water Aquifer Surface-Water Pond

Box 3

Ground-Water Well

I = Infiltration

Precipitation/Evaporation

Irrigation Domestic Use decom-025.ppt 041403

No retardation is assumed to occur in the aquifer; however, radioactive decay is taken into account. A volume of contaminated water equivalent to the irrigation volume is assumed to be returned annually to the source zone. The concentration of radionuclides in the irrigation water is assumed to remain constant during the year. Radionuclides deposited on the vegetation are assumed to be removed at a constant rate. The DandD ground-water model should generally provide a conservative representation of the ground water system because it allows very little dilution and nominal attenuation.

The generic surface-water conceptual model in DandD assumes that radionuclides are uniformly mixed within a finite volume of water representing a pond. Radionuclides are assumed to enter the pond at the same time and concentration as they enter the ground water. Accordingly, there is assumed to be no transport of radionuclides through the ground water to the pond and thus no additional attenuation (besides the initial ground water dilution) is assumed for transport in the ground water. The surface-water model within DandD should provide a conservative dose estimate as long as a small volume is assumed for the surface-water pond. Because the parameters in DandD are selected to provide a conservative dose estimate, the generic conceptualization of the surface-water pathway should generally provide a conservative representation of transport of radionuclides through the surface-water pathway. Figure I.3 shows the generic ground-water and surface-water conceptual model within DandD.

Figure I.3 DandD Conceptual Model of the Ground-Water and Surface-Water Systems (from NUREG/CR–5621).

I-37 NUREG-1757, Vol. 2, Rev. 1

The generic conceptual model of the air pathway in DandD assumes an equilibrium distribution between radionuclides in the air and soil. The concentration in air is assumed to be a function of the soil concentration and a constant dust loading in the air. Accordingly, all radionuclides in the air are assumed to be in a particulate form. The air pathway model within DandD is very simple and should generally allow a conservative dose estimate as long as a conservative particulate concentration is assumed. Because the default parameters in DandD are geared to be conservative, in general the air pathway in DandD should allow a conservative dose estimate.

In general, the conceptual models within DandD are expected to provide a conservative representation of site features and conditions. Therefore, for screening analyses, NRC staff should consider such generic conceptual models to be acceptable provided it is acceptable to assume that the initial radioactivity is contained in the top layer (building surface or soil) and the remainder of the unsaturated zone and ground water are initially free of residual radioactivity. In using DandD for site-specific analyses, it is important to ensure that a more realistic representation of the site that is consistent with what is known about the site would not lead to higher doses. Some site features and conditions that may be incompatible with the generic conceptual models within DandD are listed in Table I.4. The relative importance of the incompatibilities varies with the scenario and radionuclides involved. More information on the assumptions of the model available in the development documentation (e.g., the NUREG/CR–5512 series).

For any site where it is known that one or more of these conditions or features are present, the licensee should provide an appropriate rationale on why the use of the DandD should not result in an underestimation of potential doses at the specific site.

NUREG-1757, Vol. 2, Rev. 1 I-38

Table I.4 Site Features and Conditions that May be Incompatible with Those Assumed in DandD

Sites with highly heterogeneous radioactivity

Sites with wastes other than soils (e.g., slags and equipment)

Sites that have multiple source areas

Sites that have contaminated zones thicker than 15 cm (6 in.)

Sites with chemicals or a chemical environment that could facilitate radionuclide releases (e.g., colloids)

Sites with soils that have preferential flow conditions that could lead to enhanced infiltration

Sites with a perched water table, surface ponding, or no unsaturated zone

Sites where the ground water discharges to springs or surface seeps

Sites with existing ground water contamination

Sites where the potential ground water use is not expected to be located immediately below the contaminated zone

Sites with significant transient flow conditions

Sites with significant heterogeneity in subsurface properties

Sites with fractured or karst formations

Sites where the ground water dilution would be less than 2000 m3 (70,000 ft3)

Sites where overland transport of contaminants is of potential concern

Sites with radionuclides that may generate gases

Sites with stacks or other features that could transport radionuclides to result in a higher concentration offsite than onsite

As example, it may be possible to demonstrate the acceptable use of DandD for analyzing sites that contain H-3 and C-14, although both radionuclides may be occur as a gas. The following approach can be used to demonstrate the acceptable use of DandD for analyzing sites that contain either H-3 or C-14 (NRC 1999a): (1) determine the area of the contaminated zone; (2) run DandD for the site with only H-3 or C-14; (3) read the associated activity ratio factor for the given area from Figure I.4; and (4) estimate the potential missed dose by multiplying the inhalation dose calculated from DandD by the activity ratio factor.

I-39 NUREG-1757, Vol. 2, Rev. 1

A ct iv ity

R at io

Contaminated area (m2)

1E+03

1E+04

1E+05

1E+06

10 100 1000 10000 100000 1000000

Activity Ratio of Vapor to Particulate A ct iv ity R at io

Contaminated area (m2)

1E+03

1E+04

1E+05

1E+06

10 100 1000 10000 100000 1000000 1E+03

1E+04

1E+05

1E+06

10 100 1000 10000 100000 1000000

Activity Ratio of Vapor to Particulate

Figure I.4 Activity Ratio of Vapor to Particulate as a Function of Contaminated Area.

I.4.3.2 Site-Specific

For site-specific analyses, the intent is to provide a more realistic assessment of doses based on more site-specific information and/or data. Presumably for such analyses, more is known about the site from which to develop a conceptual model. For site-specific analyses, the licensee should provide a schematic or verbal description of the problem that it is attempting to analyze.

Even when using a computer code that has a predefined conceptual model, it is important for the licensee to identify any site features or conditions that may differ from those assumed in the code. In developing a site-specific conceptual model or identifying potential limitations with a predefined conceptual model, the issues listed in Table I.5 should be considered.

Because conceptual models are developed based on limited data, in most cases more than one possible interpretation of the site can be justified based on the existing data. This uncertainty should be addressed by developing multiple alternative conceptual models and proceeding forward with the conceptual model(s) that provides the most conservative estimate of the dose and yet is consistent with the available data. Consideration of unrealistic and highly speculative conceptual models should be avoided. Consistent with the overall dose modeling framework of starting with simple analyses and progressing to more complex modeling, as warranted, it may be advisable for the analyst to begin with a simple, conservative analysis that incorporates the key site features and processes and progress to more complexity only as merited by site data. It is important to stress that a simple representation, of the site, in itself does not mean that the analysis is conservative. It is incumbent on the licensee to demonstrate that its simplification is

NUREG-1757, Vol. 2, Rev. 1 I-40 justified, based on what is known about the site and the likelihood that alternative representations of the site would not lead to higher calculated doses.

Table I.5 Issues To Be Considered in Developing a Site-Specific Conceptual Model

Whether a more realistic representation of the site would lead to higher doses

Whether the conceptual model accounts for the most important physical, chemical, and biological processes at the site

Whether the conceptual model adequately represents responses to changes in stresses

Whether the conceptual model includes consistent and defensible assumptions

In general, there are two primary areas of the dose analysis where the conceptual model is expected to change from one site to another; these are related to the source term and environmental transport. Aspects of the analysis related to the exposure pathways in the biosphere and dosimetry are largely determined by the scenario and the assumed behavior of the critical group. Accordingly, models related to the exposure pathways in the biosphere and dosimetry should not change from one site to another unless there is a significant change in the scenario and associated critical group. The principal environmental transport pathways that should have to be considered in a dose assessment are ground water (including transport through the unsaturated zone), surface water, and air.

The conceptual model of the source area should describe the contaminants and how they are likely to be released into the environment. Specifically, it should describe key features and processes such as the infiltration of water into the source area, the geometry of the source zone, the distribution of contaminants, release mechanisms, the physical form of the contaminants, near-field transport processes, and containment failure. If the contaminants are assumed to be uniformly distributed, this is an important assumption that needs to be justified because in general contaminants may not be uniformly distributed (see discussion under Section I.2 of this appendix). The source description should clearly identify how the contaminants are assumed to be released from the media. Common release mechanisms are diffusion, dissolution, surface release, and gas generation. The source description should also identify key processes and features that may retain or limit the release of contaminants from the source area (e.g., solubility and sorption). In addition, the description of near-field transport should state assumptions made regarding the dimensionality. In general, the assumption of one-dimensional vertical flow should be appropriate, unless there is some type of barrier present that may hinder flow in the vertical direction. The description of the source term should also describe failure mechanisms for any containment (e.g., corrosion, concrete degradation, or cover degradation) if containers or other forms of containment are present.

I-41 NUREG-1757, Vol. 2, Rev. 1

The conceptual model of the ground water pathway should describe how contaminants could migrate through the unsaturated and saturated zones to potential receptors (e.g., a well, spring, or surface-water bodies). Essential features that should be included in the conceptual model include hydrostratigraphic units; information on the geometry of the pathway (i.e., boundaries and boundary conditions); the physical form of the contaminants (i.e., dissolved, suspended sediment, gas, etc.); structural features of the geology (i.e., those that influence contaminant transport such as fractures, faults, and intrusions); and physical and chemical properties.

Important processes that should be characterized include the dimensions and state conditions (e.g., steady-state) of flow; dimensions and state conditions of transport (e.g., dispersion);

chemical and mass transfer processes (e.g., sorption, precipitation, complexation); and transformation processes (e.g., radioactive ingrowth and decay). Although contaminant migration through both the unsaturated and saturated zones is best represented in three dimensions, it may be appropriate to assume only one or two dimensions, if this provides a more conservative representation of contaminant migration, and/or if it can be demonstrated that migration in one or more other directions is not expected to result in exposure to potential receptors.

The conceptual model of the surface-water pathway should describe potential contaminant migration through surface-water bodies such as lakes, streams, channels, or ponds to potential receptors. Essential features that should be included in the conceptual model include: the geometry of the surface-water body (i.e., boundaries and boundary conditions); the physical form of the contaminants (e.g., dissolved or solid); and physical and chemical properties. Key processes that should be described include: the dimensions and state conditions of flow and transport; chemical and mass transfer processes (e.g., sorption, precipitation, volatization); and transformation. One key boundary condition that should be described is how the contaminants are expected to initially mix or interact with the surface water.

The conceptual model of the air pathway should describe potential contaminant migration through the air to potential receptors. Essential features that should be included in the conceptual model are similar to those for the other environmental pathways—namely, the geometry (i.e., boundaries and boundary conditions); form of contaminants (e.g., particulates or gases); and physical and chemical properties. Key processes that should be described include the dimensions and state conditions of flow and transport, and transformation processes.

I.4.3.2.1 Site-Specific Computer Codes

Two common computer codes used for site-specific analyses are RESRAD and RESRAD–BUILD. Both these computer codes have predefined conceptual models. Therefore, in using these codes, it is important for the licensee to demonstrate that key site features and conditions are consistent with the modeling assumptions within the codes or, where they are not consistent, the analysis may not result in an underestimation of potential doses.

NUREG-1757, Vol. 2, Rev. 1 I-42

I.4.3.2.1.1 RESRAD–BUILD

The RESRAD–BUILD code can be used to evaluate doses for the building occupancy scenario.

It considers exposure from external radiation at the source and air submersion, inhalation of airborne material, and inadvertent ingestion of radioactive material. Exposure to direct radiation at the source is calculated using surface source dose rate factors from Federal Guidance Report No. 12. RESRAD–BUILD incorporates correction factors to account for a finite area source, for any offset of the receptor from the axis of the disk of residual radioactivity, and for shielding by material covering the residual radioactivity. Exposure to external radiation from air submersion is calculated as an infinite cloud of material using dose rate conversion factors for an infinite cloud. RESRAD–BUILD models airborne concentration of radionuclides using a dynamic model that accounts for the kinetic introduction and removal of radioactive material to and from indoor air. Exposure to incidental ingestion of radioactive material is modeled using a constant transfer rate.

I.4.3.2.1.2 RESRAD

RESRAD can be used for analyzing the resident farmer scenario. As with the generic conceptual models used by DandD for analyzing the resident farmer scenario, the conceptual models in RESRAD (see Figure I.5) are more complex than those in RESRAD–BUILD. RESRAD models external exposure from volume soil sources when the person is outside, using volume dose rate factors from Federal Guidance Report No. 12. Correction factors are used to account for soil density, areal extent of residual radioactivity, thickness of residual radioactivity, and cover attenuation. When the person is indoors, exposure from external radiation is modeled in a similar manner except that additional attenuation is included to account for the building.

Exposure through ingestion of contaminated animal and plant products is modeled simply through the use of transfer factors.

The generic source-term conceptual model in RESRAD assumes a time-varying release rate of radionuclides into the water and air pathways. Radionuclides in the contaminant zone are assumed to be uniformly distributed. No transport is assumed to occur within the source zone, but radioactive decay is accounted for. In terms of containment, the radioactive material is not assumed to be contained (or containers are assumed to have failed). RESRAD does allow inclusion of a cover over the contaminated area. However, the cover is not assumed to limit infiltration of water, and is assumed to function only in terms of providing shielding from gamma radiation. Release of radionuclides by water is assumed to be a function of a constant infiltration rate, time-varying contaminant zone thickness, constant moisture content, and equilibrium adsorption. The contaminant zone is assumed to decrease over time from a constant erosion rate. RESRAD assumes a uniform release of tritium and C-14 gases, based on a constant evasion loss rate. Particulates are assumed to be instantaneously and uniformly released into the air as a function of the concentration of particulates in the air, based on a constant mass loading rate.

I-43 NUREG-1757, Vol. 2, Rev. 1

Figure I.5 Conceptualization Modeled by RESRAD (from ANL/EAD/LD–2).

The RESRAD generic conceptual ground-water model assumes one or more horizontal homogeneous strata for the unsaturated zone. Transport in the unsaturated zone is assumed to result from steady-state, constant vertical flow, with equilibrium adsorption, and decay, but no dispersion. RESRAD has two different ways of modeling radionuclides once they reach the saturated zone. In the mass-balance approach, radionuclides entering the saturated zone are assumed to be instantaneously and uniformly distributed over a constant volume equivalent to the volume of water removed by the hypothetical well (as long as the pumping rate is larger than the rate of leachate entering the ground water—if not, no dilution is assumed to occur in the ground water). For the mass-balance approach, radionuclides are assumed to enter a well pumping immediately beneath the contamination zone. The mass-balance approach is very similar to the ground-water modeling approach in DandD. In the nondispersion approach, transport in the saturated zone is assumed to occur in a single homogeneous stratum, under steady-state, unidirectional flow, with a constant velocity, equilibrium adsorption, and decay. It assumes no dispersion; however, radionuclides are assumed to be diluted by clean water as a function of the assumed capture zone of the hypothetical well, in relation to the width of residual radioactivity and the depth of residual radioactivity, in relation to the depth of the hypothetical well. Radioactive decay and equilibrium adsorption are assumed to occur for the nondispersion approach. Radionuclides are assumed to enter a well located at the immediate downgradient

NUREG-1757, Vol. 2, Rev. 1 I-44

Uw v dw

A len

I v len dw and

(I–1)

Uw A len len dw v d

I v w⋅ and

(I–2) edge of the contamination zone. For the nondispersion model, the calculated width of the effective pumping zone could be a factor of 2 larger than what one would predict from a steady-state capture zone analysis; this could lead to a slight overestimation in the amount of dilution (NRC 1999a).

In determining which of these two conceptual models to use, consideration should be given to where the hypothetical well may be located (i.e., either at the center of the residual radioactivity or at the edge of the residual radioactivity); the relative half-life of the radioactivity; and the potential capture zone of the hypothetical well. Use of the nondispersion model will generally result in lower estimated doses. Both models assume that radionuclides enter the well as soon as they reach the water table. However, the nondispersion model, unlike the mass-balance model, calculates the time it takes for the peak concentration to occur after the initial breakthrough.

Accordingly, the nondispersion model accounts for radioactive decay during the interval between the initial breakthrough and arrival of the peak concentration. Generally, the amount of decay should be small unless the radionuclides have short half-lives and are retarded. In addition, unlike with the mass-balance model, for the nondispersion model no assumption is made that all radionuclides released from the contaminated zone are withdrawn through the well.

Therefore, the nondispersion model may include dilution. The only way that dilution is not considered is if the expected capture zone of the hypothetical well is small in relation to the width and thickness of the residual radioactivity. Because the nondispersion model will generally give a lower estimated dose than the mass-balance model, it is important for the licensee to justify the use of this model for the specific analysis. Use of the mass-balance approach should always be acceptable. In Equations I–1 and I–2, use of the nondispersion model should be acceptable, without additional justification, for modeling long-lived radionuclides (i.e., where radioactive decay is not important) when either one of the following conditions are met:

or

I-45 NUREG-1757, Vol. 2, Rev. 1

Aw Uw

I = (I–3) where Uw = pumpage rate from the well (m3/y);

= ground-water darcy velocity (m/y);

A = area of residual radioactivity (m2);

dW = depth of well intake below water table (m);

len = length of residual radioactivity parallel to ground water flow (m); and I = infiltration rate (m/y).

As a general rule, use of the nondispersion approach should be acceptable when the area of residual radioactivity is known to be larger than the assumed capture area of the hypothetical well. Assuming an essentially flat water table and steady-state conditions, the capture area of the hypothetical well can be calculated in Equation I–3 as follows:

where AW = area of well capture (m2);

UW = pumpage rate from the well (m3/y); and I = infiltration rate (m/y).

The generic conceptual model of the surface-water pathway in RESRAD assumes that radionuclides are uniformly distributed in a finite volume of water within a watershed. For example, the default watershed area in RESRAD Version 5.91 is 1x106 m2 (250 ac).

Radionuclides are assumed to enter the watershed at the same time and concentration as in the ground water. Accordingly, no additional attenuation is considered as radionuclides are transported to the watershed. In the surface water, radionuclides are assumed to be diluted as a function of the size of the contaminated area in relation to the size of the watershed. The RESRAD surface-water conceptual model assumes that all radionuclides reaching the surface water are derived from the ground water pathway. Thus, transport of radionuclides overland from runoff is not considered. In addition, additional dilution from overland runoff is not considered.

The generic conceptual model of the air pathway in RESRAD uses a constant mass loading factor and area factor to model radionuclide transport. The area factor, which is used to estimate the amount of dilution, relates the concentration of radionuclides from a finite area source to the concentration of radionuclides from an infinite area source. It is calculated as a function of particle diameter, wind speed, and the side length of a square-area source. The conceptual model assumes a fixed particle density, constant annual rainfall rate, and constant atmospheric stability.

No radioactive decay is considered. See Chang, et al., (1998) for more detail. Tritium and C-14 gases are assumed to be uniformly mixed in a constant volume of air above the contaminated zone. RESRAD does not model the transport of tritium and C-14 as particulates in the air.

NUREG-1757, Vol. 2, Rev. 1 I-46

I.4.3.2.2 Limitations of Site-Specific Computer Codes

In…