Page 1 of 6

The methodology in this document has been earlier described as screening level in terms of theoretical sophistication, but site specific in its application. Chapter 2 described concepts of exposure and assigned values to exposure parameters which define, for purposes of demonstration, a central and a high end exposure pattern. Chapters 3 and 4 described algorithms for the fate, transport, and transfer of dioxin-like compounds, and also assigned parameter values for purposes of demonstration. The methodology was demonstrated in Chapter 5, using exposure and fate and transport parameters which had been laid out in earlier chapters. Those who wish to use the methodology for further analysis of incremental exposures to sources of dioxin-like compounds are now in a position to use the same algorithms, perhaps many of the same parameter values. The purpose of this chapter is to provide guidance on some key issues for potential users.

Section 6.2 discusses the use of the parameter values selected for the demonstration scenarios in Chapter 5 for other applications. Section 6.3 is a sensitivity analysis exercise on the parameters required for algorithms estimating exposure media concentrations. Section 6.4 addresses the issue of mass balance with regard to the source strength terms of the four source categories.


Table 6-1 lists all the parameters, including names, definitions, and units, that are required for the methodologies of this assessment except the exposure parameters. Exposure parameters are given in Table 2-1 of Chapter 2. Table 6-1 also gives four additional pieces of information for each parameter listed. Three are numerical values which were used in the sensitivity analysis exercises that are described in Section 6.3. below. The parameter values labeled "selected" were the ones used in the demonstration of the methodologies in Chapter 5. Section 6.3. below justifies the high and low values of parameters selected for sensitivity analysis.

table Table 6-1. Parameters used to estimate exposure media concentrations for this assessment.  

Other users of this methodology may wish to view these high and low values as reasonable high and low possible values for their applications; note however that the chemical specific parameters are those only for 2,3,7,8-TCDD.

The fourth piece of information is a qualitative judgement on the part of the authors of this document as to the appropriateness of using the "selected" parameter values for other assessments.

This judgement is categorized in three ways:
expand table Table V3 6-1

1) First Order Defaults, or FOD:
As defaults, these parameters are independent of site specific characteristics and can be used for any assessment. Also, as first order defaults, it is felt that the values selected for the demonstration scenarios carry a sufficient weight of evidence from current literature such that these values are recommended for other assessments. Several of the chemical specific parameters, such as the Henry's Constant, H, and the organic carbon partition coefficient, Koc, fall into this category. The qualifier above, "current literature", indicates that new information could lead to changes in these values.

2) Second Order Defaults, or SOD:
Like the above category, these parameters are judged to be independent of site specific characteristics. However, unlike the above category, the current scientific weight of evidence is judged insufficient to describe values selected for demonstration purposes as first order defaults. SOD parameters of principal note are the bioconcentration parameters specific to the chemicals, such the Biota Sediment Accumulation Factor, or BSAF. This parameter translates a bottom sediment concentration to a fish tissue concentration. The science is evolving for this parameter, including thought on the extent to which BSAFs generated for one species at one site can be generalized to other sites and/or species, the differences in BSAF between column and bottom feeders, the differences between past and ongoing contamination, and so on. Users should carefully review the justification for the SOD values selected for the demonstration scenarios before using the same values.

3) Site Specific, or SS:
These parameters should or can be assigned values based on site-specific information. The information provided on their assignment for the demonstration scenarios, and for selection of high and low values for sensitivity analysis testing, is useful for determining alternate values for a specific site. A key class of SS parameters which were not fully included in Table 6-1 above are the source strength terms - the soil concentrations, effluent discharge rates, and stack emission rates. There are likely to be site-specific applications of this methodology for which detailed information is unavailable. Often the midrange values selected for the demonstration scenarios are suitable for site specific applications when data is unavailable.

The exposure parameters have not been categorized as have the contaminant fate and transport/transfer parameters. Assignment of these values are critical as LADD estimates are linearly related to parameter assignments - doubling exposure duration assumptions double LADDs, and so on. All exposure parameters were developed based on information and recommendations in EPA's Exposure Factors Handbook (EPA, 1989) and Dermal Exposure Assessment: Principals and Applications (EPA, 1992). Some of the exposure parameters of Table 2-1, Chapter 2, are appropriately described as FOD.

These include:
lifetime, body weights, water ingestion rates, inhalation rates, and an exposure duration for a childhood pattern of soil ingestion. All of the other exposure parameters are better described as either SOD or SS. Attaining site-specific information is recommended for them. However, this is often difficult for site specific assessments and impractical if the procedures in this assessment are used in general assessments. In the absence of site specific information, the following parameters can be considered SOD: adult exposure durations of 9 years for central scenarios (whether they be modeled after "residential" settings or not) and 20 years for high end scenarios (whether "farming" be the model for high end exposures or not), childhood soil ingestion rates, the fruit/vegetable food ingestion rates, the fraction of fruit/vegetable consumption that comes from a home garden, and the fractions of time spent at home (which are applied to inhalation and water ingestion pathways).

The remaining exposure parameters pertain to the exposure pathways evaluated as most critical to dioxin exposures. For this reason, users should either pursue site specific information or carefully justify parameter selections in the absence of site specific information. These include the rate of beef, milk, and fish ingestion and the fraction of these food products which are impacted by the source. Fish ingestion rates for the demonstration of methodologies in this assessment were 1.2 g/day for central scenarios and 4.1 g/day for high end scenarios.

These were developed using an approach recommended in the Handbook when site specific data was unavailable. Specifically, a meals per year of fish recreationally caught from the impacted water body was assumed, and then this was translated to a grams per day consumed. These rates are both less than a national average estimate of fish consumption that was published in an water quality criteria document for 2,3,7,8-TCDD, 6.5 g/day (EPA, 1984). The setting for the demonstration scenarios was a rural setting which contained farm and non-farm residences, but which did not contain a major water body for frequent recreational or subsistence fishing purposes.

Rather, a smaller size water body which allowed for more occasional recreational fishing was assumed. In a setting where more substantial water bodies exist which do supply fish for commercial and recreational use, fish ingestion rates from these water bodies would be higher. The other parameters are the ingestion rates and contact fractions for beef and milk ingestion. The ingestion rates for these food products were 50% ingestion rates given in the Handbook. The contact fractions assigned for the high end scenarios were developed from a USDA survey (USDA, 1966) of rural farm households, some of which home produced. For home producers only, the percent of their total ingestion of beef and milk which was homegrown was 44% and 40%, respectively.

In addition to the above qualifications, the parameters of this methodology have been categorized in terms of their role in the methodology. The following is a brief description of three principal categories.

Category 1. Human behavior exposure parameters
These are the contact rates, contact fractions, exposure durations, lifetime and body weights used in the following equation for lifetime average daily dose:

Equation V3 6-1

Category 2. Fate, transport, and transfer parameters
These parameters are all the parameters required to estimate exposure media concentrations, except those specifically associated with a contaminant - chemical-specific parameters are included in Category 3 below. All fate, transport, and transfer parameters are listed, defined, and further subcategorized in Table 6-1. Not included in the discussions in Section 6.3 are perhaps the most important terms in this category, and these are critical source strength terms: the concentrations of dioxin-like compounds for the soil source categories (onsite and offsite source categories), and the release quantities of dioxin-like compounds into the air for the stack emission source category and into the surface water for the effluent discharge source category.

A general comment that can be made for fate and transport parameters is that values for the demonstration scenarios were selected to be midrange and plausible, and that this document provides information on selecting alternate values for site-specific applications. Most of the parameters in this category fall under the SS qualification. Subcategories within the fate and transport category include:

- Contaminated and exposure site characteristics:
These are areas, soil properties, and depths of tillage (which are depths to which residues transported by erosion or deposition are mixed in conditions of tillage such as agriculture or gardening, and no tillage). Like the soil concentration term, the area of contamination is a site-specific and critical parameter. Soil properties were assigned to be midrange and typical of agricultural soils. Depths of mixing for tilled and untilled circumstances are not known with certainty, and these two parameters were characterized as SOD.

- Soil and sediment delivery parameters:
These include parameters associated the erosion of contaminated soil from a site of contamination to a nearby site of exposure and/or to a nearby surface water body. All but one of the parameters in this subcategory are physical, site-specific parameters which should be evaluated for site specific applications. The one parameter not of this description is the enrichment ratio, which describes the enrichment of eroded soil with dioxin-like compounds, and was assigned a rating of SOD. Geometric parameters include watershed drainage area, water body volumes, and distances. Physical parameters include soil loss estimates, organic carbon contents, water body suspended solids, and background watershed contaminant concentrations.

- Volatilization and dust suspension parameters:
These parameters are associated with suspension, dispersion, and transport of contaminants from contaminated soils. One parameter included in this category is the exposure duration, which appears to be misplaced. In fact, the exposure duration is used to determine the average vapor phase air concentration - this is further discussed in Section 6.3 below. Parameters in this category are site-specific and should be evaluated for specific methodology applications.

- Bioconcentration and biotransfer parameters:
These include parameters describing the biota and the media surrounding the biota which influence the transfer of dioxin-like compounds from the media to the biota. Some of these parameters are site-specific, although obtaining values may be difficult. Included here are annual rainfall, fish lipid contents, a fresh to dry vegetable weight conversion factor, and yields and intercept fractions for vegetation categories.

Others are theoretical; values for these were determined from the literature and can be used for other assessments if better information is unavailable. Included here are atmospheric deposition velocities of particles, washout of wind-suspended particles from the atmospheric, the retention of wet particle depositions on vegetations, empirical correction factors for atmospheric to plant and soil to plant transfers, and the bioavailability of soil as compared to vegetation as a vehicle of transfer of dioxin-like compounds to cattle.

These were given a rating of SOD. A third group describes exposure of cattle to dioxin-like compounds through their diet. These include fractions of cattle diet which are soil, pasture grass, and cattle feed, and the extent to which these three are impacted by the source of contaminant. Sensitivity analysis below shows how beef concentrations are impacted by changes in assumptions of how cattle are exposed to dioxin-like compounds through their diet. Since beef and milk dietary exposures are most critical for human exposure, the cattle exposure assumptions made for demonstrating the methodologies of this assessment should be carefully considered before using them for other assessments.

- Effluent discharge source category:
These are three physical parameters that can be determined on a site-specific basis, and include flow rates of the effluent and receiving water body, organic carbon contents of suspended solids in the effluent and the receiving water body, and suspended solids content of the effluent and the receiving water body.

- Stack emission source category:
In fact, most of the parameters required to evaluate the impact of stack emissions to a nearby site of exposure have been included in other categories. Sensitivity analysis only focuses on parameters and issues unique to this category. One set of input values are contaminant wet and dry deposition rates. Three depositions are required: one for the site of exposure, one to represent depositions on watershed soils which drain into the water body, and one to represent direct deposition onto the water body.

These were all generated using the COMPDEP model, as described in Chapter 3. Two other key inputs generated by the COMPDEP model are the ambient air vapor phase and particle phase concentrations of contaminant at the site of exposure. All such quantities are a function of that model's algorithms and parameter input requirements, particularly the release rate from the stack. Information on the COMPDEP model and its application is given in Chapter 3 and not discussed further in this chapter. Users can determine air concentrations and contaminant deposition rates in other ways, and use those in the methodologies to determine impacts and exposures.

The no-till depth of mixing at the site of exposure, dnot, is required for the off-site soil source algorithm as well. It's selected value for the stack emission source category was 1 cm in contrast to 5 cm assumed for the off-site soil source category; hence its impact is examined twice in Section 6.3. below. The only other unique parameters not included in other subcategories are the average watershed mixing depth (used for determining watershed soil concentrations, which are then used to determine impacts to water bodies) and the fraction of particles depositing on water bodies which remain in suspension. These are both theoretical values and can be used in other assessments lacking better information.

Category 3. Chemical properties of dioxin-like compounds
The ten chemical-specific parameters required for the algorithms of this assessment fall under two categories, FOD and SOD. As such, they are all independent of the specifics of the site. The parameters deemed FOD are chemical fate and transport parameters, some of which are common and often determined in laboratory conditions. These include the Henry's Constant, the organic carbon partition coefficient, molecular diffusivity in air, a plant weathering rate constant for contaminated particles, and the soil dissipation rate for eroding or depositing contaminants.

The selected values for these parameters are, in the authors' opinion, the best values derivable from current data. A second set of chemical specific parameters are associated with bioconcentration/biotransfer algorithms. Some of them are determined from field data (data on dioxin-like compounds or other compounds), and others are determined by experimentation and with that experimentation, development of empirical relationships between a critical transfer factor and the chemical's octanol water partition coefficient.

The authors cannot be definitive in a judgement that values given to these parameters be considered default, hence the SOD rating. For these compounds, field/experimental data is conflicting or there simply is a lack of appropriate data. Parameters included in this category are a soil to below ground vegetation transfer factor, two air to plant factors: the air-to-leaf vapor phase transfer coefficient and the plant washoff rate constant, two water body to fish parameters: the biota to sediment accumulation factor and the related biota to suspended solids accumulation factor, and a beef/milk bioconcentration factor.


Sensitivity analysis was undertaken in order to evaluate the impact of model results with changes in model parameters. The following sections describe the limitations, methodology and parameter selections, and results.

6.3.1. Limitations of the Sensitivity Analysis Exercises

The exercises were not comprehensive and/or definitive. Following are some key limiters:

. The COMPDEP model was not evaluated in this section. Chapter 3 describes the COMPDEP model. No sensitivity analysis runs were performed on COMPDEP model output for this chapter. This section does evaluate the impact of different deposition rates and modeled ambient air concentrations on exposure sites soils, surface water, and biota.

. Sensitivity to changes in exposure parameters was not evaluated. The basic equation for evaluating lifetime average daily dose was given above as Equation (6-1).

Chapter 2 described all terms in this equation except the exposure media concentration, which was the focus of Chapter 4. Because LADD estimates are a linear function of all exposure parameters, sensitivity analysis was not performed on LADD exposure estimates. The focus of this section instead is on the fate, transport, and bioconcentration/biotransfer algorithms used to estimate the exposure media concentration term in Equation (6-1).

. The analysis was not exhaustive in its coverage. Principal algorithms in the fate, transport, and transfer of dioxin-like compounds were evaluated, and all parameters required for algorithms were tested at least once. However, not all possible tests were conducted. Before noting those, following is a list of algorithms which were tested:

- Volatilization/suspension and transport of vapor/particle phase airborne residues from a site of soil contamination to a nearby site of exposure (using algorithms developed for the off-site soil source category);

- Volatilization/suspension and dispersion of vapor/particle phase contaminants for the circumstance where soil contamination is at the site of exposure (on-site soil source category);

- Transport via erosion of contaminants at a site of soil contamination to a nearby site of exposure to impact exposure site soils (off-site soil source category);

- Transport via erosion of contaminants at a site of soil contamination to a nearby surface water body, to impact bottom sediments, water, and fish (off-site soil source category);

- Transfers of contaminants from soils to below ground vegetables and from air to above ground vegetations (on-site soil source category);

- Transfers of contaminants from soils and vegetation to beef (on-site soil source category);

Direct discharges of dioxin-like compounds into surface water bodies, and the effect of surface water and effluent parameters on fish and water concentration estimation (effluent discharge source category); and

- Particle depositions and ambient air concentrations, which result from stack emissions, onto exposure site soils, watershed soils, surface water bodies, and biota (stack emission source category);

The exercise was purposefully limited since several possible exercises would have been duplicative. For example, impacts to beef and milk in the off-site soil source category are, of course, modeled within this assessment, but specific sensitivities to beef and milk concentration predictions with parameter changes within the off-site soil source category are not explicitly evaluated below.

Parameters required for the beef bioconcentration algorithm are evaluated in the context of the on-site source category, and these are the same ones required for all three source categories which include beef impacts (the effluent discharge source category does not include beef and milk impacts). Further, only impacts to the beef algorithms were tested. The milk bioconcentration algorithm was not tested because principal conclusions from the beef exercise are generally true for the milk algorithm. Any generalizations from the on-site source category exercises are transferable to the other two source categories.

A related limitation has to do with the cascading effect of certain parameters. For example, a key contaminant parameter is the organic partition coefficient, Koc, which impacts (among other concentrations) vapor phase air concentrations. Air concentrations are used to estimate above ground vegetation concentrations, including those of grass and cattle feed. Beef concentrations are a function of concentrations in grass and cattle feed. What is not done for this example (and many others like it) is to evaluate the impact of changes in Koc to beef concentrations. What is done, however, is as follows.

The sensitivity of air concentration predictions to changes in the partition coefficient are evaluated. Then, the sensitivity to grass and cattle feed concentrations to plus and minus one order of magnitude differences in estimated vapor phase air concentrations are evaluated. In this way, any possible parameter change(s) which influences air concentrations within a plus/minus order of magnitude range is evaluated for grass and feed concentrations. Finally, beef concentration estimations are evaluated within a similar plus/minus order of magnitude change for grass and feed concentrations. With some examination, therefore, the effect of cascading impacts can be determined.

The impact of changing soil concentrations to estimates of exposure media concentrations (air, water, biota) is linear and direct in all cases - i.e, increasing soil concentrations by a factor of five increases all impacted exposure media by the same factor of five. For this reason, soil concentrations are not displayed in the sensitivity graphs displayed in the next section, with one exception. This was in the estimation of beef concentrations from soil contamination.

Beef concentrations are a function of concentrations in the dry matter diet of the cattle, including soil, grass, and cattle feed. Therefore, if soil concentrations were to change and concentrations on the other intakes were to not change, than beef concentrations would not be a linear and direct function of soil concentrations. However, and in the context of this sensitivity analysis, when changing only soil concentrations, vegetative concentrations are linearly and directly impacted by the same order of magnitude change. Therefore, beef and milk concentrations turn out to be linearly related to soil concentrations.

A final limitation to note is that this exercise does not evaluate the multiple effects of changing more than one independent parameter simultaneously. Other numerical methods, particularly Monte Carlo, can be used to evaluate the impact of simultaneous changes to model parameters. Applications of this technique to dioxin exposure assessments are discussed in Chapter 7 of this volume.

There are instances where parameters were evaluated as dependent and changes were made simultaneously. One example is in three parameters which are related to the size of a watershed (also termed the "effective drainage area" since such an area might be smaller than a surrounding river system watershed), and which are important in determining the impact of a bounded area of soil contamination to a nearby surface water body.

These three include the watershed size, the watershed sediment delivery ratio (which decreases as watershed size increases), and the surface water body volume (which increases as watershed size increases, assuming sources of water - surface runoff, interflow, and groundwater recharge - remain the same on a per unit area basis). To test the impact of watershed size to surface water and sediment concentrations, all three parameters were changed simultaneously in modeling a small and a large watershed.

One set of parameters which might not be independent, but which were treated as such in the sensitivity testing, are the chemical specific parameters. For example, a higher organic carbon partition coefficient might be associated with a lower Henry's Constant - tighter binding to soils means less of a tendency to volatilize. Empirical relationships between such chemical specific parameters have not been established, and since there is uncertainty in precise values selected for the dioxin-like compounds, chemical specific parameters were treated as independent parameters.

. Only a high and a low value for model parameters were tested; no discussions of likelihood for parameter values or distributions of parameter values are included.
Certainly the identification of all model parameters and the justification for assignment of high and low values will be helpful to others using the methodology. Assignment of parameter values for purposes of demonstrating the methodologies in Chapter 5 should be carefully considered when users apply this methodology for specific purposes or specific sites.