of the sediment delivery ratio equation - developed from data on construction sites near surface water bodies - are more appropriate. The key source strength terms tested, the area of contamination, ASC, and the soil loss rate from the site of contamination, SL_s, both have a direct linear impact on the both sediment and surface water concentrations. The other soil loss term, the erosion rate for the watershed, SL_w, also has a direct linear impact. What also appears to be critical in this algorithm is the size of the watershed. The reduction of impacts seen with the "large" watershed necessitated a y-axis spanning two orders of magnitude.

The algorithm seemed fairly insensitive to the remaining four parameters tested. The average watershed concentration, initialized at 0.0 in order to just show the incremental impact from the contaminated site, was increased to 1 ppt. This was also the concentration used to demonstrate the on-site source category, and was identified as a possible background concentration of dioxin-like compounds. If this is a reasonable selection for a background soil concentration, it is seen in Figure 6-4 that background soils have a marginal impact on a water body which is impacted from a site of elevated soil concentrations. The impact of the organic carbon partition coefficient, Koc, on bottom sediments appears small despite the fact that the Koc range spans two orders of magnitude. This is an indication that it is so high for the dioxin-like compounds, that (at least in the algorithm of this assessment), its assignment is not critical for sediment concentration estimations. The same lack of impact appears to be the case for the organic carbon content of water body sediments, and the level of suspended solids in the water column.

6.3.3.5. Estimation of fish tissue concentrations Fish tissue concentrations for three of the four source categories of this assessment are a direct function of bottom sediment concentrations; the one source category where this is not true is the effluent discharge source category, where fish tissue concentrations are a function of suspended sediment concentrations. As laid out in Section 4.3.4.1, Chapter 4, whole fish tissue concentrations are estimated as: (C_sed/OC_sed) * BSAF * f(lipid). Therefore, any parameters which impact bottom sediments for the three pertinent source categories impact fish tissue concentrations. All the sensitivities to bottom sediment concentrations displayed in Figure 6-4 and discussed in the above section follow through with fish tissue concentrations. In Figure 6-5 and from the formulation above, it is seen that fish tissue concentrations are a direct linear function of bottom sediment concentrations. There is the same linear relationship between whole fish tissue concentrations and the other three parameters displayed on Figure 6-5. The linear relationship is direct for C_oc, BSAF, and f(lipid), and inverse for OC_sed. It is noted that the concentration on bottom sediments, C_sed, is impacted by the value assigned to OC_sed. However, as described in the previous section, the impact to C_sed with changes to OC_sed is marginal and in the same direction. For example, reducing OC_sed from its baseline of 0.03 to 0.01, reduces C_sed by a small amount. The impact to fish tissue from changes in OC_sed is more pronounced and essentially in an inverse linear manner, as shown by the formulation above and in

Figure 6-5.

6.3.3.6. Estimation of on-site air concentrations in the vapor phase The impacts of parameter changes for this algorithm are shown in Figure 6-6. These impacts are very similar to those for the algorithm estimating exposure site concentrations from an off-site area of soil contamination, which is shown in Figure 6-1. The principal difference in the algorithms is that the on-site algorithm has a near-field dispersion algorithm, whereas the off-site algorithm has a far-field dispersion algorithm. For all those parameters which only impact the volatilization flux only, the impacts are the same in the on-site and off-site categories. These include the exposure duration, ED, the organic carbon content, OC_sl, soil porosity, E_slp, particle bulk density, P_soil, and the three chemical-specific parameters, Henry's Constant, H, organic carbon partition coefficient, Koc, and molecular diffusivity, D_a. The impact of area is interestingly different in the onsite as compared to the off-site algorithm. For the off-site algorithms, the area term, ASC, impacts the source strength, with an order of magnitude increase in ASC increasing exposure site air concentrations by a little over 2 times (>200%). For the on-site algorithms, the area term, AES, impacts the dispersion algorithm, and the same order of magnitude increase in area only increases concentrations by around 30%.

6.3.3.7. Estimation of on-site air concentrations in the particulate phase The impacts of parameter changes to this algorithm are shown in Figure 6-7. Like the similarity between on-site and off-site impacts for concentrations in the vapor phase, these

impacts are also similar to those for the algorithm estimating exposure site particle phase concentrations from an off-site area of soil contamination, which is shown in Figure 6-2. And also similar is the difference in the role of the area terms for the off-site versus the

on-site particulate algorithms; the area term in the off-site algorithms impacts the source strength and in the on-site algorithm, it impacts the dispersion.

6.3.3.8. Vapor-phase transfers and particle phase depositions to above ground vegetations Concentrations in above ground vegetations are a function of vapor-phase transfers and particle phase depositions. Vapor and particle reservoirs originate from contaminated soils as volatilization and wind erosion, respectively. Atmospheric dispersion and deposition modeling delivers concentrations and depositions, respectively, from a stack to a site of exposure. This section focuses on the sensitivities of the transfer algorithms for the contaminated soils source categories. The same general trends would occur for the off-site soil and the stack emission source categories. The principal difference is in the relative proportions of the contaminant which are in the vapor and particle phases. As discussed below, more contaminant is delivered via particle depositions for the stack emission source category as compared to the soil contamination source categories.

Vapor transfers and particle depositions are evaluated in Figures 6-8 and 6-9. Three vegetations are modeled for this assessment, including vegetables/fruit, grass, and cattle fodder. The latter two are for the beef/milk bioconcentration algorithm, the first for human exposure via consumption of unprotected fruits or vegetables.

For vapor-phase impacts shown in Figure 6-8, it would appear that changes to total vegetation concentrations are critically a function of parameters specific to the vapor transfer algorithm. There is between one and two orders of magnitude range of plant concentrations predicted over the range of the vapor phase transfer coefficient, B_vpa, tested. This parameter is uncertain as well as very influential in this methodology. Also influential and uncertain is the empirical parameter introduced to model the difference between the leaves of the experiment for which B_vpa was developed and the bulky vegetation to which the B_vpa is applied, the VG parameters (VG_veg, VG_gr, and VG_fod). The need for such a correction factor is justified given the evidence that dioxin-like compounds to do not translocate into vegetations. The leaf concentrations in the experiments for which B_vpa was derived are likely to be analogous only to the outer layer concentrations in bulky vegetations, not the whole plant (or whole fruit/vegetable) concentrations. This empirical parameter was set to 0.01 for bulky fruits/vegetables, but was set at 1.00 for grass, under the assumption that grass is similar to leaves, and 0.50 for cattle fodder, which is assumed to contain some bulky (grains) and leafy (hay) vegetations. In any case, the impact to vegetation concentrations with changes to the VG parameters is significant and displayed on Figure 6-8.

A dry weight to fresh weight conversion factor, FDW, is required for estimating above ground concentrations of vegetable/fruits. This is because the algorithms estimate

above ground vegetative concentrations on a dry weight basis, and the concentrations need to be diluted since fruit and vegetable consumption are given in this assessment on a fresh weight basis. The impact to concentrations is direct and linear, and since the range of likely FDW is small, the impact is small as well. This parameter is also required for

the particle deposition algorithm, but is left out of Figure 6-9 for clarity. In fact, FDW is applied once vapor phase and particulate phase contributions to vegetable/fruit concentrations are already summed; in other words, it is not tied to either the vapor or particle phase algorithms.

The impact of all the particle phase parameters to overall plant concentrations is less than that of vapor transfers, as seen in a comparison between Figures 6-8 and 6-9. For the parameters including rainfall amount (R), washout factor (W_p), denseness of vegetation (as modeled by yield, Y, and intercept fraction, INT), velocity of particle deposition (V_p), and plant weather dissipation rate, kw, results in Figure 6-9 are for vegetable/fruits and not grass or fodder. Vegetables/fruits are more impacted by particle depositions than grass/fodder, and as seen, there is less than half an order of magnitude impact from the range of values for these parameters tested. The impact of depositions on vegetable/fruit concentrations occurs because the correction factor for vegetables, VG_veg, is equal to 0.01, which minimizes the vapor-phase contributions to vegetable concentrations in comparison to the contributions of the vapor phase concentrations for grass and fodder concentrations, which have correction factors of 1.00 (for grass) and 0.50 (for fodder).

Model results on the proportion of above ground plant concentrations that are due to air-to-leaf transfer and particulate deposition were examined for the soil contamination and stack emission source categories for 2,3,7,8-TCDD, and results are summarized in Table 6-2. Results show that vapor phase transfers tend to dominate vegetative concentrations, although particle phase concentrations are important for bulky fruits and vegetables. Again, the critical difference in the two plant types is the use of an empirical VG which reduces the magnitude of vapor phase impact for bulky vegetations. Results also show that the relative impact of vapors and particles is a function of distance for the stack emission source category. For the central stack emission Scenario, #4, where the site of exposure is 5000 meters from the stack, vapor transfers generally have more of an impact to vegetation as compared to the high end Scenario, #5, where the site of exposure is 500 meters away.

It is possible that the impact of particle depositions is being underestimated, for at least four reasons:

• The wind erosion algorithm estimating air-borne contaminant concentrations for soil contamination source category only estimates concentrations of PM-10, or inhalable size particulates, those 10 m m size diameter and less, while the COMPDEP

Table 6-2. Contribution of above ground vegetation concentrations of 2,3,7,8-TCDD from air-to-leaf transfers and particulate depositions.¹

Air-to-leaf Particulate

Description² Transfers Deposition

Scenarios 2: On-site

Soil Source Category

vegetables/fruit 49 51

pasture grass 94 6

fodder 95 5

Scenarios 3: Off-site

Soil Source Category

vegetables/fruit 68 32

pasture grass 97 3

fodder 98 2

Scenarios 4: Stack Emission

Source Category

vegetables/fruit 55 45

pasture grass 94 6

fodder 95 5

Scenarios 5: Stack Emission

Source Category

vegetables/fruit 32 68

pasture grass 89 11

fodder 88 12

The precise impact of these factors might be investigated more fully in a later assessment with additional models. Tests were run for this sensitivity analysis by increasing the amount of particulate phase contaminants depositing onto vegetation by an order of magnitude to the on-site demonstration scenarios, without changing the vapor phase contributions. The vapor phase/particulate phase contributions to above ground fruits and vegetables, originally 49%/51% (from Table 6-2 above), changed to 9%/91% with an order of magnitude increase in particulate phase contributions. Vegetable concentrations increased by a factor of 6. The impact was less for grass and fodder, with concentrations increasing by a factor of 1.7. The impact was comparable for the off-site soil contamination source category.

One possible theoretical shortcoming of the plant concentration algorithm for the soil source categories is that contaminants which volatilize from soil are assumed to remain in the vapor phase. Given the affinity of the dioxin-like compounds to soil, it seems possible, if not likely, that a portion of the volatilized flux will sorb to airborne particles and become part of the particulate reservoir. As discussed in Chapter 3, Section 3.4.5.5, Bidleman (1988) provides an algorithm to estimate what fraction of a total "available" reservoir of airborne contaminant would sorb to airborne particles. Assuming that volatilized residues would comprise an available reservoir, this sorbed fraction ranges from 10 to 70% (roughly), depending on the density of particles in the air. A test was conducted to see what impact partitioning of the vapor phase reservoir partly into the