PUBLIC HEALTH ASSESSMENT

LOWER DUWAMISH WATERWAY
SEATTLE, KING COUNTY, WASHINGTON

APPENDIX C: EXPOSURE ASSUMPTIONS

This section provides calculated exposure doses and assumptions used for each completed exposure pathway. The dose estimates for each pathway are discussed under the Pathways Analysis/Public Health Implications section. The following exposure parameters and dose equations were used to estimate exposure from fish consumption and direct contact with sediments at the LDW site.

Fish Consumption

Average and upper-bound exposure scenarios were evaluated for consumption of fish from the LDW site. Because site-specific consumption rates for fish caught from the LDW study area were lacking, surrogate rates were taken from a recent survey of the Suquamish Tribe using data gathered from fish consumers only, and a study of recreational anglers in urban embayments of Puget Sound. Exposure assumptions given in Table C1 below were used with Equations C1 and C2 to estimate contaminant doses associated with fish consumption.

a = consumption rates from Landolt et al were reported as g/day during the fishing season. A body weight of 72 kg was used to convert the units to g/kg/day.

Contaminant concentrations used to estimate exposure to contaminants in Duwamish River fish are given below in Table C2. These concentrations are based on data compiled from various sources as part of the scoping-phase risk assessment in preparation for the Remedial Investigation. For finfish, only data generated from analysis of fillets without skin were used. Crab data included analytical results from muscle tissue only.

A majority of the finfish samples were analyzed as composite samples meaning that several fish were collected and blended together in order to obtain an approximation of the mean contaminant concentration in a group of fish. This method reduces the cost of environmental analyses, but decreases the knowledge about the variability of contaminant levels in a fish population. This health assessment relied on both composite and individual samples. As a result, a weighting method was used to calculate mean contaminant concentrations in finfish. Estimates of variance around the mean were also calculated in order to estimate confidence intervals around the mean. The equations and examples are shown below.

Weighted mean (grouped mean)

The weighted mean is equal to the sum of composite sample concentration times the number of fish per composite divided by the sum of fish per composite in all samples.

For example:

The weighted mean concentration of the following data set is shown below.

Weighted mean concentration
= ( 2*100 + 8*52 + 10*48) / 20
= 54.8

Grouped Variance

Using the above example

Grouped variance ={ (2*(100)² + 8*(52)² + 10*(48)²) - (((2*100)² + (8*52)² + (10*48)² )/20)}/19

The 95 UCL was calculated for groups of composited fish samples using the weighted mean and the grouped variance. In general, composite sampling is designed to get a better approximation of the mean, therefore, the 95 UCL was not much greater than the weighted mean.

a = Arsenic concentrations are given as total arsenic. Ten percent of this value was used in dose calculations to represent the amount of inorganic arsenic to be consistent with EPA's RfD and cancer slope factor both of which are based on exposure to inorganic arsenic.
b = Sum of Aroclors. The predominant Aroclors detected in Puget Sound fish are Aroclor 1254 and Arolcor-1260
c = Mercury concentrations are given as total mercury. All measured mercury is assumed to in the methylmercury form for comparison with EPA's RfD for methylmercury.
f = Represents maximum value detected.
g = striped perch

Note: Rounding of hazard quotients could impact the hazard index. Doses are reported in scientific notation format. For example, 1e-02 is the same as 1 x 10^-2 or 0.01.

a = Arsenic concentrations are given in units of mg/kg while PCB values are in ug/kg.
Note - Cancer risks are reported in scientific notation format. For example, 1e-02 is the same as 1 x 10^-2 or 0.01.

Note - Doses are reported in scientific notation format. For example, 1e-02 is the same as 1 x 10^-2 or 0.01.

Next Section      Table of Contents